36872 Federal Register Vol. No. Friday, July 9, 1993 ......George Jett or Ms. Debra Nicoll,...

Federal Register / Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regulations

ADDRESSES: The workshop will be heldat the EPA National Vehicle and FuelEmissions Laboratory, 2565 PlymouthRoad, Ann Arbor, Michigan 48105,telephone (313) 668-4200.FOR FURTHER INFORMATION CONTACT:William Rutledge, Technical SupportBranch, Manufacturers OperationsDivision (6405J), U.S. EnvironmentalProtection Agency, 401 M Street SW.,Washington, DC 20460. Telephone:(202) 233-9297.

SUPPLEMENTARY INFORMATION:

I. BackgroundOn April 21, 1993, EPA published

final Retrofit/Rebuild Requirements for1993 and Earlier Model Year UrbanBuses (58 FR 21359). The retrofit/rebuild program, intended to reduce theambient levels of particulate matter(PM) in urban areas, is limited to 1993and earlier model year (MY) urbanbuses operating in metropolitan areaswith 1980 populations of 750,000 ormore, whose engines are rebuilt orreplaced after January 1, 1995.Operators of the affected buses arerequired to choose between twocompliance options: Option 1 setsparticulate matter emissionsrequirements for each urban bus in anoperator's fleet whose engine is rebuiltor replaced; Option 2 is a fleet averagingprogram that sets out a specific annualtarget level for average PM emissionsfrom 1993 and earlier MY urban busesin an operator's fleet.

A key aspect of the program Is thecertification of retrofit/rebuildequipment. Emissions requirementsunder either of the two options dependheavily on the availability of retrofit/rebuild equipment certified for eachengine model. To be used for Option 1,equipment must be certified as meetinga 0.10 g/bhp-hr PM standard or asachieving a 25 percent reduction in PM.Equipment used for Option 2 must becertified as providing some level of PMreduction that would in turn be claimedby urban bus operators when calculatingtheir average fleet PM levels attainedunder the program. Technology must becertified in order for urban busoperators to take credit for a reductionin PM provided.

The certification process outlined inthe retrofit/rebuild rule is based onexisting regulations for aftermarket partscertification for light-duty vehicles andtrucks. However, in order for equipmentto be a trigger under Option I for anengine model, additional informationregarding cost must be submitted in theapplication for certification. Thecertifier must guarantee that theequipment will be offered for sale under

an appropriate life cycle cost ceiling toall affected urban bus operators forwhich the equipment is certified to bea trigger under Option 1.

In order to provide sufficient leadtime for urban bus operators to makeplans and procure retrofit/rebuildequipment, it is important to havecertified equipment available to urbanbus operators as soon as possible. Forthis reason, and because most potentialcertifiers have had no previousexperience with the similar aftermarketparts certification program or the lifecycle cost requirements, EPA feels itwould be worthwhile to provide andexplain standardized formats formeeting the rebuild/retrofit equipmentcertification requirements as quicklyand efficiently as possible.

II. Workshop StructureEPA will arrange the workshop

agenda to provide an overview of theretrofit/rebuild program followed by adetailed explanation of the equipmentcertification program and thestandardized certification applicationformats. Ample time will be allowed forquestions. Specific written questions orsuggestions are encouraged to beprovided in advance to the ContactPerson. Further, all those planning toattend are requested to notify theContact Person in advance to facilitateour planning.Robet D. Brenner,Acting Assistant Administratorfor Air andRadiation.(FR Doc. 93-16312 Filed 7-8-93; 8:45 am!elLUNG CODE S04-

40 CFR Part 414[FRL-4610-7]RIN 2040--AB65

Organic Chemicals, Plastics andSynthetic Fibers Category; EffluentUmitations Guidelines, PretreatmentStandards, and New SourcePerformance Standards

AGENCY: Environmental ProtectionAgency (EPA).ACTION: Final rule.

SUMMARY: EPA is promulgatingamendments limiting effluentdischarges to waters of the United Statesand the introduction of pollutants intopublicly owned treatment works(POTWs) by existing and new sources inthe organic chemicals, plastics, andsynthetic fibers (OCPSF) point sourcecategory.

EPA is adding Subpart J limitationsbased on the Best Available Technology

Economically Achievable (BAT) andNew Source Performance Standards(NSPS) for 19 additional pollutants aswell as Pretreatment Standards forExisting Sources (PSES) andPretreatment Standards for New Sources(PSNS) for 11 of these 19 pollutants.These amendments respond to the U. S.Fifth Circuit Court of Appeals' remanddecisions on the OCPSF regulation,Chemical Manufacturers Association v.U.S. EPA.

EPA is also correcting the criteria fordesignating "metal-" and "cyanide-bearing" waste streams, and is adoptingtwo nonsubstantive formatting changes.DATES: These regulations are effectiveAugust 23, 1993. In accordance with 40CFR 23.2, this regulation shall beconsidered promulgated for purposes ofjudicial review at I p.m. Eastern timeJuly 23, 1993.

The compliance date for PSES is assoon as possible but no later than July23, 1996. The compliance dates forNSPS and PSNS is the date the newsource begins operation. Deadlines forcompliance with BAT are established Inpermits.

Under section 509(b)(1) of the CleanWater Act, judicial review of thisregulation can be had only by filing apetition for review in the United StatesCourt of Appeals within 120 days afterthe promulgation date of today'sregulation for purposes of judicialreview. Under section 509(b)(2) of theClean Water Act, the requirements inthis regulation may not be challengedlater in civil or criminal proceedingsbrought by EPA to enforce theserequirements.ADDRESSES: The supporting informationand all comments and responses on thisamendment to 40 CFR part 414 will beavailable for inspection and copying atthe EPA Water Docket, room L-102,Waterside Mall, 401 M Street SW.,Washington, DC 20460. For access toDocket materials, call (202) 260-3027between 9 a.m. and 3:30 p.m. for anappointment. The basis for thisamendment is detailed in thesupplement to the OCPSF record whichis also in the Water Docket. Foradditional information contact GeorgeM. Jett, Project Officer, ChemicalsBranch, Engineering and AnalysisDivision (WH-552), EnvironmentalProtection Agency. 401 M Street, SW.,Washington, DC 20460.FOR FURTHER INFORMATION CONTACT:George M. Jett, (202) 260-7151, forinformation regarding the technicaldata, and Dellra Nicoll, (202) 260-5386for information regarding the economicdata. Copies of the supplementaldevelopment document and

36872


supplemental economic analysis may beobtained by writing or calling Mr.George Jett or Ms. Debra Nicoll,respectively, Engineering and AnalysisDivision (WH-552), U.S. EPA, 401 MStreet SW., Washington, DC 20460.SUPPLEMENTARY INFORMATION:

Organization of this Document:

I. Legal AuthorityII. Background and Rationale for

AmendmentsIll. Now Source Performance Standards and

Pretreatment Standards for New SourcesIV. BAT SubcategorizationV. BAT Subpart J and Corresponding

AmendmentsA. BackgroundB. Final Regulatory AmendmentsC. Basis for Economic AnalysisD. Summary of Economic ImpactsE. Small Plant AnalysisF. Cost-Effectiveness Analysis

VI. Correction of Criteria for Designating"Metal-" and "Cyanide-Bearing" WasteStreams

VII. Nonsubstantive Format ChangesVIII. Public Participation and Summary of

Selected ResponsesA. Scope of the RemandB. Appropriate Technology Basis for New

Source StandardsC. Appropriate BAT SubcategorizationD. Applicability of the Revised Pass-

Through MethodologyE. Land AvailabilityF. Guidance for Laboratory Analysis of

Complex MatricesG. Guidance for the Appropriate Flow

Basis for Converting Concentrations intoMass-Based Limitations and Standards

IX. Executive Order 12291X. Regulatory Flexibility AnalysisXI. Paperwork Reduction Act

I. Legal Authority

The amendments to 40 CFR part 414described in this notice are promulgatedunder authority of sections 301, 304,306, 307, 308, and 501 of the CleanWater Act (the Federal Water PollutionControl Act Amendments of 1972, asamended (33 U.S.C. 1251 et seq.)), alsoreferred to as "the Act" or "CWA."

II. Background and Rationale forAmendments

EPA's explanation of the backgroundand rationale for today's amendmentsare contained in the December 6, 1991proposal (56 FR 63897), the January 21,1992 extension of comment period andcorrection notice (57 FR 2238), and theDecember 1, 1992 notice of availabilityand request for comments (NOA) (57 FR56883), as supplemented by theinformation and references in thefollowing sections of this preamble andby the rulemaking record.

Briefly, these amendments respond tothe U.S. Fifth Circuit Court of Appeals'remand decisions on the November 5,1987 OCPSF regulation, Chemical

Manufacturers Association v. U.S. EPA,870 F.2d 177 (5th Cir.), modified, 885F.2d 253 (5th Cir. 1989), cert. denied,PPG Industries, Inc. v. U.S. EPA, 495U.S. 910 (1990). The Court remandedthree aspects of the OCPSF guideline:the New Source Performance Standards(NSPS) and the Pretreatment Standardsfor New Sources (PSNS) forconsideration of whether zero dischargelimits would be appropriate for newplants in the OCPSF industry based onrecycle of wastewater; thesubcategorization of the industry intotwo subcategories imposing differinglimitations based on Best AvailableTechnology Economically Achievable(BAT), on the ground that the Agencydid not provide sufficient notice of thescheme; and limitations for 19 of the 20BAT Subpart J pollutants that werebased upon in-plant biological treatmenttechnology (and the corresponding NewSource Performance Standards (NSPS)for these pollutants, as well as 13corresponding Pretreatment Standardsfor Existing Sources and PretreatmentStandards for New Sources (PSES andPSNS respectively)), on the ground thatthe model treatment systems used toestimate the cost of compliance hadshorter detention times than the systemson which the limitations were based.

In reconsidering new sourceperformance standards, EPA hasdecided not to revise the NSPS andPSNS standards that were promulgatedin the OCPSF guideline because, amongother things, EPA's database does notdemonstrate that total recycle is ademonstrated technology.

EPA also has decide dnot to revise theBAT subcategorization scheme forSubpart I and Subpart J. The Agencyconcluded that this is the mostappropriate approach for the OCPSFindustry.

EPA is promulgating the samenumerical effluent limitations andstandards that were proposed onDecember 6, 1991 for the 19 remandedBAT Subpart J and NSPS pollutants andfor 11 of the 13 corresponding PSES and

' PSNS pollutants based on revisedestimates for the cost of compliancederived from revised model in-plantbiological treatment system designs.Pretreatment standards for phenol and2,4-dimethylphenol are not beingpromulgated because, based on therevised pass-through methodologypresented in the December 1, 1992NOA, EPA has concluded they do notpass through POTWs.

These amendments also correct thecriteria for designating "metal-" and"cyanide-bearing" waste streams, andadopt two nonsubstantive formattingchanges. These actions do not arise out

of the litigation; rather, they result fromindependent EPA review of theregulation.I. New Source Performance Standards

and Pretreatment Standards for NewSources

In the 1987 OCPSF promulgation, theAgency promulgated NSPS for all directdischarging sources based on the bestavailable demonstrated technology, asrequired by CWA section 306 (52 FR at42545). NSPS was established for thethree conventional pollutants regulatedunder the OCPSF guideline on the basisof BPT model treatment technology, andfor the 63 OCPSF-regulated prioritypollutants on the basis of BAT modeltreatment technology. The numericalstandards are equivalent to the BPT andthe BAT limitations (52 FR 42545). EPAalso promulgated PSNS on the sametechnology basis as PSES; the numericalstandards for 47 priority pollutants thatwere determined to pass through orotherwise interfere with the operation ofpublicly owned treatment works(POTWs) are equivalent to the PSESstandards (52 FR 42549).

The Natural Resources DefenseCouncil (NRDC) challenged the finalNSPS and PSNS standards arguing, inpart, that the Agency failed to giveadequate consideration to betterpollution control technologies thatcould be used by new sources.

On March 30, 1989, the Fifth Circuitrejected all but one of NRDC'schallenges to the NSPS standards andremanded the NSPS standards to EPA"for consideration of whether zerodischarge limits would be appropriatefor new plants in the OCPSF industrybecause of the existence of recycling"(870 F.2d at 264). However, the Courtleft the standards in place during theAgency's response to the remand (870F.2d at 266).

The Agency has reconsidered theissues related to establishing new sourcezero discharge standards based onprocess wastewater recycle and, asproposed, has decided not to revise theexisting NSPS and PSNS standards. EPAreceived comments from NRDC urgingEPA to promulgate zero dischargestandards based on recycle of processwastewater, and numerous industrycomments supporting EPA's proposal toretain the existing NSPS and PSNSstandards. As explained more fully inSection VIII.B., below, the Agency hasconcluded that it has no basis to imposea zero discharge technology-based NSPSstandard on any OCPSF source, andthat, even if it were to undertake anextensive data collection and technicaldevelopment effort, it is unlikely EPAcould impose a zero discharge standard

36873


on more than a few of the 25,000product/processes in the OCPSFindustry. First, the "concentration-based" approach which forms theframework of the OCPSF guidelinelimits the opportunities for thepromotion of recycling and re-use ofwastewater through a nationalguideline, in contrast to the "mass-based" approach adopted in otherguidelines. The Agency explicitlyrecognized this limitation duringtheguideline development process, butopted for this approach nonethelessbecause it provided the basis for aguideline with more expansivecoverage. This was a rational regulatorydecision made by the Agency.Moreover, because the OCPSF recordwas imprecise with regard to its use ofthe term "recycle," both NRDC and theFifth Circuit in its remand ordermisinterpreted the support in thedatabase for zero discharge throughrecycling. In fact, the record containsvery few reports of complete recycle anddoes not demonstrate that recycle is ademonstrated technology on which EPAcan base a zero discharge standard.

IV. BAT SubcategorizationThe original OCPSF guideline had

two technology-based BATsubcategories for the control of toxicpollutants: one for any direct discharge

oint source that uses end-of-pipeiological treatment or installs end-of-

pipe biological treatment to complywith BPT effluent limitations (Subpart I,§ 414.90), and one for any directdischarge point source that does not useend-of-pipe biological treatment or doesnot install end-of-pipe biologicaltreatment to comply with BPT effluentlimitations (Subpart J, § 414.100).Subparts I and J set limits for 63 and 59pollutants, respectively. Of the 59Subpart J Maximum for MonthlyAverage limitations, 9 are Identical to,20 are more stringent than, and 30 areless stringent than the correspondingSubpart I limitations.

As explained in the proposal, EPAestablished this scheme based, in part,on its conclusion that there are plants inthe OCPSF industry whose wastewatershave such low levels of BiochemicalOxygen Demand (BOD) that they willnot be able to operate biologicaltreatment systems effectively and do notneed biological treatment systems tocomply with the BPT BOD effluentlevels (56 FR at 63899). Biologicaltreatment systems rely onmicroorganisms to biodegrade or "eat"the organic pollutants in thewastewater. BOD, a measure of theorganic pollution strength in water orwastewater, is determined by measuring

the oxygen used by microorganisms to* oxidize or "eat" the organiccontaminants of a sample.Consequently, BOD measures theamount of substrate or "food" availablefor the survival of microorganisms (id.).Biological treatment systems thereforerequire sufficient BOD levels to operate(id.).

NRDC challenged the BATsubcategorization scheme in thelitigation over the OCPSF guideline,arguing that the Agency had failed topresent its BAT subcategorizationscheme for comment and also assertingthat this type of BAT subcategorizationviolated the CWA because It allowed adischarger who chooses not to employend-of-pipe bi6logical treatment to besubject to fewer and less stringent BATSubcategory J limitations, rather thanthe more stringent Subcategory Ilimitations which apply to plants withend-of-pipe biological treatmentsystems. NRDC also argued that, if ithad had an opportunity to comment, itwould have urged EPA to establish araw waste BOD "floor" above whichplants would not be able to qualify forSubpart J, or to limit the applicability ofSubpart J to those categories of OCPSFproduction that tend to have low rawwaste BOD levels (NRDC 6/30/88 Briefat 54).

On March 30, 1989, the Fifth CircuitCourt of Appeals, without ruling onNRDC's substantive arguments,remanded the BAT subcategorization ofthe industry for notice-and-commentproceedings. The Court left the schemein effect pending further rulemaking,reasoning in part that the notice-and-comment proceedings may disclose thatthe BOD floor urged by NRDC is neithernecessary nor feasible (870 F.2d at 236).

The Agency has reconsidered theissues related to revising the BATsubcategorization scheme or otherwiselimiting the applicability of Subpart Jand has decided not to revise theexisting scheme. The schemeaccommodates the complexity of theIndustry and encourages source controland rational waste managementdecisions. In addition, EPA does notbelieve revision of the scheme isnecessary. Plants must comply with lowBPT limits, and plants that need toachieve significant BOD reductions willgenerally install biological treatment

* because other treatment alternatives aresignificantly more expensive. EPA doesnot believe plants' treatment decisionswill be motivated by the desire to besubject to Subpart J. In any event,Subpart J is not significantly lessstringent than Subpart L

Moreover, the Agency does not havea technical basis to determine which

plants can sustain biological treatmentecause of the lack of a theoretical BOD

floor for sustaining biological treatmentand the great variability of OCPSFproduction and wastewatercharacteristics. For these reasons, asexplained more fully in response toNRDC's comments, below, in the 1991proposal, and in the rulemaking record,the Agency his decided not to establisha BOD floor or otherwise limit theapplicability of Subpart J.V. BAT Subpart I and CorrespondingAmendmentsA. Background

In the 1987 OCPSF guideline, EPApromulgated toxic pollutant effluentlimitations based on the twosubcategory scheme described inSection IV above. Subpart J establisheddirect discharge toxic pollutantlimitations for plants that comply withBPT limitations without the use of end-of-pipe biological treatment or contracthauling. The Subpart J toxic pollutantnumerical limitations were based on theperformance of in-plant wastewatertreatment technology including steamstripping to remove volatile prioritypollutants, chemical precipitation formetals, alkaline chlorination forcyanide, and in-plant biologicaltreatment for removal of selectedpriority pollutants includingpolynuclear aromatics, phthalate esters,acrylonitrile, phenol, and 2,4-dimethylphenol (52 FR 42538--45;Final "Development Document forEffluent Limitations Guidelines andStandards for the Organic Chemicals,Plastics, and Synthetic Fibers PointSource Category," (EPA 440/1-87/009),October 1987 (hereafter referred to as1987 DD), Vol. 1, pp U1-8 to 11).

Numerical standards for 20 of theSubpart J pollutants were based on theperformance of three biologicaltreatment systems with detention timesof 1.6, 3.5, and 17.2 days. In contrast,detention times between I and 2.1 dayswere used to estimate the costs ofcompliance based on the model in-plantbiological treatment systems (1987 DD,p VII-189; R.93970-4020; EPA 9/23188Response Brief at 244-59).

The Chemical ManufacturersAssociation (CMA) challenged theSubpart J limitations based on in-plantbiological treatment arguing, in part,that the plants used by EPA to derivethe limitations based on in-plantbiological treatment have more.treatment in place than EPA's modeltreatment technology used to estimatecosts of compliance and that EPAtherefore significantly underestimatedthe costs of installing in-plant biological

36874


treatment (CMA's 4/25/88 Brief at 58-76).

After the Fifth Circuit initially upheldthese Subpart J limitations (870 F.2d at240-42), CMA petitioned forreconsideration, again arguing, in part,that the Agency underestimated thecosts of compliance due to thedifferences between the detention timesof the plants that provided the basis forthe numerical standards and thedetention times of the model technologythat provided the basis for estimatingthe engineering costs of compliance(0A's 5/3/89 Brief on Petition forRehearing, pp 8-11). The Courtconcluded that the detention time wasa key variable in determining theeffectiveness of biological treatment andthat EPA had failed to demonstrate areasonable basis to conclude thatbiological systems with a 1 or 2.1 daydetention time would control pollutantsas effectively as the biological systemswith the 3.5 and 17.2-day detentiontimes (885 F.2d at 265).

The Court remanded limitations forthe Subpart J pollutants based on thetwo plants with these longer detentiontimes. In a June 29, 1990 revocationnotice (55 FR 26691), the Agencywithdrew the BAT limits for the 19 ofthe 20 Subpart J limits that were basedon these two plants. EPA left in effectthe limitations for acrylonitrile, whichwere based upon the treatment systemwith the 1.6 day detention time. In thisnotice EPA also withdrew the 19corresponding NSPS standards, and the13 corresponding PSES and PSNSstandards that were based on theremanded Subpart J limits.

The remand was based on thediscrepancy between the detentiontimes of the systems that provided thetechnical basis for the Subpart J limitsand the detention times of the costedmodel in-plant systems, and not on thetechnical achievability of the limitsgenerally. EPA therefore proposed onDecember 6, 1991 and January 21, 1992the same numerical standards withrevised estimates of costs of compliance.The revised compliance costs werebased on revised model in-plantbiological treatment systems withincreased detention times as a functionof reported or projected raw waste toxicpollutant concentrations.

A large number of the comments onthe proposal challenged EPA'sdetermination irr the original 1987OCPSF promulgation that phenol--oneof the 13 pollutants for whichpretreatment standards wereremanded-passes through POTWs.Several comments raised the same issuewith respect to 2,4-dimethylphanol-another of the 13 pollutants. Based on

EPA's assessment that these commentshad merit, EPA announced in a noticeof availability (NOA) published in theFederal Register on December 1, 1992,that it was considering revising itsdetermination that phenol and 2,4-dimethylphenol pass through POTWs,based on a proposed modification to theAgency's traditional pass-throughmethodology (57 FR 56883). The revisedmethodology as proposed appliedscientific and engineering judgment inconjunction with biological treatmentperformance data to determine thatphenol and 2,4-dimethylphenol do notpass through POTWs.

EPA collected additional POTWphenol removal data and reviewed it inconjunction with the data that EPA usedin the 1987 pass through analysis, andperformed a chemical and engineeringassessment of the fate of phenol and 2,4-dimethylphenol in biological treatmentsystems. EPA has concluded that thesepollutants are highly biodegradable andthat the removals of these pollutantsachieved by POTWs are essentiallyequivalent to those achieved by directdischargers. In addition, since phenoland 2,4-dimethylphenol are lowvolatility pollutants, the removalsachieved by POTWs do not simplyresult from the transfer of the pollutantsto the air.

Based on these conclusions, today'samendments are based on revisedengineering costs of compliance andpollutant loading reductions for 11 ofthe 13 remanded pollutants. Finalpretreatment standards for phenol and2,4-dimethylphenol are not beingpromulgated today because the Agencyhas concluded they do not pass throughPOTWs.

B. Final Regulatory Amendments

As explained above, EPA is addingPSES and PSNS standards for 11additional pollutants in the tableappearing in the new § 414.111 (seeSection VII). The 11 pollutants areacenaphthene, anthracene, bis(2-ethylhexyl) phthalate, di-n-butylphthalate, diethyl phthalate, dimethylphthalate, fluoranthene, fluorene,naphthalene, phenanthrene, and pyrene.

EPA is adding BAT and NSPSlimitations and standards for 19additional pollutants in the table in§ 414.101. The 19 pollutants areacenaphthylene, benzo(a)anthracene,benzo(a)pyrene, 3,4-benzofluoranthene,benzo(k)fluoranthene, chrysene, 2,4-dimethylphenol, phenol, and the 11pollutants listed in the previousparagraph.

C. Basis for Economic Analysis

The economic analysis of today's finalBAT and PSES limitations is based onrevised compliance costs for the sameBAT numerical limitations and, withthe exception of phenol and 2,4-dimethylphenol, for the same PSESnumerical standards that werepromulgated in 1987 and proposed inDecember 1991. Phenol and 2,4-dimethylphenol are not regulated underPSES because EPA has determined thatthey do not pass through POTWs (seeSection V.A.).

EPA has revised its costing analysis inresponse to comments by CMA that theAgency underestimated the costs ofcompliance because the technical database was not revised to reflect theinformation gathered in EPA's April1991 survey of the 84 direct dischargeplants that did not report the use of end-of-pipe biological treatment in theoriginal OCPSF CWA Section 308survey. The principal changes based onthe April 1991 survey reflect the shift indischarge status for 14 plants fromdirect to indirect discharge and therevised projection that 47 rather than 23plants are subject to the BAT Subpart Jlimitations (May 1993 Supplement tothe Development Document for EffluentLimitations Guidelines and Standardsfor the Organic Chemicals, Plastics andSynthetic Fibers Point Source Category,(EPA 821-R-93-007)). These changesare reflected in the economic analysisperformed by EPA, described below.The changes do not materially affect.EPA's analysis, nor do they affect thedetermination that today's rule iseconomically achievable.

In response to the remand of theSubpart J limitations, EPA consideredwhether the limitations areeconomically achievable given therevised compliance costs. The Agency'sanalysis of the revised costs parallelsthe economic analysis conducted for the1987 rulemaking. The methodology forthat analysis was described in thepreamble to the 1987 final rule (52 FR42550) and in the Agency's economicimpact analysis that was published insupport of that rule ("Economic ImpactAnalysis of Effluent LimitationsGuidelines and Standards for theOrganic Chemicals, Plastics andSynthetic Fibers Industry," (EPA 440/2/87-007), September 1987). Theeconomic analysis for today's filial ruleis documented in a report: "Re-Evaluation of the Economic ImpactAnalysis of Effluent LimitationsGuidelines for the Organic Chemicals,Plastics, and Synthetic Fibers Industry,"May 1993. This report is available from

36875


EPA; see the contacts identified at thebegnning of this notice.PA undertook a revised plant impEact

analysis, a revised regulatory flexibii tyanalysis, and a revised cost-effectiveness analysis to evaluate theeffects of the final compliance costs.The plant impact analysis is the primarybasis for evaluating economicachievability. The regulatory flexibilityanalysis provides information todetermine whether small plants aredisproportionately affected by therevised costs. The cost-effectivenessanalysis provides information about therelative efficiency of the control optionselected for reducing pollutantdischarges.

The methodology for assessing plantimpacts Is the same as was used for the1987 final rule. The Impact of thecompliance costs on OCPSF plants wasevaluated using the following criteria:total annualized cost of the treatmenttechnology, potential plant closures,potential product line closures,significant sales or profit impacts, andthe job losses associated with closures.Additional Information regarding thecalculation of these impact measuresand their significance is found in theeconomic impact analysis prepared forthe 1987 final rule.

D. Summary of Economic ImpactsThe costs used to evaluate today's

final rule are based on the capital andannual operating costs of the modeltreatment technology, as described Inthe May 1993 final "Supplement to theDevelopment Document for EffluentLimitations Guidelines and Standardsfor the Organic Chemicals, Plastics andSynthetic Fibers Point SourceCategory." In the following discussion,all cost estimates are reported in 1986dollars to facilitate comparisons to the1987 promulgated rule. Cost estimatesin that preamble were reported in 1986dollars.

For approximately 265 directdischargers for which EPA hasestimated compliance costs, the totalannualized costs of the OCPSFguideline, as amended, are $231.1million. This is a $6.9 million (3.1percent) increase over the cost of therule as promulgated in 1987. However,today's rule results in fewer incrementalImpacts (i.e., plant closures, productline closures, or employment losses)than those estimated for the 1987promulgation. Two of the plants thatwere projected to close under BAT in1987 were among the 14 plants thatswitched their discharge status toindirect and are now projected to closeas indirect discharge plants undertoday's rule. The Agency estimates that

9 OCPSF direct discharging plants orproduct lines may close as a result ofthe compliance costs imposed bytoday's rule. These closures represent 3percent of all direct discharging plants.For the 1987 rule and the 1991 proposal,the Agency estimated that 11 OCPSFplants or product lines would close asa result of the compliance costs. Theemployment reduction associated withthe closures for today's rule is 1,060 (0.6percent of total.OCPSF employment).

For approximately 380 indirectdischargers for which EPA hasestimated compliance costs, the totalannualized costs of the OCPSFguideline, as amended, are $254.4million. This is a $50.1 million (24.5percent) increase over the cost of therule as promulgated in 1987. Today'sfinal cost estimate is $7.0 million (2.8percent) higher than the costs estimatedfor the December 1991 proposed rule;the net change in estimated costs reflectthe shift in discharge status for 14 plantsthat were direct discharge facilities in1987 but now discharge to POTWs, andthe exclusion of phenol and 2,4-dimethylphenol in the finalpretreatment standards. Overall, plantmpacts for indirect dischargers increase

slightly when compared to the 1987 and1991 results. Sixty (60) plant andproduct line closures are projected toresult from compliance with the OCPSFguideline as amended by today's rule.As originally promulgated, the numberof plant and product line closures was52. For the 1991 proposal, the estimateof plant and product line closures was56. (Two of the additional four closuresover the 1991 estimate result from theswitch to indirect discharge status oftwo direct discharging plants that wereprojected to close.) When expressed inpercentage terms (relative to the numberof indirect dischargers), the closure ratefor today's final pretreatment standardsIs 16 percent; at promulgation in 1987,the closure rate for indirect dischargerswas 14 percent. The employmentreduction associated with the closuresfor today's rule is 2,946, which is anIncrease (of 0.4 percent of total OCPSFemployment) from 2,190 at the 1987promulgation.

The Agency finds that the Impactsimposed by the revised compliancecosts for both BAT and PSES are notsignificantly different from the impactsprojected in 1987, and that today'samendment is economically achievable.

In addition, EPA received commentsthat its approach understates the actualcost of compliance with today's rule.EPA disagrees and believes the costs ithas presented are an accurate industry-wide estimation of compliance costs.However, EPA has conducted two

sensitivity analyses examining theprojected impacts of increasedcompliance costs with today's rule, onewhich doubled the projected costs ofcompliance, and one based onassignment to individual plants ofadditional treatment unit operationsthat commenters claimed would berequired (see Response to CommentsSection of the Public Record). Thesehighly conservative sensitivity analysesproject one additional and fouradditional plant closures, respectively,which EPA finds would still beeconomically achievable. EPA believesthat the closures estimated using thesensitivity analyses overstate the actualclosures that will result fromcompliance with today's rule.

E. Small Plant AnalysisA regulatory flexibility analysis

addresses the burden of regulatoryactions on small entities. For today'sfinal regulation, as in the 1987 finalrule, the regulatory flexibility analysisexamined whether small plants, asdefined by a plant production thresholdof 5 million pounds, weredisproportionately affected by theregulation. The assessment includesconsideration of plant and product lineclosures and profit and sales impacts.The assessment reflects the shift indischarge status for 14 plants, therevised compliance costs, and, forindirect discharge plants, the exclusionof phenol and 2,4-dimethylphenol fromthe PSES standards, as explained inSection V.C., above.

In the 1987 final rule, based on itssmall plant analysis, EPA set BAT equalto BPT for plants whose annual OCPSFproduction is less than or equal to 5million pounds (52 FR 42539); EPA didnot establish different PSES for anysector of indirect dischargers (52 FR42548).

At promulgation in 1987, the Agencyprojected that 79 percent of the smalldirect discharging plants would beaffected by plant or product lineclosures or profit or sales impacts; fortoday's rule, 77 percent of the smalldirect discharging plants are projectedto be affected.

At promulgation in 1987, 61 percentof the small indirect discharging plantswere projected to be affected; for today'srule, 63 percent of the small indirectdischarging plants are projected to beaffected by plant or product lineclosures or profit or sales impacts.

These plant impacts on small directand indirect dischargers are notsignificantly different from the impactsevaluated for the 1987 final rule, andthe basis for establishing BAT andPSES, as presented in the 1987 final

36876

Federal Register I Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regulations

rule, is unchanged by the revisedeconomic analysis. Thus, there is nochange in the small plant analysisfindings.F. Cost-Effectiveness Analysis

EPA conducted a cost-effectivenessanalysis for the 1987 final rule andreported the results in the preamble andin supporting documents. EPA's cost-effectiveness analysis compares theincremental cost of a control option (in1981 dollars) to the pounds ofpollutants removed by the controloption, where those pounds areweighted by their relative toxicity.Additional descriptions of the cost-effectiveness methodology are found inthe preamble to the 1987 final rule (52FR 42552) and in a document includedin the administrative record for thatrule: "Cost-Effectiveness Analysis forthe Organic Chemicals, Plastics, andSynthetic Fibers Industry," September1987 (R. Sec. VI-11, pp 5155 to 98).

For today's final rule, EPArecalculated the cost-effectiveness ratiosfor BAT and PSES using the revisedcompliance costs and revised pollutantremoval data to account for theexclusion of phenol and 2,4-dimethylphenol In the finalpretreatment standards. The cost-effectiveness of the OCPSF BATlimitations, as amended, is $4 perpound equivalent removed. The cost-effectiveness of the OCPSF PSESstandards, as amended, is $39 perpound equivalent removed.

The cost-effectiveness of the BATlimitations as amended is virtually thesame as the result reported for the 1987final rule. The notice of proposedrulemaking in December 1991 statedthat as part of the assessment forpublishing the final rule, EPA wouldconsider the cost-effectiveness ratio ofBAT Subpart J (56 FR 63905). The resultof that calculation is a cost-effectivenessratio of $6 per pound equivalentremoved for the final BAT limitationsfor Subpart J direct dischargers. Thecost-effectiveness ratio for Subpart Idirect dischargers is, when examinedseparately, $4 per pound equivalentremoved.

The cost-effectiveness ratio of PSES,as amended, is $39 per poundequivalent removed; this result iscomparable to the $38 cost-effectivenessratio reported in the December 1991proposal. The result reported for the1987 final rule was $34 per poundequivalent removed. The cost-effectiveness of PSES as amended is not-significantly different from the cost-effectiveness of PSES as originallypromulgated.

VI. Correction of Criteria forDesignating "Metal-" and "Cyanide-Bearing" Wase Streams

To control chromium, copper, lead,nickel, zinc, and total cyanide, the 1987OCPSF guideline establishedconcentration-based limitations thatapply only to metal-bearing or cyanide-bearing waste streams. Other wastestreams have a zero discharge allowancefor these pollutants. EPA listed theproduct/processes considered to havemetal-bearing or cyanide-bearingprocess wastewater in Appendix A ofthe regulation. However, EPArecognized that at some sites processwastewaters not listed In Appendix Amay contain significant levels of metalsor cyanide. In such cases, EPA intendedfor the regulations to authorize thepermit writer or control authority todesignate such waste streams as "metal-bearing" or "cyanide-bearing" on a case-by-case basis and to apply theconcentration limitations set forth in theregulation to these waste streams.

The 1987 final regulation includedlanguage intended to require separatetreatment of Appendix A waste streamsand waste streams designated as metal-bearing or cyanide-bearing on a case-by-case basis, unless combination of suchwaste streams prior to treatment wouldresult in substantial reduction of themetals or cyanide contained in thewaste streams. The requirement wasintended to prevent facilities fromcombining waste streams containingdifferent metals prior to treatment,thereby complying, in whole or in part,with the promulgated limits throughdilution rather than through treatment.However, as promulgated, the regulationIncorrectly provided that a non-Appendix A waste stream, even if itcontained significant quantities ofmetals or cyanide, could not bedesignated as metal-bearing or cyanide-bearing unless the combination of suchwaste stream with an Appendix A wastestream prior to treatment would resultin substantial reduction of thepollutants. In other words, theprohibition against combined treatmentinadvertently appeared as anindependent restriction on thedesignation of non-Appendix A wastestreams as "metal-bearing" or "cyanide-bearing," rather than a restriction onhow those streams must be treated oncethey have been determined to be metal-or cyanide-bearing. The Agency did notintend to narrow the scope of wastestreams that can be designated as metal-or cyanide-bearing, and intends that anywaste streams with significant levels ofmetals or cyanide should be regulated asmetal-bearing or cyanide-bearing,

respectively. The Agency is adoptingthe proposed revision that correctlyreflects EPA's intent to restrictcombined treatment, rather than tonarrow the scope of metal-bearing andcyanide-bearing waste streams.

VII. Nonsubstantive Format Changes

The Agency is adopting the twononsubstantive formatting changes thatwere proposed to improve theorganization and utility of 40 CFR part414 (56 FR 63904). First, the Agency isrevising the order of the toxic pollutantlistings in the regulatory tables to listthe limitations and standards,alphabetically by pollutant name.Second, the Agency is deleting multiplelistings of the same table. With theexception of the footnotes for the zincpretreatment standards in the tables for§§ 414.25 and 414.35, each of the tablesIn §§414.25, 414.35, 414.45,414.55,414.65, 414.75, and 414.85 is identical.To consolidate the regulation, theAgency is adding Subpart K to list thepretreatment standards in one table withintroductory text and an appropriatefootnote for zinc (identical to Footnote2 to the table in § 414.101), and deletingthe tables of pretreatment standards anda portion of their introductory text from§§ 414.25, 414.35,414.45, 414.55,414.65, 414.75, and 414.85. Appropriatereferences to Subpart K are replacing thetables in §§ 414.25, 414.35,414.45,414.55, 414.65, 414.75, and 414.85. EPAis making corresponding changes to§§ 414.26, 414.36, 414.46, 414.56,414.66. 414.76, and 414.86, so that boththe PSES and PSNS regulationscorrectly refer to the consolidated table,rather than to the former, separatetables.

VIII. Public Participation and Summaryof Selected Responses

The Agency received comments from28 separate commenters on theDecember 6, 1991 proposal and January21, 1992 extension of the commentperiod. These included three tradeassociations, two POTWs, 22 individualcompanies, and NRDC. The Agencyreceived comments from 26 separatecommenters on the December 1, 1992NOA. These included four tradeassociations, four POTWs, the City ofPhiladelphia, and 17 individualcompanies.

The Agency's responses to commentsare contained in the "CommentSummary and Response" section of therulemaking docket. This sectionpresents the Agency's responses to theprincipal comments relating to theremand issues. In addition, this sectionprovides guidance in response to two

36877


general OCPSF implementation issuesraised in the public comments.

A. Scope of the Remand

Contrary to the assertions of CMA andother commenters (including ICIAmericas, Inc., the Eastman ChemicalCompany (Kodak), Union CampCorporation, and Allied-Signal, Inc.),EPA believes it correctly interpreted theFifth Circuit's remand as limited to there-costing of in-plant biologicaltreatment systems based on thedetention times used by the end-of-pipesystems on which the Subpart Jlimitations and correspondingpretreatment standards were based, andto the land availability issues associatedwith the larger systems EPA has costed(56 FR at 63903). The Court originallyupheld the limits based on in-plantbiological treatment, rejecting all ofCMA's and other petitioners' argumentsand concluding that the petitioners"have failed to demonstrate that end-of-pipe biological treatment systems aresufficiently different to make EPA'sreliance on end-of-pipe data irrational"(CMA v. EPA, 870 F.2d at 240 (emphasisadded)). As explained in Section V.A.,above, the Court remanded 19 of the 20BAT Subpart J limits on rehearing,based on the fact that the two plantsused to derive those 19 limits hadtreatment systems with detention timesthat were significantly longer than thedetention times of the model in-plantsystems costed by EPA (CMA v. EPA,885 F.2d at 265). The Court concludedthat EPA had not adequatelydemonstrated that variation of otherfeatures of a plant's biological treatmentsystem, such as the concentration ofbiodegrading organisms (MLVSS), couldcompensate for the shorter detentiontimes. On this basis, the Courtconcluded that EPA "failed todemonstrate a reasonable basis for itsconclusion that in-plant treatment caneliminate pollutants as effectively as theend-of-pipe systems of Plants 1293T and948F [the two plants with the long-detention-time treatment systems]" (id.).EPA has fully addressed the remand byrecosting the limits in question byprojecting the cost of installing in-plantbiological systems with the samedetention times as the end-of-pipesystems on which the limits were based.

The commenters argue that theCourt's remand requires a broaderexamination and demonstration of thetechnical achievability of the limits.First, they point out that the Court'sremand language was not expresslylimited to cost, but rather stated thatEPA "failed to demonstrate" that in-plant biological treatment systems "canremove pollutants as effectively as the

end-of-pipe systems of Plants 1293T and948F" (see, e.g., CMA comment at 15).The commenters thus conclude that theCourt's remand re-opened the generalissue of whether EPA rationally reliedon data from end-of-pipe biologicaltreatment systems to establish limits forwhich the model treatment technologyis in-plant biological treatment (id. at16). Commenters raise numerousarguments related to the technicalachievability of the limits generally,focusing largely on purporteddifferences between the in-plant modeltechnology and the end-of-pipe systemson which the limits are based, and onthe purported differences between in-plant and end-of-pipe waste streams(see, e.g., id. at 16-38).

The commenters' reading of theremand is wrong. The Court remandedonly the limitations based on the twoplants with the end-of-pipe treatmentsystems that had significantly longerdetention times (3.5 and 17.2 days) thanthe model systems costed by EPA. TheCourt left in place the limitations foracrylonitrile, which were based on theend-of-pipe biological treatment systemwith the shorter detention time (1.6-day)that was within the range of detentiontimes (1-2.1 days) used to estimate thecost of compliance for the modeltreatment systems (885 F.2d at 253; seealso January 3, 1990 SettlementAgreement between CMA and EPA(agreeing that remand left in placelimitations for acrylonitrile)). Had theCourt generally rejected EPA's approachof establishing limits based on end-of-pipe treatment systems, it would haveremanded all of the Subpart J limits andanalogous pretreatment standards. It isclear from the scope of the remand andthe Court's discussion that it remandedthe 19 Subpart J limits and 13 analogousPSES standards based solely on thediscrepancy in detention times betweenthe costed model treatment systems andthe two end-of-pipe systems. That is tosay, the Court concluded that EPA hadnot demonstrated the achievability ofthe limits only in the sense that EPAhad not demonstrated that the short-detention-time systems it costed couldachieve the same removals as thelonger-detention-time systems on whichit based the limits. The Court's remandbased on the limited issue of detentiontime did not otherwise negate its earlierdetermination that in-plant and end-of-pipe systems were not "sufficientlydifferent" to make EPA's approachirrational (870 F.2d at 240). By recastingthe model in-plant technology based onthe longer detention times employed atPlants 1293T and 948F, EPA has fullyaddressed the only basis for the remand.

CMA in addition argues that theCourt's remand should be interpreted asre-opening the technical achievability ofthe limits generally on the ground thatCMA's challenges to the limits incomments during the original OCPSFrulemaking and in litigation raisedissues relating to technical achievabilitygenerally (CMA Comments at 13-15).CMA points out, among other things,that it cited in litigation a number ofconcerns regarding the feasibility andeffectiveness of in-plant biologicaltreatment (id. at 13). However, the scopeof the Court's remand, of course, isdetermined by the'Court's decision, notby the issues raised by the litigants. TheCourt considered CMA's arguments andrejected them on the grounds that in-plant and end-of-pipe biological systemsare not appreciably different (270 F.2dat 240). On reconsideration, the Courtremanded the limits based solely on thediscrepancy in detention times (285F.2d at 265). It is clear that, aside fromthis single issue, the Court consideredand rejected the range of argumentsraised by CMA.

CMA and a number of othercommenters raise essentially the sameachievability arguments in thisrulemaking as they had raised in theirchallenge to the 1987 OCPSF guideline.In addition to pointing out thediscrepancy in detention times betweenthe costed and the end-of-pipe systems,CMA argued extensively before the FifthCircuit that the end-of-pipe systemsinvolved more extensive treatment(including pre- and post-biologicaltreatment and equalization for flow andconcentration) than the costed system(e.g., CMA Brief at 51-60), and that theinfluent to in-plant biological systemswould have different characteristics(specifically, that it would have higherinfluent concentrations) than theinfluent to end-of-pipe systems (CMAReply brief at 58 n.108, 62 n.13).

In addition, CMA and othercommenters raise in comments now thesame kinds of costing issues arisingfrom their technical critique as they didin their challenge to the 1987 OCPSFguideline (see, e.g., CMA Brief at 56n.94, CMA Reply Brief at 61 n.112 (EPAhas grossly underestimated cost ofcompliance and economic impactsbecause it did not cost sufficientlyextensive treatment systems)). Theseissues have been litigated and decidedin EPA's favor, and were not re-openedby the Court's remand. Rather, theissues opened by the remand arewhether EPA has accurately re-costedthe model in-plant technology to reflectthe longer detention times assigned tothe plants and whether EPA has

36878


adequately accounted for landavailability.

EPA has exhaustively addressed thetechnical and related costing issuesraised by the commenters during theOCPSF rulemaking and litigation.Nevertheless, EPA has again addressedthese issues raised by CMA and othercommenters in the response tocomments document accompanyingtoday's rule, but it does not therebyconcede that the issues raised are withinthe scope of the Court's remand orwaive its position that the issuesaddressed are not open to furtherchallenge in future litigation (seeNatural Resources Defense Council, Inc.v. Thomas, 838 F.2d 1224. 1252 (D.C.Cir. 1988)).0. Appropriate Technology Basis forNew Source Standards

NRDC in its comments challenges thesufficiency of EPA's response to theCourt's remand of NSPS. Contrary toNRDC's suggestion, EPA did take a"hard look" at the practicability ofimposing a zero discharge NSPS andconcluded that the administrativerecord fo.r the guideline does notsupport recycle as a demonstratedtechnology for the promulgation of atechnology-based NSPS zero dischargestandard within the OCPSF industry.Simply put, recycle of wastewater is nota zero discharge technology. It is a waterconservation method that may beimplemented on a plant-by-plantbasis-and is widely used in the OCPSFindustry-but rarely, if ever, results inzero discharge.

EPA thoroughly reviewed its databaseand performed a detailed technicalanalysis of the potential candidates forzero discharge standards, and identifiedonly three products--out of 25,000 inthe industry--that held the potential toachieve zero discharge based on totalrecycle of process wastewater.(Technically, even for these products.EPA does not believe that zero dischargecan be based on recycle alone since, asexplained in the proposal, wastewaterthat is vacuum stripped from thereaction vessels must be partiallyevaporated before it can be recycledback to the reaction vessel (56 FR63909)). As EPA explained, the recorddoes not disclose the means actuallyemployed to achieve zero discharge atthe plants reporting zero discharge forthese products (id.). Thus, even for thesethree products, EPA does not have atechnical basis to impose a zerodischarge NSPS.

Overall, NRDC faults EPA for notundertaking a major new data collectionand technical development effort inresponse to the remand. The commenter

argues that EPA improperly rejected theimposition of zero discharge NSPS onthe ground that it was impracticable forEPA to develop such standards, whereasthe proper inquiry, they assert, is thepracticability to the industry ofachieving zero discharge. However,NRDC ignores EPA's point that, withinthe OCPSF industry, recycle is not a"technology" that can be applied acrossa range of OCPSF production, but,rather, is the result of a range of wateruse practices and process modificationsthat are practicable to a greater or lesserextent for individual product/processes(56 FR at 63907).' Recycle can bebroadly divided between general waterconservation practices--such as recycleof vent or air scrubber water-andrecycle of water that is actually used inor generated from chemical reactions ata plant. This latter category can befurther divided into recycle withinindividual processes and recycle ofcombined waste streams back into oneor more processes (id.). Water In the firstcategory is generally used to remove asubstance from a medium, e.g., in an airpollution control device, to removehydrochloric acid from the emissionsfrom a reactor vessel. This type ofrecycle never results in zero dischargebecause the recycled water becomesprogressively more saturated with theby-products or contaminants it isremoving and must be released througha "blowdown" and replaced by cleanwater in order to perform its function(id.). The frequency of blowdown andquantity released is highly dependenton the specific process and the functionof the water.

The second general category ofrecycle will seldom, and only underplant-specific circumstances, achievezero discharge. Wastewater virtuallyalways contains contaminants that,except in very rare cases, will preventcomplete recycle (id. at 63908-09). Aswater is recycled, It becomesprogressively more contaminated and,in almost all circumstances the Agencyis aware of, will have to be released andreplaced periodically or progressively.The extent to which recycle can beemployed, and the frequency andquantity of discharge, will vary greatlyfrom one product/process to another,depending on the types and amounts of

I In this respect, EPA disagrees with industrycommenten who suggested that total recycle is notdemonstrated because it is used by only 0.012% ofthe Industry, e.g., CMA at 52. The application of atechnology by a single industry plant, or even apilot plant, might well provide a demonstratedbasis for NSPS. EPA has concluded that totalrecycle is not demonstrated not simply becausethere are so few eaimples of It. but because the fewexamples are not transferable to the manufacture ofother products.

contaminants in the wastewater, thesensitivity of the product/process tocontamination, the grade of productrequired by a manufacturer's customers(the more contaminated the water inputto a process, the lower the grade productthat will be produced), and other factors(id.)

In general, OCPSF products areproduced through chemical reactions inreaction vessels that are carefullymaintained to maximize production ofthe desired product. Chemical processesalmost never convert 100 percent of rawmaterials into the desired product; theyinevitably result in the manufacture ofby-products because there are a varietyof "reaction pathways" which result ina variety of outputs of the chemicalreaction. Depending on the nature of theby-products produced and a givenfacility's operations, the facility may beable to use some of the by-products inother processes or sell them; some of theby-products may need to be disposed of.In order to minimize production ofundesirable by-products, as well as lowgrade products and "off-spec" productsthat may require disposal, OCPSFfacilities typically "fine tune" theoperations of reaction vessels bycontrolling the purity of raw materialinputs, the physical condition of thechemical species (gas, liquid, or solidstate), method of adding and mixing thereactants, temperature, pressure, themix and quantities of solvents andcatalysts, and the configuration of the

rocess equipment. The appropriatealance of these factors wil favor a

particular reaction pathway andmaximize the conversion of the rawmaterials into the desired product, thusminimizing the amount of raw materialthat is wasted or used inefficiently bybeing converted into undesirable by-products.

The chemical reactions used toproduce the 25,000 OCPSF products areextremely complex and varied, and thecomplexity increases as the purity of theraw material inputs decreases (e.g.,through recycle of concentrated,contaminated wastewater). The additionof impurities can affect chemicalprocesses in ways that are Alifficult topredict. As explained in the proposal,the attempt to develop a zero dischargeNSPS would require a detailed studyand technical development effort on aproduct/process-by-product/processbasis to determine the feasibility andeffects of wastewater recycle. This typeof major research and developmenteffort is far beyond the scope of whatEPA can practicably accomplish, and farbeyond the inquiry into availabletechnologies that EPA has taken in pastguideline development efforts.

i I

36879


Moreover, virtually all chemicalreactions will reach a tolerance level foraccumulated contaminants in recycledwastewater and will require a dischargeat some point, unless the contaminatedwater can periodically be incorporateddirectly into the facility's product. Thiswill result in lower grade product, andthe ability of individual facilities to dothis will depend on customer demands.Since most OCPSF products are sold asintermediate products that will in turnbe part of the raw material for aconsumer product, customer demandsvary greatly, depending on the purity ofthe raw material needed for themanufacture of the consumer product.In addition, although it is possible thatindividual facilities may be able toadjust other variables in their chemicalprocesses to compensate to some extentfor increased impurity from recycledwastewater, these adjustments would behighly facility-specific and would likelynot reduce the production of by-products and/or off-spec products to thelevels that could be achieved withoutthe added contamination from therecycled wastewater. Some of theresulting by-product and off-specproduct would likely be hazardous andrequire treatment and disposal underSubtitle C of the Resource Conservationand Recovery Act. Again, even if zerodischarge were attainable, the extent towhich the imposition of a zerodischarge standard would result in non-water quality environmental impacts,and how those impacts would compareto the reduction of pollutant loadings toreceiving waters, could only beevaluated on a product/process-by-product/process basis. (See Rohm andHaas Co. Comment at 2 (identifyingnon-water quality impacts that couldresult from total recycle, includingimpacts from increased use of solventsto replace water in cleaning operationsand energy consumption resulting fromevaporation or off-site transfer ofwastewater, as well as from increasedproduction of off-grade materialrequiring disposal)).

In the case of the three productsidentified in EPA's technical analysis aspotential zero discharge candidates-melamine, phenolic, and urea resin-water that is produced in the processcan be removed, partially evaporated toraise the concentration of raw materials,and reintroduced to the reactor vessel(56 FR 63909). (Again, even for theseproducts, zero discharge is notachievable through recycle alone, butrequires an evaporation step.) Even inthis situation, the concentration ofcontaminants in the water will build asit is recycled and evaporated, and the

need to discharge (or otherwise disposeof) the wastewater could be eliminatedonly if the partially-evaporated water or"syrup" could be periodicallyincorporated directly into the facility'sproduct (id.). This incorporation wouldresult in lower grade product (whichmay or may not be saleable, dependingon customer demands), and possibly off-spec product that would have to bedisposed of. Therefore, EPA does notconsider these products to be goodcandidates for a national zero dischargestandard; rather, they are the onlyproducts EPA is aware of that appear tohold the potential to achieve zerodischarge through "total" recycle underparticular circumstances. Moreover,EPA believes that the potential for theseproducts to be manufactured withoutwastewater discharge based on totalrecycle is limited by the unique processchemistry involved and would not betransferable to other industry product/processes (id. at 63909).

In addition, even if EPA were to studythe plants that reported zero dischargebased on recycle for these products, themost EPA could do would be to developa technical basis for a zero dischargestandard for the particular chemicalprocesses employed by these plants toproduce these products. The threeproducts in question are producedthrough a variety of processes. TheOCPSF rulemaking record Identified sixreaction pathways or process routes forthe manufacture of melamine resins,eight process routes for urea resinmanufacture, and 14 process routes forthe manufacture of phenolic resins.With respect to new sources, this list ofprocesses is almost certainly not acomprehensive list of possibleprocesses; there are numerous ways tomake most OCPSF products, andcompanies often choose or developspecific processes to maximize the useand re-use of raw materials and torespond to customer specifications. Forexample, a company may produce aproduct through a process that creates aby-product that can be used as an inputfor the manufacture of another product;the use of this by-product as an inputwould dictate the range of processoptions for production of the secondproduct. Similarly, a plant may chooseto employ a particular process to makea given product because the processcreates a by-product that the plant cansell or use in another process. Thedevelopment of new, valuable productsthrough the beneficial use of by-products has been characteristic of thedevelopment of the OCPSF industry inthis century (1983 DD, pp IMI-1 to 2).

Each of the processes used to 'manufacture the three candidate resins

differs significantly in terms of the rawmaterial inputs to the reaction vessels,and likely differs significantly in termsof other variables associated withreactor vessel conditions. Even if EPAcould determine that zero discharge wasdemonstrated for a particular processused to manufacture urea resin forexample, that would not demonstratethat the manufacture of the sameproduct through a different chemicalprocess could achieve zero discharge,and would not provide the technologybasis to impose a zero dischargestandard on new plants using differentprocesses to manufacture urea resin. Itis entirely possible that no new plantwill ever manufacture any of the threecandidate products through the specificprocesses used by the plants identifiedin EPA's database. For EPA to go furtherand impose a zero discharge standardon manufacture of the product itself,without regard to process, would likelysignificantly constrict themanufacturing options open to futureplants. EPA would be extremelyreluctant to do this, given the flexibilitythat has characterized the developmentof the OCPSF industry and the inabilityto project the impacts of such aconstriction, including non-waterquality environmental impacts.

Kodak supports EPA's conclusions.Kodak states that the principal technicalconstraints on complete recyclinginclude product specifications thatrequire the use of very pure water,chemical reactions that generate water,and the need to eliminate from thewater recycle loop trace contaminantsthat accumulate to undesirableconcentrations when water iscontinually recycled and evaporationconcentrates the contaminants. Kodakalso notes that very few chemicalprocesses share the necessarycharacteristics that might permit totalrecycle of wastewater-" * * *processes where any process wastewaterand associated contaminants can becontained in the products andbyproducts without infringing onproduct quality, and * * * processes inwhich all excess process water isevaporated * * "(Kodak Commentsat 3-4).CMA in its comments also states,

"The need for blowdown (a wastewaterdischarge) is characteristic of all recycleprocesses-accumulation ofcontaminants present in the processfeedstocks and catalysts, generated bythe chemical reactions, or brought inwith the makeup water will eventuallyaccumulate to unacceptable levels andadversely affect the product. The onlytime a process is truly zero discharge iswhen the blowdown is accomplished by

36880


letting contaminants leave with theproducts. For basic organic chemicals,plastics, and synthetic fibers, this israrely an acceptable approach" (CMAComments at 50).

EPA cannot rule out the possibilitythat the manufacture of one or more ofthe remaining 25,000 products in theindustry could achieve zero dischargethrough recycle, but EPA believes thenumber would be very small and theinvestment in resources to develop thetechnology basis would be enormous.As explained, it would require adetailed technical study or individualproduct/processes, treating separateproduct/processes as separatesubcategories, and then determining thefeasibility of total recycle andestablishing design and operatingparameters for each subcategory (56 FR63906-07). EPA rejected this product/process-by-product/process approach asunworkable in the 1983 proposal thatformed the basis for the entire OCPSFregulation, deciding to regulate theOCPSF industry through concentration-based rather than mass-based limits dueto the complexity and variableproduction within the industry (id. at63907). EPA recognized at the time thatthe concentration-based approachinvolved a trade-off, permitting broaderindustry coverage but diminishing theopportunity to promote recycle on anational level. This diminishedopportunity arises because theconcentration-based approach does notfocus on specific product/processes butregulates through end-of-pipe limitsimposed on broad, general subcategoriesof production (id.). EPA concluded atthe time, and continues to believe, thatthe attempt to regulate the OCPSFindustry through the analysis andregulation of individual product/processes is simply infeasible.

As EPA has explained, theconcentration-based approach did notmean that recycling andreuse ofwastewater would not be encouraged atindividual plants. Rather, the guidelineis structured so that recycling and reuseevaluations will be made on a facility-by-facility basis (which is the only waythey can be made) by the permit writeror control authority, rather than throtighthe national guideline (id.). In both theDevelopment Document accompanyingthe final OCPSF guideline and insubsequent guidance, EPA providedguidance to permit writers in using flowreduction as the basis to set permitlimits (id.). Elsewhere in today'spreamble, EPA discusses this guidancefurther and urges permit writers tocarefully evaluate the opportunities forwastewater recycle and reuse atindividual plants.

EPA believes the detailed technicalreview it has performed fully complieswith the Fifth Circuit's remand and theClean Water Act. EPA developed a600,000-page record for the OCPSF ruleand, based on this record, hasdetermined that the potentialopportunity for total recycle in theOCPSF industry is extremely limited.2

In this situation, EPA believes it hasdiscretion to determine the utility ofundertaking extensive additional datacollection and technical analyses and,in this sense, believes it is appropriateto consider the "practicability" to theAgency of developing the technologybases for zero discharge standards. TheFifth Circuit, based upon itsunderstanding of the OCPSFadministrative record, ordered EPA toreconsider the appropriateness of zerodischarge standards through recycling.EPA has done that and does not believethe remand imposed an obligation toundertake significant new datacollection and analysis where, based onits review of the administrative record,EPA does not believe that a zerodischarge standard based on recyclewould be feasible for more than a fewof the product/processes in the OCPSFindustry. See CMA v. EPA, 870 F.2d at209 (upholding EPA determination notto establish BPT limitations based onpolishing ponds following biologicaltreatment where EPA determined that,although 17 plants successfullyemployed ponds, the experience ofthese plants could not be transferred toother plants in the OCPSF industry dueto the diversity of OCPSF wastewatercharacteristics).

C. Appropriate BAT SubcategorizationIn its comments, NRDC argues that

EPA's establishment of the BATsubcategorization scheme violates theClean Water Act because it allows thedischarger to choose the technologybasis for the BAT limitations based onwhether the discharger installs end-of-pipe biological treatment to complywith the BPT limitations. EPA disagrees.As a general matter, NRDC is correct instating that it is EPA's responsibility toidentify the best available technologyand to establish limitations based on its

2 EPA notes that the 1987 Development Documentaccompanying the final OCPSF guideline stated that"Irlecycling systems can achieve significantpollutant load reductions or zero discharge atrelatively low cost," at VII-8. Based on EPA'scomprehensive re-evaluation of the OCPSF recordand technical analysis in response to the remand,this statement is incorrect. Recycle is widelypracticed in the industry and, depending onindividual plant processes and configurations, canachieve significant pollutant load reductions.However, recycle alone, except in very limitedcircumstances, does not achieve zero discharge.

application. However, as EPA explainedin its proposal, the Agency concludedthat it was not feasible, necessary ordesirable to eliminate or limit theapplicability of the Subpart J, non-end-of-pipe biological treatmentsubcategory. Based upon thesedeterminations, as described in theproposal and below, EPA believes thesubcategorization scheme represents thebest approach for the OCPSF industryand is a lawful application of EPA'sdiscretion in selecting BAT. EPA doesnot believe the Clean Water Act requiresthe Agency to develop a scheme whichis not technically defensible and whichwould create undesirable treatmentincentives within the regulatedcommunity.NRDC identifies three alternatives tothe present scheme that It considersvalid. Each is addressed in turn.

1. BOD FloorFirst, as explained above in Section

IV, NRDC suggests that EPA shoulddevelop a BOD "floor" (i.e., a minimumBOD level) to limit the applicability ofSubpart J. However, as EPA explainedin the proposal, the development of afloor is technically infeasible due to thelack of a theoretical minimum BODlevel for sustaining biological treatmentand the great variability of OCPSFproduction and wastewatercharacteristics.

EPA received a number of commentssupporting its conclusion that it wouldbe infeasible to establish a BOD floor forthe OCPSF industry. CMA points outthat there are a number of factors thataffect a plant's ability to operate abiological treatment system effectively,including not only the plant's BOD loadbut also the variability of the plant'swaste concentration, wastewater flows,and chemical composition andtreatability (CMA Comments at 9). TheSynthetic Organic ChemicalsManufacturing Association (SOCMA)states that, even though it may betheoretically possible to establish a BODfloor under steady-state laboratoryconditions for a particular waste stream,It would be futile to try to establishcutoffs for the large number of full scaleproduction facilities in the highlycomplex and diverse OCPSF industry(SOCMA Comments at 11). The Societyof the Plastics Industry (SPI) agrees withEPA's view, presented in the proposal,that process streams in the OCPSFindustry "may change frequently anddramatically" (SPI Comments at 2).

EPA continues to believe that thecomplexity of the industry and thefrequent changes in production andwastewater characteristics, as describedin the proposal and in comments, make

36881


it infeasible to establish a ROD floor forthe operation of biological treatmentsystems. While a given plant may beable to operate a biological system at agiven long-term average HOD level, thatdoes not assure that another plant withthe same long-term average ROD level,but with a different waste streamcomposition or varying ROD levels, willalso be able to operate a biologicalsystem.

The feasibility of establishing a flooris further limited by the complexity ofthe OCPSF industry and EPA'sconsequent inability to develop adetailed knowledge of the productionpractices and other factors affectingwastewater treatment at each OCPSFplant. In the face of this complexity.EPA based the OCPSF guidelines on anumber of simplifying assumptions andconclusions (see, e.g., 56 FR at 63901).For example, in view of thefundamental, delicate inter-relationships among conventional,toxic, and non-conventional pollutantcontrols, EPA developed a schemeunder which it recognized that differentplants would exercise discretion toadopt plant-specific treatmentconfigurations in order to comply withthe promulgated BPT and BAT limits(id. at 62901-02). Without a moredetailed working knowledge ofindividual plants, which would beextraordinarily difficult to develop, theestablishment of a ROD floor or series offloors that would require plants toinstall end-of-pipe biological treatmentsystems would likely interfere with thecomprehensive waste managementsystems that have been appropriatelyinstituted at individual plants.

Moreover, as EPA explained in theproposal, the Agency does not believe itis necessary to establish a BOD floorbecause plants that need to achievesignificant ROD reductions willgenerally be motivated by economicconsiderations to install biologicaltreatment over the more costlyalternatives. In addition, as explained inthe proposal, the existence of Subpart Jwill not result in significantly increaseddischarges of pollutants to theenvironment over Subpart L Based onEPA's 1991 survey of the 84 directdischarge plants that did not have end-of-pipe biological treatment at the timeof promulgation of the OCPSFguideline, EPA concluded that 47 of theplants have not installed biologicaltreatment to comply with the OCPSFguideline (April 26, 1991 memorandumto the December 6, 1991 Public Record:"Current Status of Direct Non-BiologicalFacilities," R.00365-428). (This is morethan the 23 plants projected at OCPSFpromulgation to comply without

installing biological treatment.) Themaximum projected increased loadingsof the OCPSF-regulated toxic pollutantsto receiving waters associated withthese plants' complying with Subpart Jrather than Subpart I is estimated to be2,471 lbs/yr for the 47 plants that havenot installed end-of-pipe biologicaltreatment as of April 1991. Theseestimates are lower than the comparableestimates presented at proposal becausethey account for the incidental removalsof the four pollutants that are notregulated under Subpart J (the proposalnoted that the estimates presented werehigh because incidental removals werenot accounted for (56 FR at 63901)).These pollutants appear in wastestreams with other pollutants that areregulated by Subpart J limits based onIn-plant biological and/or activatedcarbon treatment, and are incidentallyremoved by these technologies (April29, 1993 memorandum to the OCPSFPublic Record "Toxic PollutantLoadings for the 47 Projected Subpart JPlants). (The total estimated industryloadings (before BAT and PSES) are24,166,480 lbs/yr; projected loadingsafter compliance are 571,723 lbs/yr, 56FR at 63901.) In addition, as explainedin the proposal 1,188 lbs/yr (48 percent)of the additional removalsaccomplished by Subpart I are projectedto result from volatilization of thepollutants into the air during end-of-pipe biological treatment.

In addition, EPA believes that a BODfloor would be undesirable in that itwould likely result In irrational andundesirable wastewater treatment andwaste management decisions. Inparticular, as explained in the proposal,EPA is concerned that plants that wouldotherwise be able to control ROD andTSS levels, and thereby comply withBPT, through in-plant product and by-product recovery or other source controlmeasures would eliminate or otherwisereduce the effectiveness of these in-plant controls in order to ensuresufficient organic matter to operate abiological system, or to operate such astem in a cost-effective fashion. Evena BOD floor could be developed, e.g.,

based on raw waste ROD load, such astructure would create incentives tomaximize ROD at the end-of-pipe. Oneof the distinguishing features of theOCPSF industry, noted throughoutEPA's development of the OCPSFguideline, is the variability ofwastewater characteristics at individualplants, including BOD levels, and theneed for plants to take protectivemeasures to ensure stable, healthy

opulations of microorganisms toiodegrade the organic pollutants in

their waste streams (e.g.. 56 FR at63900). Plants that are required tocomply with BAT limits most readilyachieved through end-of-pipe biologicaltreatment will often have the incentiveto maximize end-of-pipe ROD levels toassure consistently adequate ROD levelsto sustain biological treatment in theface of variability.

In-plant source controls can recovervaluable manufacturing by-products(thus eliminating these by-products aswastes), reduce or eliminate wastestreams, reduce in-plant and end-of-pipe treatment costs, and reduce oreliminate pollutants inhibitory or notamenable to end-of-pipe treatmentsystems. Process modifications,Improved instrumentation, addedoperator training, and solvent recoveryas well as water reuse, recovery, andrecycle are the first considerations inthe waste management and wastewatertreatment process (Proposal"Development Document for EffluentLimitations Guidelines and Standardsfor the Organic Chemicals and Plasticsand Synthetic Fibers Point SourceCategory," (EPA 440/1-83/009-b),February 1983 (hereafter referred to as1983 DD, Vol. I pp 119 to 122, Vol I ppVII-1 to 2; 1987 DD pp VII-1 to 10).Simple good housekeeping, such asproper maintenance and clean-uppractices, contribute to reducedpollutant loadings. In addition to thesewater use techniques and bestmanagement practices, certain in-planttreatment techniques, such as steamstripping, recover valuable products andeffectively eliminate pollutants from aplant's waste stream.

EPA believes these in-plant wastemanagement practices are generallypreferable to the alternative of lettingpollutants remain In the waste stream to

e removed by end-of-pipe treatment. Inthe 1983 OCPSF proposal, EPAdiscussed the advantages of a regulatoryscheme that focuses on in-plant controlsrather than end-of-pipe treatment:"Such an emphasis would result in areduction of the overall pollutantrelease through various environmentalmedia that might otherwise occurthrough a heavier reliance on end-of-pipe biological treatment. For example,

iological treatment can, in someinstances, cause the transfer of somevolatile organic pollutants from thewastewater to the air, and theadsorption of some other organicEollutants, as well as metals, to the

iological sludge, which is thendisposed of through methods whichmay affect other media. While some in-plant physical/chemical controls maysimilarly transfer pollutants to othermedia * *, other in-plant controls

36882


and treatments return at least somepollutants to the process, therebyminimizing total environmentalreleases" (48 FR 11838; March 21,1983).

In addition, the reduction of pollutionor treatment of pollution inconcentrated waste streams close to theprocess source is more efficient andoften less energy-intensive than end-of-pipe biological treatment of combinedwaste streams (which is a relativelyenergy-intensive treatment method), andEPA intended under the OCPSFguideline that plant managers, who willhave a far better working knowledge ofthe optimal treatment configurations forindividual plants than EPA can acquire,have the discretion to weigh suchfactors. EPA is reluctant to establish ascheme that discourages plants fromdoing what it would otherwiseencourage them to do, which is toreduce pollutants at the source, recyclewaste streams to the extent practicable,or treat pollutants close to the sourcewhen that is the most efficient treatmentalternative. SOCMA, in its comments, at12, raises a similar concern, stating thatlimitation of Subpart J would preventdischargers from selecting alternativein-plant controls and techniques such asrecycling, waste reduction and materialrecovery.

EPA is especially reluctant to changethe structure of the OCPSF guideline assuggested by NRDC at this point, morethan five years after promulgation. Mostplants in the industry are now requiredto be in compliance with all of theOCPSF limitations and standardsincluding the BPT limits. Many of theseplants may not have sufficient BODevels to operate end-of-pipe biological

treatment systems. Compliance with theOCPSF subcategorical "maximum formonthly average" and "maximum forany one day" BOD limitations requiresplants to achieve long-term annual-average BOD effluent levels rangingfrom 12 to 41 mg/1 depending on thesubcategory or subcategory mix for anyplant (12 mg/l BOD for Other Fibers, 16for Rayon and Thermoplastics, 20 forCommodity Organics, 23 for BulkOrganics, 30 for Specialty Organics, and41 for Thermoplastics; p VII-176, Vol. I,1987 DD). These levels are below thelevels for which the Agency has datademonstrating the feasibility ofbiological treatment. The OCPSF recordhas BOD data for 98 plants with end-of-pipe biological treatment. The influentong-term average BOD levels for these

plants ranged from 60 mg/l (a singleplant) to 9,420 mg/l, with only 12 plantsbelow 125 mg/l. (The Agency's fieldsampling efforts identified oneadditional plant with a two-week

average influent value of 37 mg/l, butthis value may not be representative ofthe plant's long-term average InfluentBOD concentration.) Thus, plants thathave achieved compliance with BPTwithout the use of end-of-pipebiological treatment would likely haveto (and, in any event, would likely havethe incentive to) eliminate the in-plantsource controls and/or cease operationof the in-plant treatment systems theyhave installed in order to elevate theirBOD levels. EPA does not believe thisis a rational regulatory result.

NRDC suggests, based on the OCPSFrecord, that even in instances of lowBOD levels, nutrient addition, pureoxygen, or extended aeration could beused to operate end-of-pipe biologicaltreatment systems. However, as EPAexplained in the December 1991proposal, these practices are notgenerally applicable substitutes forsufficient BOD to sustain biologicaltreatment. In their comments, CMA, atpp 11-12, and Ashland Chemical, at p3, agree. The several brief referencesidentified by NRDC to practices used inthe OCPSF Industry, out of the 600,000-page OCPSF record, do not provide abasis to treat these techniques as"available technologies" on which EPAcould base limits.

EPA agrees with CMA's assessmentthat the pure oxygen variation ofactivated sludge biological treatmentoffers no particular advantages becauseit is not used to treat low-strength (i.e.,low BOD) wastewaters (CMA Commentsat 11). The pure oxygen activated sludgevariation generally treats widelyfluctuating organic loadings and high-strength wastewaters more effectivelythan "standard" activated sludgedesigns. CMA also notes that highoxygen transfer rates in the aerationbasins due to the use of pure oxygenassure that sufficient oxygen is availablefor biological removal of organics withhigh biomass populations. However,high biomass concentrations can only beachieved when ample biodegradableorganic compounds (BOD) are availablein the wastewater. Pure oxygenactivated sludge treatment is not asolution for biological treatment ofwastewater with low raw waste BODconcentrations.

In the case of extended aeration,which is a variation of the complete mixactivated sludge design, low organicloadings and long aeration times permitmore complete wastewater degradationand partial aerobic digestion of themicroorganisms (p VII-63, Vol. I, 1987DD). However, as CMA also notes,effective extended aeration operationrequires a wastewater with sufficientorganic substrate to form a biomass that

can be flocculated and kept in thetreatment system for extended periodsof time. Wastes with low BOD may notgenerate enough biomass that can besettled from the treated effluent bygravity and recycled to the aerationbasin. In the case of insufficient rawwaste BOD, the biological solids wouldwash out of the system, and extendedaeration operation would not work.(CMA comment at 12). Extendedaeration treatment is not a substitute foran adequate BOD level, and, asdiscussed above, whether an extendedaeration system or any other type ofbiological system were used, the BODlevel that would be sufficient to sustainbiological treatment would vary fromplant to plant,

NRDC also misunderstands thefunction and purpose of "nutrientaddition." As generally practiced in theOCPSF industry, nutrient addition is theprocess of adding nitrogen orphosphorous in a chemically combinedform to a waste stream. (p XV-14, Vol.11, 1987 DD) These nutrients areanalogous to vitamins; they may help aviable biological system operateoptimally, but they are not a substitutefor adequate substrate or "food." CMA'scomments state that it is possible to addsupplemental organic substrate to atreatment system, but this practice is atodds with the purpose of effluenttreatment. CMA also correctly observesthat the addition of readilybiodegradable substrate, such asmolasses, to a low strength wastewaterdoes not ensure the growth of anactivated sludge that is capable ofeffectively removing organic toxicpollutants. The biomass must beacclimated to the same or similarcompounds that must be removed. Thismay create the need to add toxicpollutants to the wastewater to assurethat they can be removed consistentlyby the treatment system. (CMAComments at 12.) EPA is not aware thatthis technique Is practiced in the OCPSFindustry, and does not encourage it. TheAgency believes that the framework ofthe OCPSF regulation should encourageplants to minimize the overallgeneration of raw wastes rather thancreate an incentive to increase thedischarge of "biodegradable"compounds to ensure the operation ofbiological treatment systems.

With respect to these suggestedsolutions to low BOD levels, NRDCargues that EPA has based itsruemaking on ignorance, and that theAgency was obligated to collectadditional information or performadditional technical analyses to evaluatewhether nutrient addition or othertechniques might be a substitute for low

36883

36884 Federal Register / Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regulations

BOD. EPA disagrees. EPA has evaluatedthese options, based on the voluminousOCPSF record and on commentsreceived, and does not believe they areviable substitutes for adequate BOD.Rather, they are approaches that, underplant-specific circumstances, can beused to improve the operation ofbiological treatment systems. EPA doesnot have an obligation to collect dataand perform extensive technicalanalyses to investigate options that itdoes not believe provide viable bases fortechnology-based limits in a nationalguideline.

More generally, NRDC argues thatEPA has performed a "paper review" inresponse to the Fifth Circuit's remandrather than collecting additional dataand performing additional technicalanalyses regarding the feasibility ofestablishing a BOD floor. NRDC believesthat EPA should have evaluated theplants that have not installed end-of-pipe systems to determine a BOD floorfor the subcategory. Again, EPAdisagrees. EPA conducted a detailedtechnical re-evaluation in response tothe Court's remand and concluded thatthe present subcategorization scheme isthe best approach for the OCPSFindustry; EPA does not believe theestablishment of a BOD floor isnecessary or desirable.

In addition, EPA does not read intothe Fifth Circuit's remand the samerequirements as NRDC does. The courtremanded the subcategory scheme "fornotice-and-comment proceedings," andremarked that such proceedings maydisclose that NRDC's suggestions areneither necessary nor feasible (870 F.2dat 236). Prior to publishing theDecember 1991 proposal, EPAconducted an in-depth review of itsOCPSF database, an extensive literaturesearch, and additional pollutantloadings analyses to evaluate NRDC'ssuggestions. The Agency does notbelieve that the Court's remand, theClean Water Act or general principles ofadministrative law require EPA tocollect data and perform extensiveadditional analyses where, based uponthe information before it, the Agencyconcludes that such efforts are not likelyto be valuable.

The Clean Water Act authorizes EPAto consider non-water qualityenvironmental impacts (includingenergy requirements) and other factorsEPA deems appropriate in setting BAT,,limits. CWA 304(b)(2)(B). EPA believesit has discretion to create a BAT schemethat does not discourage plants fromcomplying with BPT through in-plantsource control and treatment, with thepotential to recover product and by-product and otherwise to reduce

pollution at the source. EPA continuesto believe that the current BATsubcategorization scheme is theappropriate scheme for the OCPSFindustry.

2. Subcategory Limited to CertainProcesses

Alternatively, NRDC argues that EPAshould limit subpart J to those processesfor which an adequate showing of low-BOD wastewater has been createdthrough industry comments, such as thechlorosolvent industry. NRDC arguesthat EPA has impermissibly extendedsubpart J beyond the scope of thecomments and data that support It. EPAdisagrees. The OCPSF record supportsthe conclusion that the phenomenon oflow BOD levels is not limited tospecified industry segments; rather, lowBOD wastewater may occur throughoutthe OCPSF industry. Therefore, EPAbelieves the availability of subpart Jshould not be limited to specificindustry segments.

As NRDC notes, the specificcomments and data during the 1987OCPSF rulemaking with respect to lowBOD levels related principally to thechlorosolvent industry. Nonetheless,EPA had already concluded that thecharacteristics associated with low BODlevels extended beyond chlorosolventmanufacture to other generic processchemistry and other product/processes.In response to the Halogenated SolventIndustry Alliance 1985 comment thatEPA should set separate BATlimitations for stand-alonechlorosolvent production plants becausebiological treatment methods are notused and would be ineffective if used,EPA declined to establish a subcategorylimited strictly to chlorosolvent plants,recognizing that other types of facilitiesare also expected to have low BODlevels (1987 Public Record commentresponse No. 254, R103,160).

In 1983, the Agency had studied 46generic OCPSF process groups and theirpotential to generate various BODoadings. The study identified 19

generic processes expected to generatewastewater of relatively low BOD whenrefractory chemical species predominatein the wastewater and when relativelyfew chemical species are present in thewastewater (1983 DD, Vol.I, p 63-68).EPA concluded at the time that themanufacture of halogen compounds Ingeneral, not just chlorosolvents inparticular (chlorosolvents are includedamong halogen compounds), tend togenerate relatively low raw waste BODlevels. Other generic process groups thattend to produce wastewater with thelowest BOD concentrations includealkylation, isomerizatlon,

polymerization (bulk & addition), andphosgenation (id.). Plants withmanufacturing activity limited to thesegroups generally, but not always,produce wastewater with relatively lowBOD concentrations.

Further analysis following the 1983OCPSF proposal revealed that, althoughgeneric process groups provide a roughcorrelation with raw waste BOD levels,individual product/processes within thegeneric process groups exhibit widelyvarying raw waste BOD levels, andgeneric process groups cannot beconfidently classified as "low-BOD" or"high-BOD" raw waste groups. Forexample, even within the low BODhalogenation group, there is asignificant variation of raw wastewaterBOD concentrations--chlorobenzeneprocessing averaged 20 mgfl BOD andtrichloroethylene averaged 36 mg/lBOD; however, 2-dichloroethaneaveraged 1,869 mg/Il BOD (Appendix S,draft "Contractors Engineering ReportAnalysis of Organic Chemicals andPlastics/Synthetic Fibers Industries,"November 1981, pp S-1839, S-3301,and S-2084, respectively). Similarvariability exists among the product/processes within the generic processchemistry groups that are projected togenerate the highest levels of BOD. Forexample, within the high BODesterification group, n-butyl acrylate/ethylhexyl acrylate processing averaged5,481 mg/l BOD and dimethylterphthalate processing averaged 1,310mg/i; however, n-butyl methacrylateprocessing averaged 4 mg/l (id., pp S-1,666, S-2,166, and S-1,770,respectively).

In supporting EPA's current BATsubcategorization scheme, CMAobserved (and EPA agrees) that theSubpart J limits cannot be confined toa small number of subcategories sincelow-BOD wastewaters may occur acrossvarious subcategories, citing data fromthe OCPSF database showing low BODlevels across a range of product groups(CMA Comment, pp 3-4).

Thus, generic reaction chemistry andgeneral product groups are incompletepredictors of the raw wastewatercharacteristics necessary to supporteffective biological treatment. EPAbelieves that the phenomenon of lowBOD levels, and the availability ofSubpart J, cannot be limited to certainindustry segments, and, as explainedabove, believes that such a limitationwould create undesirable incentives inother industry segments.3

NRDC similarly argues that BOD levels In theOCPSF industry generally appear high. and lowBOD wastewater appears concentratd In the meoor alternative diwarge categories (ciring the 167


3. FDF Variance AlternativeFinally, NRDC argues that EPA should

eliminate Subpart J altogether andaddress low-BOD situations throughfundamentally different factors (FDF)variances (or maintain the Subpart butapply it only where a site-specificshowing of necessity is made). However,as explained above, EPA believes thecurrent subcategorization schemeappropriately reflects the potential forlow BOD levels throughout the OCPSFindustry. Moreover, the currentsubcategorization scheme does notdiscourage source control and other in-plant waste management techniques.EPA has discretion in determiningwhether to account for industrycharacteristics throughsubcategorization or through the FDFprocess. In this instance, EPA's decisionto subcategorize the industry wasrational and within its discretion, forthe reasons explained above.

D. Applicability of the Revised Pass-Through Methodology

A number of industry commenterssupported the Agency's proposedconclusion presented in the December1, 1992 NOA that phenol and 2,4-dimethylphenol do not pass throughPOTWs, but urged that the modifiedpass through analysis used to reach thatconclusion be applied to the remaining11 remanded PSES pollutants todetermine that they also do not passthrough. EPA disagrees, for the reasonsexplained below.

DD at 13-30-36 and V-32-33). However, the tablescited by NRDC display mean, aggregated BODvalues for plants within different subcategories, anddo not reflect variability of plants within thesubcategories. For example, Table V-20, 1987 DDat V-32, shows a mean BOD value for 62 directdischarge thermoplastics plants (Part 414, SubpartD) of 725.190 mg/l. Despite this relatively highmean, individual plants within the Subcategorywill exhibit a broad range of BOD values, and themean reveals nothing about whether individualplants will have sufficient BOD levels to sustainbiological treatment. As explained above, varyingBOD levels occur throughout the industry, whetherthe average POD level for a particular subcategoryis low or high. In any event, the tables cited byNRDC contain too little zero/alternative dischargeplant data for any meaningful comparison betweenregulated and zero discharge plants. For example,the data base used to display the average raw wasteBUD levels presented in Table V-20 contained onlyone zero discharge plant with 2.28 percent of itsproduction in the Specialty Organic ChemicalManufacturing subcategory. At this plant. 280 mg/I BOD in raw wastewater was attributed to SpecialtyOrganic Chemical production. In contrast, the entiredata base used to display average raw waste BODlevels for the industry contained the equivalent of60.37 Specialty Organic Chemical plants on aproduction-weighted basis. The data derived from2.28 percent of one zero discharge plant cannot bemeaningfully compared with the data from 60.37discharging plants to draw any conclusion aboutthe BOD levels exhibited by zero discharge asopposed to discharging facilities.

1. BackgroundUnder section 307(b) of the Clean

Water Act, EPA is required topromulgate categorical pretreatmentstandards for pollutants which aredetermined not to be susceptible totreatment by POTWs or which wouldinterfere with the operation of POTWs(33 U.S.C. 1317(b)(1)). The methodologyEPA used to evaluate whether apollutant is susceptible to treatment(i.e., whether it "passes through"POTWs) for purposes of establishingcategorical pretreatment standards forthe OCPSF point source category, aswell as in previous guidelines,compared the median percent removalof each pollutant of concern achieved bydirect dischargers employing BAT-leveltreatment to the median percentremoval achieved by well-operatedPOTWs with secondary treatment (1987DD, pages VI-22 to 32). The source ofEPA's data on POTW removals used todetermine pass through was a 1982study of the performance of 50 PQTWs(the 50 PQTW Study; EPA 440/1-82/303).

Where EPA had data on pollutantremovals from both POTWs and directdischargers, EPA relied exclusively on acomparison of the percent removalsdemonstrated by the data (1987 DD,page VI-23). If the data showed thatdirect dischargers with BAT-leveltreatment achieved a higher percentremoval of a pollutant than well-operated POTWs, then the pollutant wasdetermined to pass through, and EPAestablished categorical pretreatmentstandards to regulate the pollutant. Ifthe POTWs showed removals equal to orgreater than the direct dischargers, thenthe pollutant was determined not topass through, except for volatile andsemi-volatile pollutants, for which EPAdetermined that some of the removalsfrom the wastewater were accomplishedas a result of "air stripping" (id., p VI-27). Because these removals were theresult of transfer of the pollutants to theair rather than treatment, EPA applied a"volatile override" to determine thatthey passed through and establishedpretreatment standards (id., p VI-37).EPA also applied the volatile override toseveral pollutants for which it lackedPOTW removal data based on itsprofessional judgment that thesepollutants volatilize based on theirphysical characteristics.

In determining the percentage of apollutant that a plant removed, EPAcompared the concentration of thepollutant in the influent to theconcentration in the effluent. Whereeffluent concentrations for either directdischargers or POTWs were below the

analytical minimum level, usually 10ppb, EPA assigned this value to theeffluent id., p VI-23). Although theactual concentrations may have beenbelow the analytical minimum level,EPA concluded that this represented areasonably conservative approach, sincethe actual levels could not bequantified. For phenol and 2,4-

imethylphenol, both the directdischargers and well-operated POTWsin EPA's data base generally achievedeffluent concentration levels that werebelow the analytical minimum level,which is 10 ppb, (id., p VII-186).Accordingly, EPA assigned 10 ppb tothe effluents. EPA's pass throughmethodology was upheld in the OCPSFlitigation (CMA v. EPA, 870 F.2d at 243-48).

Allied Signal, Inc. and othercommenters on the December 6, 1991proposal argued that phenol and 2,4-dimethylphenol are highlybiodegradable and are treated byPOTWs to the same degree as directdischargers, and that EPA's passthrough analysis for these pollutantswas therefore overly conservative. Thecommenters argued that the apparentdifference between direct dischargerand POTW performance arises from thefact that the direct dischargers in EPA'sdatabase have significantly higherinfluent concentrations than thePOTWs; as a result, the directdischargers show higher removals thanthe POTWs because their wastewaterconcentration is treated from a highconcentration down to 10 ppb, whereasthe POTWs' wastewater is treated froma comparatively lower concentrationdown to 10 ppb. The commenters alsosubmitted performance data to supporttheir claims, which consisted of pilot-scale biological treatment studies aswell as actual sampling data fromPOTWs that receive industrialwastewaters containing these twogollutants showing their highiodegradability.

EPA concluded that the commenters'arguments might have merit. Using 2,4-dimethylphenol as an example, themedian percent removal of thispollutant demonstrated by directdischargers was 99.8 percent. This wasbased on data from four OCPSF plantswith average influent concentrationsranging from 697 to 29,868 ppb, andwith 30 of 37 effluent values below theanalytical minimum level and thereforeassigned values of 10ppb. For POTWperformance, EPA had a singleobservation of a POTW with an averageinfluent concentration of 20.5 ppb andan average effluent concentration belowthe analytical minimum level, whichwas also assigned a value of 10 ppb.

36885


Thus. POTW removal was calculated at51.2 percent, and the pollutant wasdetermined to pass through. In this case,the pass through determination may bean artifact of the differing influentconcentrations and does not necessarilyreflect a real difference in removals.4

In the December 1, 1992 NOA, EPApresented a comprehensive assessmentof the available data with respect tophenol and 2,4-dimethylphenol, as wellas a chemical and engineering analysisof the fate of these two pollutants inbiological treatment systems, Based onthis analysis, EPA proposed toconclude, and today has concluded, thatthese pollutants are treated toessentially the same levels by directdischargers and POTWs and, therefore,do not pass through. However, as theAgency explained in the NOA, EPAgenerally is continuing to apply themedian percent removal methodologyused to determine pass through atpromulgation of the OCPSF guideline(57 FR at 56885). This methodology wasupheld in litigation as an appropriate,conservative approach to determiningpass through (870 F.2d at 243-48), andEPA continues to believe it is the correctapproach as a general matter. EPAdetermined that the approach is overlyconservative for the highly-biodegradable phenol and 2,4-dimethylphenol, but believes it isappropriate for the other 11 remandedpollutants. As explained in the NOAand below, EPA believes thesepollutants are less biodegradable and,consequently, less readily treatable byPOTWs, which typically have biologicaltreatment systems with much shorterdetention times than the systemsemployed by direct dischargers.

4 EPA acknowledged this phenomenon indeveloping the OCPSF rule and proposed severalmodifications of the pass through analysis,including applying a "removal differential" underwhich EPA would determine that a pollutantpassed through only if the analysis found adifference between direct discharger and POTWremovals that exceeded 5% or 10% (see. e.g., 50 FR29084-5 (July 17, 1985)). However, after carefullyreviewing comments arguing, among other things,that this approach would bias the analysis againsta finding of pass through, EPA decided to employits historical approach to pass through, with onevariation. In the final OCPSF rule, EPA edited itsdatabase to exclude POTWs with influentconcentrations of less than ten times the analyticalminimum level (typically 100 ppb), unless therewas no plant in the data base with influentconcentrations that high, in which case EPAretained the 20 ppb cut-off used in previousguidelines (1987 DD at VI-33). This mitigated theunderestimation of removals that could occur whencomparing very low influent concentrations to theanalytical minimum level.

2. Assessment of the RemandedPhthalate Esters and PolynuclearAromatics

In the NOA and accompanyingTechnical Support Document (TSD;"Technical Support Document for theOrganic Chemicals, Plastics andSynthetic Fibers Point Source CategoryNotice of Availability of NewInformation, November 30, 1992), EPAdid perform a data review and technicalanalysis for the other 11 remandedpollutants similar to that performed forphenol and 2,4-dimethylphenol. TheAgency reviewed the available data onthe removal of the two phenols as wellas the two other genera pollutantcategories covering the remaining 11pollutants, phthalate esters (PEs) andpolynuclear aromatics (PNAs). TheAgency also reviewed the availableliterature on the biochemicalmechanisms of biodegradation for all 13pollutants, and Investigated theadequacy of biological treatmentsystems at POTWs in effectively treatingthese pollutants via biodegradation. TheAgency included all of its performancedata from various data sources as wellas information collected from theliterature on the biochemicalmechanisms of biodegradation of thesepollutants in the Record supporting theNOA.

EPA's decision to modify itstraditional pass through methodologyfor phenol and 2,4-dimethylphenol isbased on EPA's conclusion that both thedata available for these two pollutantsand the chemical and engineeringanalysis performed by EPA indicatedthat the OCPSF pass throughmethodology is overly conservative forthese pollutants. The data and technicalanalyses do not support a similarconclusion for the other 11 pollutants.

EPA's analysis focused first on thedata from the OCPSF Record relating tophenol removal. A comparison ofmedian removals (the original OCPSFmethodology) indicated that phenolpasses through POTWs (TSD at 11,Table 11-2). However, when EPAarrayed all of the direct discharge andPOTW data for phenol, it becameapparent, as explained in the NOA, thatthe pass through conclusion was strictlyan artifact of the higher Influentconcentrations for direct dischargers inEPA's database. Viewing the data as awhole, POTWs appeared to achieveremovals that are essentially equivalentto those achieved by direct dischargers(57 FR 56886-87). This conclusion wasconfirmed by additional data EPAsolicited from three POTWs, thatdemonstrated phenol removals fromvery high influent concentrations (e.g.,

4,043 ppb at the Sheboygan RegionalWastewater Treatment Facility) to belowthe analytical minimum level. Inaddition, as explained in the NOA andthe accompanying TSD, EPAdetermined that 2,4-dimethylphenolwould be removed by POTWs to thesame degree as phenol, given its similarmolecular structure.

Three of the remaining elevenpollutants-fltoranthene, bis(2-ethylhexyl)phthalate and di-n-butylphthalate-were detected in POTWeffluent in the 50 POTW Study (TSD at11, Table 11-2). For these pollutants, theresults of the pass throughdetermination clearly are not merely anartifact of differing influentconcentrations but reflect worseperformance by POTWs. EPA has nobasis to conclude that these pollutantsdo not pass through.

With respect to the remaining eightpollutants, EPA does not have datacomparable to the data that provided abasis to modify the pass throughmethodology for the phenols. Inaddition, EPA's technical analysisconfirmed that phenol and 2,4-dimethylphenolare the most readilytreatable by POTWs of the 13 pollutants.EPA noted that while phenols arerapidly biodegraded in biologicaltreatment systems due to their simplemolecular structure, PEs and PNAswould be expected to biodegrade at amuch slower rate because of the

-additional time required to convertthese pollutants into a form that can bereadily biodegraded (TSD at 6).

Biodegradation does not commenceuntil a pollutant is "sorbed" by (i.e.,attached to) the microorganisms in thebiological treatment system that degradethe pollutant ('OCPSF Remand Issues-Assessing the POTW Compatibility ofRemanded Organic Pollutants Regulatedby PSES," November 20, 1992memorandum, Item No. 73 of theOCPSF December 1, 1992 NOA PublicRecord). Once sorbed, pollutantsdegrade at different rates that depend onstructural complexity. In order to bebiodegraded, a pollutant must be able topass through the cell wall of amicroorganism. This transfer will occuronly if the pollutant is compatible withthe proteins in the cell wall. Whilesmall, simple molecules are generallycompatible, the more complexstructures typical of PE and PNAorganic pollutants must first be brokendown into smaller chemical units byextra-cellular enzymes secreted by themicroorganisms. Thus, biodegradationdepends on the ability of themicroorganisms to structurally alterpollutants outside the cell wall whilethey are sorbed.

36886


As EPA explained in the NOA, thephenols have simple chemicalstructures that permit them to be rapidlytransferred through the call wall andbiodegraded (57 FR 56888). Thismolecular-level analysis is confirmed bythe fact that wastewaters containingEhenol and 2,4-dimethylphenol have

igh "biodegradation rate constants"(id. at 56887). (As explained in the NOA(57 FR at 56887), "biodegradation rateconstant" is a measure of how rapidlya compound or mixture of compoundsbiodegrades). In addition, these twopollutants have the highest compound-specific estimated biodegradation rateconstants of the 13 remanded pollutants(TSD at 11, Table 11-2)(biodegradationrate constants can be assigned to bothindividual pollutants and to wastestreams containing mixtures ofpollutants). In contrast, as furtherexplained in the Supplement.to theDevelopment Document accompanyingtoday's rule, the phthalate esters andpolynuclear aromatics are structurallymore complex, and require additionaltransformation steps before they can betransferred through the cell wall of thebiodegrading microorganisms andbiodegraded. These steps requireadditional time in the aeration basin ofa biological treatment system that isgenerally available at OCPSF directdischarge facilities, which typicallyhave detention times that exceed 24hours, but may not be available atPOTWs, where aeration basin detentiontimes are usually four to eight hours.

Thus, based on rate of biodegradation,EPA believes that phenol and 2,4-dimethylphenol are more readilytreatable by POTWs than the eightremaining pollutants. EPA recognizes,as several commenters on the NOApointed out, that organic pollutants maybe removed from wastewater bybiological treatment systems to varyingdegrees by removal mechanisms otherthan biodegradation. In particular,pollutants may be removed byvolatilization and by adsorption tosludge. However, as explained below,EPA believes that a pollutant'sbiodegradation rate is the most accurateindicator of whether the pollutant willpass through POTWs.

In general, volatile pollutants are notreadily treated in POTWs; rather, thesepollutants are volatilized or "stripped"to the atmosphere. As EPA explainedabove, EPA applied the volatile overridein the 1987 OCPSF guideline todetermine that several volatile andsemi-volatile pollutants pass throughwhere POTWs showed equal or betterpercent removals than direct OCPSFdischargers or where no POTW removaldata were available. In determining

whether to apply the volatile override,EPA considered total estimatedvolatilization of a pollutant after leavingan indirect discharge facility-i.e.,volatilization in both the aeration basin(i.e., the treatment basin) of thebiological treatment system andvolatilization in the sewer systems andpre-biological unit treatment operationsthat convey the pollutant to the aerationbasin (1987 DD at VIU-281).

For five of the PNAs that wereremanded-naphthalene, acenaphthene,anthracene, fluorene, andphenanthrene-EPA would haveapplied the volatile override in the 1987OCPSF rule to determine thesepollutants passed through if the percentremoval analysis had not shown passthrough. These pollutants have overallvolatilization rates comparable to therates for which the override wasapplied. For example, EPA applied theoverride to hexachlorobenzene,hexachloroethane andhexachlorobutadiene in promulgatingthe 1987 guideline (1987 DD at VIII-279). These pollutants have a 5 to 10percent estimated volatilization rate inthe aeration basin; the pre-biologicalvolatilization rates forhexachlorobenzene, hexachloroethaneand hexachlorobutadiene are estimatedto range from 19 to 39 percent, 59 to 66percent, and 48 to 73 percent,respectively (1987 DD at VIII-281).Similarly, the estimated aeration basinvolatilization rates for the fiveremanded PNAs at issue range from 10to 30 percent, and the estimated pre-biological volatilization rates range from12 to 82 percent (id.) EPA notes thatestimated volatilization rates forindividual pollutants vary depending onthe source of the estimate, and theaeration basin volatilization rates thatappear in the TSD, at 11, Table 1-2,vary from those presented in the 1987Development Document because theyare based on different technical studies.TSD Table 11-2, however, does notaccount for pre-aeration-basinvolatilization, and the overall estimatedvolatility of the five pollutants at issueis comparable to the estimated volatilityof the pollutants to which EPA appliedthe volatile override in 1987. Becausethese pollutants are chemically morecomplex than phenol and 2,4-dimethylphenol and, EPA believes,therefore less readily biodegradable inPOTWs, and because much of the"removal" of these pollutants prior toand during POTW biological treatmentis likely the result of volatilization, EPAcontinues to conclude, based on itstraditional methodology, that these fivepollutants pass through POTWs.

EPA believes the remaining threepollutants--diethyl phthalate, dimethylphthalate, and pyrene--are likelyadsorbed to sludge in the biologicaltreatment system. A compound'spropensity to separate from the waterphase and adsorb to sludge (whichincludes the microorganisms thatdegrade the compounds) is predicted byits "octanol/water partition coefficient."Pyrene, in particular, has a highestimated octanol/water partitioncoefficient, and would be expected toadsorb rapidly to the sludge in abiological system (TSD at 11, Table U-2). However, pollutants that are Initiallyadsorbed onto the sludge may become"desorbed" (i.e., may detach from thesludge) and pass through into thereceiving-stream if they are not rapidlytransferred through the cell wall andbiodegraded.

The ability of complex, organicpollutants such as phthalate esters andpolynuclear aromatics to remainabsorbed prior to being converted tosimpler compounds for transfer throughthe cell wall can be affected by manyconditions in the treatment system,including the presence of otherpollutants, electrolytes, oils and greasesand other more highly adsorbentcompounds ("Report to Congress on theDischarge of Hazardous Wastes toPublicly Owned Treatment Works,"February 1986, (EPA/530-SW-86-004),p 4-5). This can cause the pollutants todesorb prior to conversion andbiodegradation and pass through thePOTW to the receiving water. EPAbelieves this phenomenon explains whyorganic pollutants which are generallyconsidered highly adsorbable cansometimes be found at detectable levelsin the POTW effluent. For example,anthracene and phenanthrene have highestimated octanol-water partitioncoefficients and therefore would beexpected to adsorb rapidly to sludge(TSD at 11, Table 11-2). PQTW Number6 from the 50 POTW Study shows anaverage influent concentration ofanthracene and phenanthrene of 62.2ppb and an average effluentconcentration of 16.2 ppb, while POTWNumber 52 has a much higher averageInfluent concentration of 225.3 ppb foranthracene and 195.8 ppb forphenanthrene, both reduced to notdetected at 10 ppb (1987 Public Recordat 115910-115976). Based in this data,the propensity of these pollutants toadsorb to the sludge does not appear tobe a good indicator of POTW removalperformance. EPA believes that externalconditions in a biological treatmentsystem can affect the ability of a POTW

36887

Federal Register I Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regulations

to remove more complex pollutants byadsorption or biodegradation.

The overall removal data for the 13remanded pollutants appears to confirmthat octanol/water partition coefficientis not a reliable indicator of passthrough. Phenol has the lowest octanol/water partition coefficient of the 13pollutants but is rapidly and virtuallycompletely removed by biologicalsystems, including POTW systems. Incontrast, bis(2-ethylhexyl)phthalate anddi-n-butyl phthalate have among thehighest octanol/water partitioncoefficients, but achieved lower POTWremoval levels (TSD at 11, Table 11-2).In fact, the only pollutants among the 13remanded that were detected In POTWeffluents in the 50 POTW study-bis(2-ethylhexyl)phthalate, di-n-butylphthalate and fluoranthene-also hadthe highest octanol/water partitioncoefficients of the 13 pollutants (TSD at11, Table 11-2).

In sum, EPA believes that apollutant's estimated biodegradationrate is the best theoretical indicator ofwhether it will pass through POTWbiological treatment systems. As aresult, EPA continues to conclude thatdiethyl phthalate, dimethyl phthalate,and pyrene pass through based on itstraditional pass through methodology.These pollutants are structurally morecomplex and consequently less readilybiodegradable than phenol and 2,4-dimethylphenol, and are therefore morelikely to pass through POTW biologicaltreatment systems. Moreover, EPA doesnot have data demonstrating that thesepollutants are adequately treated byPOTWs.

E. Land AvailabilityCMA in comments asserts that EPA

overestimated the land available for theconstruction of biological treatmentsystems in its survey of eight indirectdischarge facilities by including in itsanalysis parcels of non-contiguous landand land that is obstructed by railroadtracks, buildings and other physicalobstacles. (CMA Comments at 39-41).This is not true. Each of the eightfacilities EPA surveyed has sufficientcontiguous, unobstructed land for theinstallation of the model biologicaltreatment system costed by EPA atproposal. Furthermore, the availableland is configured such that it canaccommodate the costed biologicaltreatment systems.

In the December 1991 proposal, theAgency recognized that the larger in-plant biological treatment systemscosted for compliance with the BATSubpart J limits and correspondingpretreatment standards would requiremore land than the smaller systems

costed for the 1987 OCPSFpromulgation. EPA investigated whetherland availability would be a constrainton the ability of OCPSF plants to installin-plant biological treatment. EPA'sinvestigation included the landrequirements for treatment of all 13 ofthe remanded PSES pollutants,including phenol and 2,4-dimethylphenol, which are not beingregulated by today's final rule. At thattime, 20 of the 242 indirect dischargeplants costed for in-plant biologicaltreatment were projected to requiremore than one acre of land. EPAprojected land requirements forindividual facilities based on themodeled raw waste concentrations forthe facilities developed by the Agencyfor purposes of costing compliance withthe 1987 OCPSF guideline. The Agencyvisited the eight indirect dischargefacilities with land estimates greaterthan one acre in the three-state area ofNew York, New Jersey, and Delaware.Indirect discharge facilities wereselected because their typical locationin urban areas makes them more likelythan direct dischargers to have land-availability constraints. EPA believesthe combination of large landrequirements and an urban settingmakes these eight plants a "worst case"sample of land availability.

Five of the plants visited hadsufficient land based on the landrequirements projected from theirmodeled raw waste concentrations (theremaining three had from 78 to 96percent of the projected requirements).The remaining three had enough landbased on their actual reported raw wasteconcentrations (the three plants hadfrom 1.9 times to 3.7 times more thanthe required land). EPA generally wasconservative in projecting raw wastecharacteristics in order to err on the sideof overestimating rather thanunderestimating plant compliance costs.EPA thus believes its raw wasteprojections will often be higher thanactual loadings (April 19, 1993memorandum to the OCPSF PublicRecord "Estimation of BAT and PSESCompliance Costs). Based on thisassessment, the Agency concluded thatland availability is not a constraint forinstalling the model treatmenttechnology (56 FR 63904; 1991Supplement to the DD, p 111-33).The three plants for which CMA

asserted the record shows insufficientcontiguous land-Plants 257, 1853, and1667--are the plants for which EPAdetermined that there is sufficient landbased on the plants' reported raw wasteconcentrations (56 FR 63904). CMAapparently overlooked this portion ofthe analysis and based its comments on

the land estimates based on the plants'projected raw waste concentrations. Asdescribed in more detail below, all ofthe plants EPA visited have more thansufficient contiguous land to install in-plant biological treatment systems tocomply with the land requirementsestimated by EPA for compliance withthe 13 remanded pretreatmentstandards.

Furthermore, based on the Agency'sdecision not to promulgate pretreatmentstandards for phenol and 2,4-dimethylphenol, the estimated landrequirements are lower for six of theeight plants visited than therequirements estimated at proposal forthese plants based on their projectedraw waste concentrations. Two plantsno longer require in-plant biologicaltreatment (257 and 2300), reducing theirland requirements to zero. Theestimated land requirements for fouradditional plants were reduced by 29,69, 75 and 74 percent (plants 814, 1667,1853 and 2485, respectively). Theestimated land requirements for theremaining two plants have not changedfrom the 1991 estimates.

Addressing the plants individually,the commenter states that the availableland claimed by the Agency for Plant257 was made up of three parcels, thatone parcel would require demolishingtwo buildings and that another parcel iscrisscrossed by railroad tracks. At thetime of the site visit, plant personnelinformed EPA that plans called for thedemolition of the two buildings inquestion and in fact demolition wasalready underway at the time of the sitevisit; the Agency reasonably concludedthat the land made available by thedemolition of these two buildingswould be available, and notes that thepretreatment standards to which thisplant was to be subject do not becomeeffective until three years after thepromulgation of today's amendments.The area made available by thedemolition of these buildings inaddition to the contiguous, open areadesignated as "A" to the left of therailroad tracks on the plot plansubmitted by the facility will more thanaccommodate EPA's land requirementestimate of 0.55 acres for Plant 257(1991 Supplement to the DD, p 111-35).This land is contiguous and is notintersected by the railroad tracks.Finally, based on the Agency's decisionnot to promulgate pretreatmentstandards for phenol and 2,4-dimethylphenol, Plant 257 no longer isprojected to install in-plant biologicaltreatment.

The commenter also claims that"part" of one of the parcels of land atPlant 1706 is unavailable because of a

36888

Federal Register / Vol. 58, No. 130 / Friday, July, 9, 1993 / Rules and Regulations

nearby flare stack. But the commenterdoes not explain, and EPA does notunderstand, how a nearby flare stackwould prevent installation of abiological treatment system. Nor did itindicate how much of the four-acreparcel in question it considered to beunavailable, and EPA has no basis toconclude that the presence of a nearbyflare stack renders unavailable the 1.8acres estimated as necessary for plant1706 to install the costed biologicaltreatment system.

The commenter also states that theAgency unrealistically utilized twoparcels of land (1 acre and 0.2 acres) tomeet the estimated land requirement of1.25 acres for Plant 1667 (CMAComment at 41). In addition to statingthat the two parcels of land are notcontiguous, the commenter states thatthe 0.2 acre parcel contains a'2 storybrick building and the 1 acre parcel hasa railroad track running through it.Again, the commenter has overlookedportions of the Agency's analysescontained in its Record. Even if therailroad track bisected the 1 acre parcel,the Agency's revised land estimate of0.38 acres based on the facility'sreported raw waste concentration (1991Supplement to the DD, p M-35) couldstillbe accommodated by either one ofthe two 0.5 acre parcels. Moreover, theAgency's Record clearly states that the2 story brick building was confirmed asnot in use and available (1991 ProposalRecord, p R01236).

In a related argument, the commenterargues that EPA has included land in itsanalysis that is unavailable because ofcontamination and related factors. EPAdisagrees with CMA's analysis of therecord, as explained below.

The commenter states that personnelfrom Plant 2756 informed EPA that theavailability of its land depended ongetting clearance from the state agencybecause contamination was suspected.However, the plant provided noinformation during EPA's site visit or incomments regarding the likelihood,nature or extent of the suspectedcontamination, the procedures involvedin obtaining clearance from the state, orthe extent to which the contaminationmight preclude the installation of abiological treatment system to complywith today's regulations within thethree years allotted. The Agency hasconservatively estimated that 32 percentof the facility's unused land (equal tothe 1.61 acres required) will be availableto accommodate the installation of in-plant biological treatment.

The commenter also states that" * * Four of the eight acres identifiedfor Plant 1853 were under investigationfor possible contamination. EPA was

told by plant personnel that theavailability of the land was dependenton the results of the investigation* * *" (CMA Comment at 41).However, the Agency's Record showsthat the uncontaminated 4-acre parcel atthe site will accommodate EPA'sestimated land requirement of 2.16acres, based on reported raw wasteconcentrations, for Plant 1853 (1991Supplement to the DD, p 111-35). Inaddition, EPA has insufficientinformation regarding the "possible"contamination to evaluate its effect oncompliance with today's amendments.

The commenter states that plantpersonnel informed EPA that of the 130acre site for Plant 2485, someunspecified portion of the plant site wasunder investigation for contaminationand 30 percent of the site wasconsidered fresh water wetlands. Since30 percent of the total plant site totals39 acres and since no accurate estimateof the extent of the contamination at the130 acre plant site could be made byplant personnel, the Agency hasconservatively estimated the amount ofland available at 20 acres or about 15percent of the total plant site, which ismore than adequate for the 6.64 acresprojected to be required at proposal.EPA also notes that no comments havebeen received to date regarding theresults of the site investigation ofpotential contamination which wasscheduled for completion in 1991.Finally, based on the Agency's decisionnot to promulgate pretreatmentstandards for phenol and 2,4-dimethylphenol, the land requirementsfor Plant 2485 have been reduced from6.64 acres to 1.68 acres.

The commenter also states that plantpersonnel at Plant 814 informed EPAthat 11 of the 13 acres EPA included inits available area was underinvestigation for possiblecontamination. Subsequentcorrespondence from Plant 814confirmed the presence ofcontamination but did not detail theextent of the contamination, only thatremediation would be necessary and"* I * a large portion of these areas

will not be available for futureconstruction other than that related toremediation * * *" (1991 ProposalRecord, p R01210). However, evenaccording to the plant's information, 2.3acres of land are not under investigationfor contamination. Although this land iscomprised of two separate parcels, thelarger of the two alone is sufficient toinstall the costed biological treatmentsystem. Based on the Agency's decisionnot to promulgate pretreatmentstandards for phenol and 2,4-dimethylphenol, this plant only requires

1.55 acres of land to install the Agency'scurrent recommended treatment system.Subtracting the smaller of the twoavailable parcels (designated as area "J"on the facility plot plan, estimated at 0.5acres) from the 2.3 acre total,approximately 1.8 contiguous,uncontaminated acres remain available,which will accommodate the currentland requirement (1991 ProposalRecord, p R01243). Moreover, only 14percent of the 11 contaminated acreswould be required to install the entiretreatment system, not counting any ofthe 2.3 acres which the commenteradmits is available. The information that"a large portion" of the 11 acres isunavailable does not provide a basis toconclude that the facility could notinstall a biological treatment system tocomply with the promulgatedpretreatment standards within threeyears.

Overall, EPA reasonably concludedthat each of the plants visited shouldhave sufficient contiguous,unobstructed, uncontaminated land toinstall the costed biological treatmentsystems. In addition, even if EPA'sanalysis indicated a lack of contiguous,available land, this would notnecessarily preclude installation of thecosted biological treatment systems.Individual pieces of a plant's treatmentsystem, including separate aerationbasins, can be physically located onnon-contiguous parcels, or on differentportions of a single parcel. In the OCPSFindustry, plant manufacturing and/ortreatment areas are sometimessegmented or separated by such thingsas utility roads, railroad tracks, canals,parking lots, warehouses, or otherunrelated parcels of land. EPA cannotperform a detailed evaluation, in anational guideline, of how individualfacilities in the industry can bestcomply with the promulgatedlimitations and standards. Especiallywith considerations as inherently plant-specific as land availability andpotential contamination andremediation requirements, EPA can onlyassess whether, for the industry as awhole, sufficient land should beavailable to comply with therequirements of the guideline. EPA hasperformed such an assessment and hasconcluded that land availability will notbe a constraint on compliance withtoday's limitations and standards. Tothe extent that an individual plantdetermines, after making a good faitheffort to use the land available to it, thatit is unable to comply with therequirements of today's rule, the plantmay apply for an FDF variance.

36889


F. Guidance for Laboratory Analysis ofComplex Matrices

Several commenters stated that theywere unable to measure some of theregulated pollutants in OCPSFwastewater at the concentrationsrequired by the regulation due to matrixinterferences. i.e., that the compositionof wastewater samples complicatesmeasurement of OCPSF-regulatedpollutants at the low levels required toshow compliance with the rule. Theysuggested that EPA provide notice thatrelief is available to the regulatedcommunity under this regulation whena permittee is unable to measurepollutants due to matrix problems.

At the time of promulgation of theOCPSF guideline in 1987, EPA foundthat for well-designed, well-operatedtreatment systems, matrix interferencesshould not present a problem. Thelimitations were based upon data thatdemonstrated that the pollutants havebeen and thus can be measured at theregulatory levels (52 FR 42563). EPA'sdetermination that the regulatedpollutants could be measured at thecompliance levels was upheld by theFifth Circuit (CMA v. EPA, 870 F.2d at231).

Since promulgation of the OCPSFguideline, the Analytical Methods Staffof the Engineering and AnalysisDivision has been assisting EPA Regionsand States in evaluating claims ofmatrix interferences and other analyticaldifficulties associated with OCPSFcompliance monitoring. Since 1990, theAnalytical Methods Staff has issued aseries of draft reports that provideguidance to control authorities andlaboratories for accommodating matrix-related problems that complicatelaboratory measurements of the analytesof interest. These documents have beenupdated and expanded in one finalpublication, the May 1993 "Guidanceon Evaluation, Resolution, andDocumentation of Analytical ProblemsAssociated with ComplianceMonitoring," (EPA 821-B-93-.001) thatis available from Mr. William A.Telliard, Chief, Analytical MethodsStaff, Engineering and Analysis Division(WH-552), USEPA, Washington, DC20460. The document includes: (1) Achecklist of laboratory data required tosupport a claim that a permittee wasunable to measure pollutants due tomatrix problems, (2) guidance foranalysts attempting to identify andquantify pollutants in wastewatersdischarged from plants manufacturingOCPSF products, (3) cost estimates forresolving matrix interferences, (4)guidance ior reviewing data from theanalysis of organic compounds using

EPA 600/1600 series analyticalmethods, (5) case histories of datasubmitted for claims of matrixinterferences under the OCPSF rule, and(6) gdance on contracting foranalytical services.

The Agency's past experience is thatnearly all matrix interference problemscan be resolved when industries andtheir laboratories apply the philosophyand techniques suggested in the draftdocuments. Based on this experience,EPA does not believe matrixinterferences will present a problem indemonstrating compliance with theOCPSF guideline.

Finally, EPA notes that this guidanceregarding matrix interference is beyondthe scope of the Fifth Circuit's remandand today's rule. As stated above, theFifth Circuit upheld EPA'sdetermination that the OCPSF-regulatedpollutants can be measured at thecompliance levels, and no issuesrelating to measurement wereremanded. The above discussion isguidance only, and it relates only toimplementation and enforcement issues;it does not provide a basis to challengetoday's amendment.

G. Guidance for the Appropriate FlowBasis for Converting Concentrations IntoMass-Based Limitations and Standards

The Passaic Valley SewerageCommissioners, referring to supportingcorrespondence from the State of NewJersey, complained about conflictingguidance and differing interpretations ofthe appropriate flow basis forcalculating the mass-based permitlimits. They requested that the Agencyclarify its guidance for (1) determiningthe appropriate flow basis forestablishing the permit limitations andstandards as well as (2) the appropriateflow basis for converting compliancemonitoring concentration data intomass-based figures.

Regarding the first issue-theappropriate flow basis for establishingpermit limits--the promulgated OCPSFeffluent limitations guidelines andstandards listed in 40 CFR 414 areconcentration-based and thus do notregulate flow. As required by theregulation, the permitting or controlauthority must multiply a reasonableestimate of a plant's regulated processwastewater discharge by theconcentration limitations to developmass limitations for each NPDES orindustrial user permit.

The appropriate process wastewaterflow to be used must be determined bythe permitting or control authority on acase-by-cas basis using currentinformation provided by the applicantand other available data. EPA strongly

urges the permit writer or controlauthority to develop an appropriateprocess wastewater flow for use incomputing the mass effluent or internalplant limitations based on waterconservation practices. The factors thatshould be considered in developing theappropriate process wastewater flowinclude: review of the component flowsto ensure that the claimed flows are, infact. process wastewater flows asdefined by the regulation; review ofplant operations to ensure that soundwater conservation practices are beingfollowed (examples includeminimization of process water uses;cascading or countercurrent washes orrinses, where possible; reuse or recycleof Intermediate process waters or treatedwastewaters at the process area and inwastewater treatment operations (e.g.,pump seals, and equipment and areawashdowns)); and review of barometriccondenser use at the process level(barometric condensers often generaterelatively large volumes of slightlycontaminated wastewater; replacementof barometric condensers with surfacecondensers can reduce wastewatervolumes significantly and result incollection of condensates that may bereturned to the process). (1987 DD, plX--9-10)

Assuming proper water conservationis being practiced, the 1987 OCPSFDevelopment Document accuratelyadvises the control authority to "use theplant's annual average processwastewater flow to convert theconcentration-based limitations intomass-based limitations" (id. at p IX-10).To clarify, the annual average flow isdefined as the average of daily flowmeasurements calculated over at least a

ear. These average flows could beased on a single year's data; however.

if available, multiple years' data arepreferable to obtain a representation ofannual average flow. The regulatedOCPSF process wastewater flows, asdefined by 40 CFR 401.11(q). are theprocess waste streams that are subject to40 CFR Part 414.

Based on current guidance issued bythe Office of Water Enforcement andCompliance. the permitting or controlauthority is advised to establish, foreach direct or indirect point sourcedischarge, a single estimate of theregulated long-term average of dailyflow measurements based on three tofive years of facility data. In the eventthat no historical or actual processwastewater flow data exists, such as fora new source, the permitting or controlauthority is advised to establish areasonable estimate of the facility'sprojected flow. Historical or projecteddaily maximum, weekly maximum, or

36890


monthly maximum flows and design-based or plant-capacity-based flows arenot recommended as appropriate basesfor determining a facility's regulatedlong-term or annual average of dailyflow measurements and correspondingmass limits. The permitting or controlauthority is advised to establish a flowrate that is expected to be representativeduring the entire term of the permit orother individual control mechanism. Ifa plant is planning for significantchanges in production during theeffective period of the permit, thepermitting or control authority mayconsider establishing multiple tiers oflimitations as a function of thesignificant, projected changes inproduction. In addition, or in thealternative, a permit may be modifiedduring its term, either at the request ofthe permittee (or another interestedparty) or on EPA's initiative, to increaseor decrease the flow basis in response toa significant change in production (40CFR 124.5,122.62). A change inproduction could be an "alteration" ofthe permitted activity or "newinformation" that would provide thebasis for a permit modification (40 CFR122.62(a) (1), (2)).

Guidance for determining appropriateprocess wastewater flow is presented inseveral documents published by theEPA Office of Wastewater Enforcementand Compliance, Washington, DC:"Guidance Manual for the Use ofProduction-Based PretreatmentStandards and the CombinedWastestream Formula," 1985 (NTISOrder No. PB92-114438) and "TrainingManual for NPDES Permit Writers, 1993(EPA 833-B-93-003).

Confusion as to the recommendedbasis for determining appropriateprocess wastewater flow'has arisen,however, due to several OCPSFguidance memoranda that presentguidance that is In conflict with theguidance presented in the OCPSFpreamble and the above-mentionedguidance documents. Specifically, twoEPA guidance memoranda recommend,as a basis for establishing long-termaverage flow, that the permit writer orcontrol authority use "the highestmonthly average flow during the pasttwelve (12) months or the highest yearlymean of the twelve monthly averageflows during the past five (5) years."These incorrect examples were listed inthe February 16, 1989 memorandum toRegional Water Management DivisionDirectors and NPDES Authorized StateDirectors from James R. Elder, Director,Office of Water Enforcement andPermits, entitled "NPDES PermittingStrategy for OCPSF Direct Dischargers"(pp. 29, 40, & 44), and in the October 12,

1988 memorandum to Regional WaterManagement Division Directors andNPDES State Directors from Mr. Elderentitled "Questions and AnswersRegarding the OCPSF EffluentLimitations Guidelines" (p. 4). Thisguidance establishes an Inappropriatebasis for determining permit limitsbecause the promulgated OCPSFmaximum daily and maximum monthlyaverage limitations were derived bymultiplying the long-term averageperformance level of well-designed,well-operated treatment systems by therespective variability factors for thetreatment system. The variability factorsalready include, among othercomponents, the variability associatedwith day-to-day and month-to-monthproduction and flow variations. As aresult, the OCPSF limits and standardsare, in general, considerably lessstringent than the long-term averagesachieved by the plants on which thelimits and standards were based, andplants that design their operations andtreatment systems to achieve the long-term averages for individual pollutantsshould be able to achieve the OCPSFlimits and standards even during high-flow days and months. The data fromany given day or month may not berepresentative of the plant's annual orlong-term flow. Use of the highestmonthly mean to set permit limitswould "double count" the effect of flowvariability, since the potential for highflow periods is already accounted for inthe promulgated limits and standards.The approach presented in the twomemoranda from Mr. Elder results in anoverly generous permit limit. Therefore,the time period of the measure ofproduction or flow should correspondto the time period used to derive thepromulgated limitations, which is anannual average or long-term averagemeasure.

Regarding the second issue-thecorrect flow basis for determiningcompliance--the Agency intends thatcompliance with the OCPSF standardsshould be evaluated based on the actualtotal applicable OCPSF-regulated flowdischarged during the period forcollecting the effluent sample, typically24 hours. The cumulative 24-hour flowcorresponding to the day on whichsampling is performed, when combinedwith concentration data from 24-hoursampling, gives the best indication ofthe actual mass of pollutants dischargedon a given day. The OCPSF mass-basedpermit limits are calculated using theregulated long-term or annual average ofdaily flow measurements, adjusteddownward as appropriate based onpotential for flow reduction, as

discussed above. The limits in 40 CFRpart 414 are expressed as maximum forany one day and maximum for monthlyaverage values. Since the limits in thepermits are mass-based, the compliancedata must also be mass-based. A dailymass value is defined as the total massdischarged over a 24-hour period(unless the operating day is less than 24hours). Similarly, the monthly averageis derived from averaging the availabledaily mass values in each calendarmonth. Compliance with the mass-basedlimits should be based on the actualtotal applicable OCPSF-regulated flowdischarged on the day of sampling, noton the long-term average flow rate thatprovided the basis for establishing thepermit limitations and standards.

Therefore, to determine compliancefor OCPSF facilities, the measuredconcentration of the pollutant inquestion in the effluent sample shouldbe multiplied by the total applicableOCPSF-regulated flow during theeffluent sampling period. For example,if analytical data from a 24-hour sampleperiod for a particular plantdemonstrates a pollutant concentrationof 0.055 mg/l, and the measured processwastewater flow for the same 24-hourperiod is 0.600 million gallons, then thelant's reported mass compliance valueor that day is 0.275 pounds of the

pollutant.EPA notes that this guidance

regardingthe proper flow basis isbeyond the scope of the Fifth Circuit'sremand and today's rule. This guidancesimply addresses conflicts in existingguidance and reaffirms that thecontemporaneous guidance presented inthe 1987 OCPSF DevelopmentDocument correctly reflects EPA'sjudgment regarding appropriateimplementation of the OCPSF guideline.The above discussion is guidance only,and it relates only to implementationand enforcement issues; it does notprovide a basis to challenge today'samendments.

IX. Executive Order 12291Executive Order 12291 requires EPA

and other agencies to perform regulatoryanalyses of major regulations. Majorrules are those which impose a cost onthe economy of $100 million or moreannually or have certain other economicimpacts. This action is not a major rulebecause the estimated cost of today'srule is $78 million annually. This figurediffers from the $61 million costpresented in the December 6, 1991proposal because: (1) The costs reflectthe shift in discharge status for 14 plantsthat were direct discharge facilities in1987 but now discharge to POTWs; (2)the costs reflect the revised projection

36891

Federal Register / Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regdations

that 47 direct discharge facilities aresubject to the Subpart J BAT limitationsrather than the 23 facilities estimated in1987; (3) the costs associated withcompliance with the pretreatmentstandards for phenol and 2,4-dimethylphenol have been eliminatedsince EPA is not promulgatingpretreatment standards for thesepollutants; (4) the cost is expressed in1992 dollars rather than 1986 dollars aswas done at proposal, because EPA hasconcluded that this approach moreaccurately reflects the cost of today'srule; and (5) the cost presented todayreflects the total projected cost of themodel in-plant biological treatmentsystems that provide the technologybasis for the promulgated limits andstandards, rather than simply theincremental cost of today's rule over thecost of the 1987 OCPSF guideline. Thislast point is explained further inresponse to the comments of Allied-Signal, Inc. Today's rule meets none ofthe criteria of a major rule as set forthin section 1(b) of the Executive Order.This rule was submitted to the Office ofManagement and Budget for review.

X. Regulatory Flexibility AnalysisThe Regulatory Flexibility Act. 5

U.S.C. 601 et seq., requires EPA andother agencies to prepare a finalregulatory flexibility analysis forregulations that have a significantimpact on a substantial number of smallentities. As noted in Section V.E., theregulatory flexibility analysis for today'sfinal rule, as in the 1987 final rule,examined whether small plants, asdefined by a plant production thresholdof 5 million pounds, aredisproportionately affected by theregulation. The Agency's assessmentconcludes that no change in the smallplant analysis findings is necessary.That is to say, the regulation continuesto set BAT equal to BPT for plants withproduction less than or equal to 5million pounds of OCPSF products peryear, and makes no special provision forsmall plants complying with PSESstandards.

X1. Paperwork Reduction ActIn accordance with the Paperwork

Reduction Act of 1980, 44 U.S.C. 3500et seq., EPA must submit a copy of anyrule that contains a collection-of-information requirement to the Directorof the Office of Management and Budgetfor review and approval. This rulecontains no additional informationcollection requirements beyond thosealready required by 40 CFR part 403 and40 CFR part 122. and therefore thereview requirement of the PaperworkReduction Act is not applicable.

List of Subjects In 40 CFX Part 414

Organic chemicals manufacturing,Plastics manufacturing, Synthetic fibersmanufacturing, Water pollution control,Water treatment and disposal.

Dated: May 28. 1993.Carol M. Browner,Administrator.

For the reasons set out in thepreamble, 40 CFR part 414 is amendedas set forth below:

PART 414-ORGANIC CHEMICALS,PLASTICS, AND SYNTHETIC FIBERS

1. The authority citation for part 414continues to read as follows:

Authority: Socs. 301, 304, 306, 307, and501, Public Law 92-500, 86 Stat. 816, PublicLaw 95-217, 91 Stat. 156, Public Law 100-4. 101 Stat. 7 (33 U.S.C. 1311, 1314, 1316,1317, and 1361).

2. Sections 414.25, 414.35, 414.45.414.55, 414.65, 414.75, and 414.85(including the tables) are revised to readas follows:

§414. Pretreatmnt standards forexisting sources (PSES) ,

Except as provided in 40 CFR 403.7and 403.13, any existing source subjectto this subpart which introducespollutants into a publicly ownedtreatment works must comply with 40CFR part 403 and achieve discharges inaccordance with § 414.111.

3. Sections 414.26, 414.36, 414.46,414.56. 414.66, 414.76, and 414.86 arerevised to read as follows:

1414,... Praetrent standards for newsources (PSNS).

Except as provided in 40 CFR 403.7any new source subject to this subpartwhich introduces pollutants Into apublicly owned treatment works mustcomply with 40 CFR part 403 andachieve discharges in accordance with§414.111.

4. In § 414.91, paragraphs (b)(1) and(b)(2) are removed and the secondsentence of the introductory text ofparagraph (b), the flush paragraph ofparagraph (b), and the table are revisedand combined to form one paragraph toread as follows:

1414.91 Toxic poihtmnt effluentlilmlatlone and etandurds for directdischarge point sourcee that ue end-of-pipe bIological treatment.

(b) * The metal-bearing wastestreams and cyanide-bearing wastestreams are defined as those wastestreams listed in Appendix A of thispart, plus any additional OCPSF processwastewater streams identified by the

permitting authority on a caie-by-casebasis as metal or cyanide bearing basedupon a determination that such streamscontain significant amounts of thepollutants identified above. Any suchstreams designated as metal or cyanidebearing must be treated independentlyof other metal or cyanide bearing wastestreams unless the permitting authoritydetermines that the combination of suchstreams, prior to treatment, with theAppendix A waste streams will result insubstantial reduction of thesepollutants. This determination must bebased upon a review of relevantengineering, production, and samplingand analysis information.

Effluent linltallonsBAT and NSPS I

Effluent charactoris- Maximumtics Maximum for forfor any ay

average

Acenaphthene .......... 59 22Acenap" ne ........ 59 22Acrylonitrile ............... 242 96Anthracene ............. 59 22Benzene .................. 136 37Benzo(a)anthracene. 59 223,4-

Benzofluorantfmen 61 23Benzo(k)fluoraithene 59 22Benzo(a)pyrene ........ 61 23BIs(2-ethythexyl)

phthalate ............... 279 103Carbon Tetrachloldde 38 18Chlorobenzene ......... 28 15Chloroethane. 268 104Chloroform ................ 46 212-Chlorophenol ........ 96 31Chrysene ................. 59 22Di-n-butyl phthalate .. 57 271,2-Dichlorobonzene 163 771,3-Dichlorobenzene 44 311,4-Dichlorobenzene 28 151,1-Dichloroethane ... 59 221,2-Dlchloroethane ... 211 681,1-Dlchloroethylone 25 161,2-trans-

Dichloroethylene ... 54 212,4-Dichloropheno ... 112 391,2-Dlchloropropne. 230 1531,3-

Dlchloropropytene. 44 29Dlethyl phthalate ...... 203 812,4-Dimethylphenol .. 36 18Dimethyl phthalale ... 47 194,6-Dlnltro-o-creol ... 277 782,4-Dkophono 123 712,4-Dlnitrotoluene ..... 285 1132,6-Dinltroto.uene.... 641 255Ethylbenzene ......... 108 32Fluoranthen. 68 25Fluorene ................. 59 22Hexachlorobnzene . 28 15Hexachlombutadlene 49 20Hexachlroethane .... 54 21Methyl Chlode ........ 190 88Methylene Chlotd.. 89 40Naphielne --- 59 22NI1robenzene ........... 66 27

36892


Effluent lhrtadensBAT and NSPS I

Effluent characterls- Maximumtics Maidmul for for

for any aawyone day motl

average

2-NltMrophenol 69 414-Nitrophonc ........... 124 72Phanthrene ........... 59 22Phenol ...................... 26 15Pyrene ...................... 67 25Teyrachlorothyeno.. 56 22Toluen .................. 80 26Total Chromium. 2,770 1,110Total Copper ......... 3,380 1.450Total Cyanide .......... 1,200 420Total Lead ................ 690 320Total Nickel .............. 3,980 1.690Total Zinc2 ............... 2,810 1,0501.2,4-

Trchlorobenzene 140 681,1,1-Trkchloroehane 54 211,1,2-Trlchloroethene 54 21Trichloroethylene ...... 54 21Vinyl Chloride .......... 268 104

SAl units are micrograms per iter.2 Total Zinc for Rayon Fiber Mamnfact

that uses the viscose process and AcrylicFiber Manufacture that uses the zinc chloride/solvent process Is 6,796 ig and 3,325 ItU1for maximum for any one day and maximumfor rnnly average, respectively.

5. In S 414.101, paragraphs (b)(1) and(b)(2) are removed and the secondsentence of the introductory text ofparagraph (b), the flush paragraph ofparagraph (b), and the table are revisedand combined'to form one paragraph toread as follows:

§414.101 Toxic pofiutant effluentlimitations and standards for dliactdischarge point sources that do not useend-of-pipe biological treastent* /* * aI a

(b) * * The metal-bearing wastestreams and cyanide-bearing wastestreams are defined as those wastestreams listed In Appendix A of thispart, plus any additional OCPSF processwastewater streams identified by thepermitting authority on a case-by-casebasis as metal or cyanide bearing basedupon a determination that such streamscontain significant amounts of thepollutants iAentified above. Any suchstreams designated as metal or cyanidebearing must be treated independentlyof other metal or cyanide bearing wastestreams unless the permitting authoritydetermines that the combination of suchstreams, prior to treatment, with theAppendix A waste streams will result insubstantial reduction of theseEollutants. This determination must be

ased upon a review of relevantengineering, production, and samplingand analysis information.

BAT effluent imite-lions and NSPS'

Effluent charactes- Maximumtics Maximum for

for any monone day average

Acenaphthene ..........Acenaphthylene .......Acronte ........Anthracene ...............Benzene ...................Benzo(a)anhracene.3,4-

BenzofluorantheneBenzo(k)fluorantheneBenzo(a)pyrone ........lls(2-9thyltwxyl)

phthalato ...............Qvbo TetrachlordeChlorobenzene .........Chloroethane ............Chloroform ...............Chrysene ............Di-n-butyl phthalate..12-Dichlorobenzone1,3-Dlchlorobenzene1,4-Dichlobenzne1,1-Dichloroethane ...1,2-Dlchloroethane ...1,1-Dichloroethylene1,2-trans-

Dichloroethylne1,2-Dichloropropane.1,3-

Dlchloropropylene.Diethyl phihalato ......2,4-Dlmethylphenol ..Dimethyl phthalate ...4,6-Dinltro-o-oresol ...2,4 9 ....trophenEthylbenzene ..........Fluoranthene ............Fluorene ...................Hexachlorobenzne.HexachlorobutadleneHexachloroethane ....Methyl Chloride ........Methylene Chloride ..Naphthalene .............Nitrobenzene ............2-Nitrophenol ............4-Nitrophenol ..........Phenanthrene ...........Phenol ...................Pyrene ...............Tetrachloroethylene..Toluene ....................Total Chroium ........Total CopperTotal Cyanide ...........Total Lead ................Total Nickel .............Total Zinc' ...............1,2,4-

Tdchlorobenzne..1,1,1-Trtchloroethane1,1,2-TrchloroethaneTrkchloroethylne ......Vinyl Chloride

4747

23247

13447

464748

258380380295325

4743

79438038059

57460

66794

7941134747

2774,291

3805447

79438079429517047

6,402231576474748

16474

2,7703,3801,200

6903,9802,610

79459

12769

172

201920

951421421101111920

19614214222

18022

25196

19646191978

1,2071422219

1961421961103619

2,23765

1621919205228

1,1101,450

420320

1,6901,050

19622322697

I Al units are micrograms per liter.

2Total Zinc for Rayon Fiber Manufacturethat uses the vlsoose process and Ac~1lcFibers Manufacture that uses the zincchlorlde/solvent process Is 6,796 pIg/ and3,325 lig9 for maximum for any one day andmaximum for monthly average, respectively.

6. Part 414 is amended by addingsubpart K to read as follows:

Subpart K-Indirect Discharge PointSources

Sec414.110 Applicability; description of the

subcategory of Indirect discharge pointsources.

414.111 Toxic pollutant standards forindirect discharge point sources.

Subpart K-indirect Discharge PointSources

$414.110 Applicability; description of thesubcategory of indirect discharge pointsources.

The provisions of this subpart areapplicable to the process wastewaterdischarges resulting from themanufacture of the OCPSF products andproduct groups defined by § 414.11 fromany indirect discharge point source.

§414.111 Toxic pollutant standards forIndirect discharge point sources.

(a) Any point source subject to thissubpart must achieve discharges notexceeding the quantity (mass)determined by multiplying the processwastewater flow subject to this subparttimes the concentration listed in the.following table.

(b) In the case of lead, zinc, and totalcyanide the discharge quantity (mass)shall be determined by multiplying theconcentrations listed in the followingtable for these pollutants times the flowfrom metal-bearing waste streams formetals and times the flow from thecyanide-bearing waste streams for totalcyanide. The metal-bearing wastestreams and cyanide-bearing wastestreams are defined as those wastestreams listed in Appendix A of thispart, plus any additional OCPSF processwastewater streams identified by thecontrol authority on a case-by-case basisas metal or cyanide bearing based upona determination that such streamscontain significant amounts of thepollutants identified above. Any suchstreams designated as metal or cyanidebearing must be treated independentlyof other metal or cyanide bearing wastestreams unless the control authoritydetermines that the combination of suchstreams, prior to treatment, with theAppendix A waste streams will result insubstantial reduction of thesepollutants. This determination must be

ased upon a review of relevantengineering, production, and samplingand analysis information.

36893

36894 Federal Register / Vol. 58, No. 130 / Friday, July 9, 1993 / Rules and Regulations

PSES and PSNS I

Effluent characteds- maximum Maximumtics for any

one day average

Acenaphthene ..........Anthracene ...............Benzene ...................Bis(2-ethylhexyl)

phthalate ...............Carbon TetrachlorideChlorobenzene .........Chloroethane ............Chloroform ................Di-n-butyl phthalate1,2-Dichlorobenzene1,3-Dlchlorobenzene1,4-Dlchlorobenzene1,1-Dlchloroethane ...1,2-Dlchloroethane ...1,1 -Dichloroethylene1,2-trans-

Dichloroethylene ...1,2-Dichloropropane.1,3-

Dichloropropylene.Dlethyl phthalate ......Dimethyl phthalate ...4,6-Dinitro-o-cresol ...Ethylbenzene ............Fluoranthene ............Fluorene ...................HexachlorobenzeneHexachlorobutadleneHexachloroethane ....Methyl Chloride ........Methylene Chloride ..Naphthalene .............Nitrobenzene ............2-Nitrophenol ............4-Nitrophenol ............Phenanthrene ...........Pyrene ......................Tetrachloroethylene ..Toluene ....................Total Cyanide ...........Total Lead ................Total Zinc 2 ... ............1,2,4-

Tdchlorobenzene1,1,1 -Tdchloroethane1,1,2-TdchloroethaneTrichloroethylene ......Vinyl Chloride ...........

4747

134

25838038029532543

79438038059

57460

66794

79411347

277380

5447

79438079429517047

6,402231576

4748

16474

1,200690

2,610

79459

12769

172

191957

9514214211011120

19614214222

18022

25196

196461978

14222.19

1961421961103619

2,23765

16219205228

420320

1,050

19622322697

I All units are micrograms per liter."Total Zinc for Rayon Fiber Manufacture

that uses the viscose process and AcrylicFiber Manufacture that uses the zinc chloride/solvent process Is 6,796 lgfL and 3,325 pg/1for maximum for any one day and maximumfor monthly average, respectively.

[FR Doc. 93-16313 Filed 7-8-93; 8:45 am]BILUNG CODE uee-W-P

DEPARTMENT OF TRANSPORTATION

Federal Transit Administration

49 CFR Part 604[Docket No. 92-ElRIN'2132-AA40

Charter Services DemonstrationProgramAGENCY: Federal Transit Administration,DOT.ACTION: Final rule.

SUMMARY: Section 3040 of theIntermodal Surface TransportationEfficiency Act (ISTEA) directs theFederal Transit Administration (FTA) toissue regulations establishing ademonstration program which wouldpermit transit operators to providecharter services for purpose of meetingthe transit needs of government, civic,charitable, and other communityactivities which otherwise would not beserved in a cost effective and efficientmanner. Section 3040 also requires theFTA to consult with a boardrepresenting public transit operatorsand privately owned charter services.Today's final rule describes thedemonstration program that the FTAhas developed in consultation with acommittee equally representative of thepublic and private sectors. During thedemonstration program, recipients ineight sites within four States will beallowed to provide direct charter serviceas determined by local officialsaccording to certain set criteria.EFFECTIVE DATE: August 9, 1993.FOR FURTHER INFORMATION CONTACT:Rita Daguillard, Attorney-Advisor, FTAOffice of Chief Counsel, (202) 366-1936,or Rosemary Woods, FTA Office ofPrivate Sector Initiatives, (202) 366-1666,SUPPLEMENTARY INFORMATION:

I. BackgroundOn April 13, 1987, the FTA, then the

Urban Mass TransportationAdministration, revised its charterservice regulation, 49 CFR part 604. Theprinciple underlying this regulation isthat federally funded equipment andfacilities may not be used to competeunfairly with private charter operators,in keeping with sections 3(f) and12(c)(6) of the Federal Transit Act, asamended (49 U.S.C. 1602(f) and1608(c)(6)). When the regulation wentinto effect on May 13, 1987, it wassubject to five limited exceptions, setout in 49 CFR 604.9. Under theseexceptions, a recipient of Federal fundsmay provide charter services if: (1) there

is no willing and able private charteroperator; (2) the private charter operatordoes not have the capacity needed for aparticular charter trip; (3) the privatecharter operator is unable to provideequipment accessible to the elderly andpersons with disabilities; (4) is non-urbanized areas, the charter service thatwould be provided would result in ahardship on users; or, (5) private charteroperators are not capable of providingservice for special events.

On December 22, 1987, the Presidentsigned the Department of Transportationand Related Agencies AppropriationsAct, 1988 (Pub. L. 100-202, 101 Stat.1329; hereafter the "FY 1988 Act"). Inthe Conference Report accompanyingthe FY 1988 Act, the FTA was directedto amend its charter service regulationto "permit non-profit social serviceagencies to seek bids for charter servicefrom publicly funded operators." (Conf.Rept., Committee Print accompanyingDepartment of Transportation andRelated Agencies Appropriations Act,1988, 100th Cong., 1st Sess. 62). Thisreport suggests that "[tihese non-profitagencies * * * be limited togovernment entities and those entitiessubject to sections 501(c) 1,3,5[sicl and19 of the Internal Revenue Code." TheReport recommends that "[in suchcases, the public operator * * * berequired to identify to the charteringorganizations any private operator tathas notified it of its willingness andability to provide comparable charterservice."

Further to this congressionaldirective, the FTA amended its charterregulation on December 30, 1988, toprovide three additional exceptions tothe general prohibition on the use offederally funded equipment andfacilities for charter service (53 FR53348).

The first exception allows the use ofFTA-funded equipment and facilities fordirect charter service with non-profitsocial service agencies that aregovernmental entities or organizationsexempt from taxation under InternalRevenue Code 501(c) (1), (3), (4) and(19), provided that the agency iscontracting for service for persons withdisabilities; is a recipient of fulids undercertain U.S. Department of Health andHuman Service ("USDHHS") programs;or has been State-certified according tothe procedure set forth in§ 604.9(b)(5)(iii) of the Charter ServiceRegulation.

The second exception provides anadditional exemption for nonurbanizedareas by allowing FTA-fundedequipment and facilities operated byrecipients in such areas to be usedincidentally in direct charter service for

36872 Federal Register Vol. No. Friday, July 9, 1993 ......George Jett or Ms. Debra Nicoll,...

Documents

Transcript of 36872 Federal Register Vol. No. Friday, July 9, 1993 ......George Jett or Ms. Debra Nicoll,...