Citation: 51 Fed. Reg. 18538 1986

Content downloaded/printed from HeinOnline (http://heinonline.org)Thu Jun 11 10:55:35 2015

-- Your use of this HeinOnline PDF indicates your acceptance of HeinOnline's Terms and Conditions of the license agreement available at http://heinonline.org/HOL/License

-- The search text of this PDF is generated from uncorrected OCR text.

Federal Register /'Vol. 51, No. 97 / Tuesday, May 20, '1986 / Rules and Regulations

ENVIRONMENTAL PROTECTIONENVIRONMENTAL PROTECTIONAGENCY

40 CFR Part 60

[AD-FRL 2915-41

Review and Amendment of Standardsof Performance for New StationarySources; Kraft Pulp Mills

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

SUMMARY: Standards of performance forkraft pulp mills were proposed onSeptember 24, 1976 (41 FR 42012), andpromulgated on February 23, 1978 (43 FR7568). On January 19, 1984, revisions tothe standards of performance for kraftpulp mills were proposed in the Federal,Register (49 FR 2448). Today's actionpromulgates these revisions andannounces the Agency's decision onother elements of the standard whichwere reviewed. The revised standardsapply to new, modified, andreconstructed kraft pulp mills, for whichconstruction was commenced afterSeptember 24, 1976. These standardsimplement Sdction 111 of the Clean AirAct and are based on theAdministrator's determination that kraftpulp mills cause, or contributesignificantly to air pollution which mayreasonably be anticipated to endangerpublic health or welfare. The intendedeffect of these standards is to require allnew, modified, and reconstructed kraftpulp mills to achieve emission levelsreflecting the best demonstrated systemof continous emission reduction,considering costs, nonair quality health,and environmental and energy impact.EFFECTIVE DATE: May 20, 1986.

Under section 307(b)(1) of the CleanAir Act, judicial review of the actionstaken by this notice is available only bythe filing of a petition for review in theU.S. Court of Appeals for the District ofColumbia Circuit within 60 days oftoday's publication of this rulte. Undersection 307(b)(2) of the Clean Air Act,the requirements that are the subject oftoday's notice may not be challengedlater in civil or criminal proceedingsbrought by EPA to enforce theserequirements.ADDRESSES: Background InformationDocument. The background informationdocument (BID) for the promulgatedstandards may be obtained from theU.S. EPA Library (MD,-35), ResearchTriangle Park, North Carolina 27711,telephone number (919) 541-2777. Pleaserefer to "Kraft Pulp Mills-BackgroundInformation for Promulgated Revisionsto Standards" [EPA 450/3-85-020]. The

BID contains: (1) A summary of all thepublic comments made on the proposedstandards and, the: Administrator'sresponse to the comments, (2) asummary of the changes made to thestandards since proposal, and (3), thefinal Environmental Impact Statementwhich summarizes the impacts of therevisions.

Docket. A docket, number A-82-36,containing information considered byEPA in development of the promulgatedstandards, is available for publicinspection between 8:00 a.m. and 4:00p.m., Monday through Friday, at EPA'sCentral Docket Section (A-130), WestTower Lobby, Gallery 1, 401 M StreetSW., Washington, DC 20460. Areasonable. fee may be charged forcopying.FOR FURTHER INFORMATION CONTACT:

For Policy Questions: Mr. Doug Bell,Standards Development Branch,Emission Standards and EngineeringDivision (MD-13), U.S. EnvironmentalProtection Agency, Research TrianglePark, NC 27711, telephone number(919) 541-5578.

For Technical Questions: Mr. KennethDurkee or Mr. James Eddinger,Industrial Studies Branch, EmissionStandards and Engineering Division(MD-13), U.S. EnvironmentalProtection Agency, Research TrianglePark, NC 27711, telephone number(919) 541-5601.

SUPPLEMENTARY INFORMATION:

I. The Standards

Standards of performance for newsources established under section 111 ofthe Clean Air Act reflect:. . application of the best technological

system of continuous emission reductionwhich (taking into-consideration the cost ofachieving such emission reduction, and anynonair quality health and environmentalimpact and energy requirements) theAdministrator deterrnines has beenadequately demonstrated [section 111(a)(1)].

For convenience, this will be referred toas "best demonstrated technology" or"BDT."

On September 24, 1976, new sourceperformance standards (NSPS) wereproposed for kraft pulp mills undersection 111 of the Clean Air Act (41 FR42012). These regulations werepromulgated on February 23, 1978 (43 FR7568). The standards limit emissions ofparticulate matter (PM) and-totalreduced sulfur (TRS) from new ormodified recovery furnaces, smeltdissolving tanks, lime kilns, digestersystems, multiple effect evapbratorsystems, black liquor oxidation systems,brown stock washer systems, andcondensate stripper systems that have

been constructed, modified, or !reconstructed after September 24, 1976.

The PM emission limits are: 0.10grams per dry standard cubic meter (g/dscm) at 8 percent oxygen for recoveryfurnaces: 0.10 grams per kilogram ofblack liquor solids (dry weight) (g/kgBLS) for smelt dissolving tanks; 0.15 g/dscm at 10 percent oxygen for lime kilnsburning gas; and 0.30 g/dscm at 10percent oxygen for lime kilns burningoil. Visible emissions from recoveryfurnaces are limited to 35 percentopacity..The TRS emission limits are: 5 partsper million by volume (ppmv) at 8percent oxygen from straight kraftrecovery furnaces; 25 ppmv at 8 percentoxygen from cross recovery furnaces; 8ppmv at 10 percent oxygen from limekilns, and 5 ppmv at the actual oxygencontent of the untreated gas stream fromdigester systems, multiple-effectevaporator systems, brown stockwasher systems, black liquor oxidationsystems, and condensate strippersystems. The TRS emissions from smeltdissolving tanks are limited to 0.0084 g/kg BLS.

The standards also require continuousmonitoring, recordkeeping, and excessemission reporting. The opacity ofrecovery furnace exhaust gases must bemonitored continuously, and a record ofthese measurements must bemaintained. The concentration of TRSemissions from recovery boilers andlime kilns must be monitoredcontinuously and a record of thesemeasurements must be maintained. Theincineration temperature of effluentgases from digesters, brown stockwashers multiple-effect evaporators,black liquor oxidizers, or condensatestrippers must be monitored. Finally, thegas stream pressure drop and liquidsupply pressure for any scrubbercontrolling emissions from lime kilns orsmelt dissolving tanks must becontinuously monitored. Records of 12-hour average TRS concentrations and12-hour oxygen concentrations must bemaintained on a daily basis. Quarterlyreports of excess TRS emissions, excessopacities, and inadequate incinerationtemperatures are required as well.

Today's rulemaking promulgates sixrevisions and two corrections to thestandards. These revisions will: (1)Exempt black liquor oxidation systemsfrom the standards: (2) revise theexisting TRS standard for smeltdissolving tanks from 0.0084 g TRS perkg of black liquor solids (g TRS/kg BLS)to 0.016 g TRS/kg BLS; (3) revise theunits of the TRS standard for smeltdissolving tanks; (4) delete therequirement to monitor the combustion

18538

Federal Register I Vol. 51, No. 97 / Tuesday, May 20, 1386 / Rules and Regulations

temperature in lime kilns, power boilers,or recovery furnaces; (5) change thefrequency of excess emission reportsfrom quarterly to semiannually; and (6)exempt diffusion washers from the TRSstandard for brown stock washersystems. The corrections would: (1)Require measurements of temperature,pressure drops and liquid feed rates forcontrol devices which must bemonitored must also be recorded; and(2) correct the reference in§ 60.284(d)(3)(ii)from § 60.283(a)(1)(ii) to§ 60.283(a)(1)(iii). In this secondinstance, the original standarderroneously referred to reporting ofexcess emissions when recovery boilersare used as incineration devices. Thecorrected standard refers to facilitieswhere incineration devices not subjectto Subpart BB are used for incinerationof TRS emissions from facilities subjectto the standard.

In the overall context of this sourcecategory, all of the changes to theexisting standards of performance areminor. Nevertheless, they areappropriate because they change thenumerical emission limit for smeltdissolving tanks to reflect theperformance of BDT, improve theoverall cost effectiveness of the existingstandards with little increase iii TRSemissions, and reduce reporting andrecordkeeping requirements.

1I. Summary of Environmental, Energy,and Economic Impacts

The revisions will not significantlyaffect nationwide particulate matteremissions, solid waste generation, waterquality, or energy consumption. Deletingthe TRS standard for black liquoroxidation (BLO) tanks may increaseTRS emissions from the only affectedkraft pulp mill by up to 16 tons per yearin the fifth year following proposal. Thefull increase of 16 tons would beequivalent to 42 percent of the mill'scontrolled TRS emissions, or about 0.5percent of its uncontrolled emissions.The exemption of diffusion washersfrom the TRS standard for brown stockwashers will cause no increase in TRSemissions. Changing the smelt tank TRSstandard will increase TRS emissions byabout 6 tons annually in the fifth yearfollowing proposal. This projection isbased on the assumptions that theaffected facility which failed previoustests for compliance with the originalNSPS will continue to perform as it hasin the past and that one similar affectedfacility will be constructed in the future.This increase represents 8 percent of amill's controlled TRS emissions, orabout 0.2 percent of its uncontr6lled TRSemissions.

There will be a maximum cost savingsof $500,000 associated with the removalof the TRS standard for BLO tanks. Thisprojection is based on our finding thatone mill may stop controlling TRSemissions from its BLO tank atpromulgation of the revised standards.The savings is in operating costs, andhas no capital component. The savingsin annual costs which will result fromexempting diffusion washers from theNSPS is estimated to be $610,000 in thefifth year. There will be no significantcost impacts associated with any'of theother revisions to the NSPS.

The environmental, energy, andeconomic impacts are discussed ingreater detail in the BID for thepromulgated standards, "Kraft PulpMills-Backgound Information forPromulgated Revisions to Standards"[EPA 450/3-85-020].

III. Public Participation

Prior to proposal of the standards,interested parties were advised bypublic notice in the Federal Register (48FR 12825, March 28, 1983) of a meetingof the National Air Pollution ControlTechniques Advisory Committee todiscuss the revisions recommended forproposal. This meeting was held onApril 27, 1983. The meeting was open tothe public and each attendee was givenan opportunity to comment on therevisions recommended for proposal.The proposed revisions were publishedin the Federal Register on January 19,1984 (49 FR 2448). The preamble to theproposed revisions discussed theavailability of the BID, "Review of NewSource Performance Standards for KraftPulp Mills" [EPA 450/3-83-017], whichdescribed in detail the regulatoryalternatives considered and the impactsof those alternatives. Public commentswere solicited at the time of proposal.and, when requested, copies of the BIDwere distributed to interested parties.To provide interested persons theopportunity for oral presentation ofdata, views, or arguments concerningthe proposed standards, a public hearingwas scheduled for February 21, 1984, atResearch Triangle Park, North Carolina.A hearing was not held becasue onewas not requested. The public commentperiod was from January 20, 1984, toMarch 9, 1984.

Twenty-eight (28) comment letterswere received relative to the proposedstandards of performance for kraft pulpmills. The comments have beencarefully considered and, wheredetermined to be appropriate by theAdministrator, changes have been madein the proposed standards.

IV. Significant Comments and Changesto the Proposed Standards. Comments on the proposed standardswere received from industry, Federalagencies, State and local air pollutioncontrol agencies, and trade associations.A detailed discussion of these commentsand responses can be found in the BID,which is referred to in the ADDRESSESsection of this preamble. The summaryof comments and responses in the BIDserve as the basis for the revisionswhich have been made to the standardsbetween proposal and promulgation.The major comments and responses aresummarized in this preamble. Most ofthe comment letters contained multiplecomments. The comments have beendivided into the following areas:Emission Control Technology andSelection of Emission Limits.

Emission Control Technology

Diffusion Washers

Two comments contain the findingsand resulting recommendations of astudy performed by an industry councilto quantify TRS emissions fromdiffusion washers. That study examined9 diffusion washer vents and the meanmass emission rate was found to be0.001 lb., or less, TRS per ton of air driedpulp (TADP). Such emission levels aretwo orders of magnitude less than thosefrom uncontrolled vacuum drum washersystems. Using the same cost estimatingprocedures employed by EPA for thecase of vacuum drum washer systems,the industry calculated the costeffectiveness (C/E) of further controllingthese emissions to be $240,000 per ton ofTRS removed. Three commenters saidthat those findings preclude EPA fromreasonably supporting the need tocontrol diffusion washer vent gases onan emission significance or economicbasis. They note that there would be noadvantage to setting mass emissionlimits and that imposing measurementand reporting requirements would beburdensome. Too commenters supportthe above findings and conclusions. Onecommenter noted that diffusion washersmay meet the existing standards withouta control device.

One commenter disagrees with theothers and says that diffusion washersshould not be exempted outright fromhaving TRS controls. This commenterbelieves each individual source shouldbe required to demonstrate thatemissions from its uncontrolled diffusionwashers can meet the same TRSstandards as controlled vacuumwashers.

The study submitted on TRSemissions from diffusion washers has

18539

Federal Register / Vol. 51, No. 97' / Tuesday, May 20, 1986 / Rules and Regulations

been reviewed by the Agency. TheAgency agrees that uncontrolled TRSemissions from diffusion washers areless than 0.001 lb TRS/TADP. This levelis orders of magnitude less than that ofuncontrolled vacuum drum washers (0.3lb TRS/TADP)' and is also many timeslower than the mass equivalent of theNSPS. The equivalent mass emissionrate for the 5 ppm NSPS, based on thevacuum drum washer, is about 0.09 lbTRS/TADP. Because of the low mass ofTRS emissions, controlled and the lowair volumes treated, requiring control ofTRS emissions from diffusion washersto the 5 ppm TRS level would result in aC/E in the range of $240,000 per ton ofTRS removed. Therefore, the Agencyhas determined that requiring diffusionwashers to meet the 5 ppmv TRSstandard would be unreasonable.

For several reasons, revision of theNSPS to a mass equivalent TRSstandard would also be unreasonable.As the nvailable data indicated,uncontrolled TRS emissions fromdiffusion washers are many times lowerthan the. mass equivalent of the NSPS.As such, requiring diffusion washers todemonstrate. compliance with a massequivalent NSPS would imposeunnecessary costs for testing andreporting requirements. In addition, anEPA reference sampling method wouldhave to be developed and promulgatedsince the present EPA Reference Method1 is insufficient for sampling the lowvelocity, low volume; and. cyclic gasstream emitted'fiom a diffusion. washer.

Development of a separate standardfor TRS emissions from. diffusionwashers would require. a majorcommitment of Agency resources tostudy a process. which produces verylow mass emissions. Such a standardwould have to include a controltechnology which, in this case, wouldundoubtedly be incineration and thecost has been estimated to be in therange of $240,000 per ton TRS removed.Because projected control costs are highand potential benefits are negligible, theAgency has concluded that developmentof an NSPS for TRS emissions fromdiffusion washers is not appropriate.

Noncontact Recovery Furnaces WithWet-Bottom Electrostatic. Precipitators(ESP's),

At the time th NSPS were developed.use of the direct contact furnace systemwas prevalent in the industry andavailable information indicated that thecontacting of furnace flue gases withunoxidized black liquor in direct contactevaporators was causing high levels ofTRS emissions. Therefore, the Agency'tested direct contact furnaces equippedwith BLO systems. Particulate emissions

from these sources were controlled bywet-bottom ESP's through which theoxidized black liquor was passed. Also,the Agency tested a noncontactrecovery furnace system, whicheliminates the contact of flue gas andblack liquor altogether, which in turneliminates the need for BLO equipment.This furnace system had a dry-bottom.ESP for control of particulate emissions.As a result of these tests, the BDT forcontrol of TRS emissions fromnoncontact furnace systems wasdetermined to be maintenance of propercombustion conditions and black liquorfiring rates and, for direct contactfurnace systems, was determined to bemaintenance of proper combustionconditions and oxidation of black liquor.For both 'furnace types, ESP's weredetermined to be BDT for achieving therequired limits for PM emissions.

Since the development of the NSPS,the paper industry's National Councilfor Air and Stream Improvement(NCASI) in 1978 investigated thepossible use of unoxidized black liquorin wet-bottom ESP's and concluded thatuse of unoxidized black liquor in wet-bottom ESP's would not cause violationsof the TRS emission limit. In 1979,another industry study concluded thewet-bottom ESP's were more reliableand less costly to operate than dry-bottom ESP's that were in use at thattime. Following these studies, wet-bottom ESP's utilizing unoxidized blackliquor were installed on ten noncontactrecovery furnaces subject to the NSPS.In 1982 it became apparent that some ofthem were having difficulty in achievingthe 5 ppmv TRS standard. During thesame time period, four noncontactrecovery furnaces were installed withdry-bottom ESP's. Two recoveryfurnaces of the direct contact designwere equipped with wet-bottom ESP'swhich used oxidized black liquor in thebottoms. All' six recovery furnaceswhich installed' the technology uponwhich the NSPS were based haveachieved those standards.• In early 1984, when the revisions

resulting from the 4-year review wereproposed, the extent of the problems,with the wet-bottom ESP's and potentialcorrective measures were not fullyunderstood. The NCASI was then in themidst.of a major study which wasconducted to identify the causes of TRSrelease from unoxidized black liquorand to develop means' of eliminatingexcess emissions from the ESP's. In theproposal, the Agency stated that it wasreasonable to delay completion of thereview of the TRS standard for recoveryfurnaces long enough to allow NCASIsufficient time to perform its study.

Seven different commenters agreedwith the EPA proposal to delay reviewof the existing TRS standards forrecovery furnace systems as theypertain to facilities which hive installedwet-bottom ESP's. All agreed that anypossible changes which would take intoconsideration the performance ofnoncontact recovery furnaces equippedwith wet-bottom ESP's using unoxidizedblack liquor should be delayed untilNCASI completed its studies of thesesystems. One commenter noted that ithas been demonstrated that wet-bottomESP's can achieve the existing TRSstandard. They conclude that anychanges to the current TRS standardshould pertain only to wet-bottom ESP'sand that any possible changes should bedelayed only until the NCASI study iscomplete. One commenter said that EPAshould resist any change in the existingstandards and that EPA should explorethe use of non-TRS bearing water in thewet-bottom ESP's.

Since proposal, much work has beendone by NCASI and by individualaffected firms in an attempt to fullyunderstand and correct the problem.The NCASI study has identified severalfactors which are contributing to theproblem. These include inlet bafflingdesign, liquor temperature, liquor level,degree of agitation, and liquorchemistry. To date, modifications tomitigate the first four factors have beenmade in most instances where theyappeared feasible. The results of themodifications differed from mill to milland were not always successful forreducing TRS emissions. Similarly,efforts by individual mills to control ormodify the chemistry of liquors used inthe wet ESP's have given mixed results.After making various combinations ofmodifications, some facilities haveachieved, or have come very close toachieving, the 5 ppmv TRS standard.However, according to industryassessments, several furnaces appearunable to consistently achieve betterthan 15 ppmv and some appear unableto consistently achieve better than 25ppmv while using unoxidized blackliquor in the ESP.

The EPA has reviewed available dataand the steps which industry havetaken. It is clear from this that NCASIand individual firms have expendedconsiderable resources in their attemptsto identify and correct the causes of TRSrelease from unoxidized black liquorused in wet-bottom ESP's. The Agencyagrees that the recovery furnace TRSstandard is probably not consistentlyachievable, at all sources when suchliquor is used, in the ESP's. However,based on its review of the industry

18540,

Federal Register / Vol. 51, No. 97 / Tuesday, May 20, 1986 / Rules and Regulations

studies, the factors which are causingexcess emissions, and of potentialremedies, EPA has concluded that thestandard for recovery furnace TRSemissions should not be revised. Inreaching this conclusion, the Agencyrecognizes that the decisions to installthe wet-bottom precipitators were madebased on the available industry datawhich indicated that the TRS emissionlimit would not be violated. But, therewere other options available and thoseoptions were employed at otherfacilities. Furthermore, retrofit optionsare available which will allow thesources with wet-bottom ESP's toachieve compliance with the TRSemission limit. For example, two millshave made piping changes which allowthem to use fresh water in wet-bottomESP's and the level of the NSPS for TRShas been achieved. In addition, millshave the option of converting thebottoms of their ESP's from the wet tothe dry design. Although each of theseoptions entails a retrofit with annualizedcosts ranging from $85,000 to $275,000per mill and the associated TRSreduction could be small, EPA believesthe costs of the retrofits are reasonable.When the annualized cost of installingand retrofitting a wet-bottom ESP arecompared to the annualized costs ofinitially installing a dry-bottom ESP. thenet difference in estimated annualizedcosts of retrofitting the wet-bottom ESPare reasonable and range from a savingsof $40,000 to a cost of $100,000.

In conclusion, therefore, the Agencybelieves that changes to the NSPS forkraft recovery furnaces would beinappropriate and that those mills nowout of compliance with the TRSstandard should take the necessarysteps to achieve compliance.

Degradation of Performance of ESP'sThree commenters disagree with the

Agency's conclusion in the BID (EPA450/3-83--017) that data from a 9-year-old ESP show that ESP's can reducerecovery furnace particulate emissionsto NSPS levels over a long period of timewhen they are properly maintained. Onecommenter operates the ESP to whichthe three referred and this commentersays the data show that, even withmaintenance, the ESP is not capable ofachieving NSPS consistently. Thecommenter also said that it isinappropriate to draw conclusions aboutlong-term performance of ESP's fromdata obtained from only one ESP.

A second commenter said that the -data provided by the previouscommenter clearly show an upwardtrend in emissions of PM with increasingage of the ESP and that EPA's judgment-concerning the ability of ESP's to meet

NSPS for particulate emissions over thelong term is an inappropriateinterpretation of data from a singlelocation. The commenter presented long-term data from two other sources withESP's designed to achieve emissionlevels similar or NSPS and said the datafrom all three sources showed anupward trend in particulate emissionswith increasing age of the ESP's. Thedata from all three ESP's also showedthat measured emissions, followingmajor rebuilds of the ESP's weresignificantly higher than those achievedwhen the precipitators were new. Thecommenter attributed the increasedemissions to such factors as buildupsand corrosion in duct work, plenumsand turning vanes, which can cause flowmaldistributions.

The second commenter maintains thatEPA has not thoroughly investigated theESP degradation issue in its NSPSreview. They also say that the Agencyhas not considered the costs of majorrebuilds or lost production due tounscheduled repairs in the cost-effectiveness calculations.

The problem of gradual deteriorationof ESP performance was investigatedduring the NSPS development and againduring the NSPS review. During theNSPS development, the ESP vendorsIndicated that a properly maintainedESP should not deteriorate over theexpected life of the unit. Problemsencountered are usually due tooperating the equipment at conditionsfor.which it was not designed (i.e.,higher gas volumes, higher inletloadings, or-lower inlet temperature).The main problem areas are corrosionand wire breakage.

The unit for which EPA obtained long-term particulate data, at-the time it wasinstalled, employed a new design whichminimized wire breakage. This unit wastested by EPA as part of the data basefor the NSPS. Additional data suppliedby the State agency during the NSPSdevelopment indicated that the unitconsistently achieved the NSPS level.During the NSPS review, the operator ofthis unit was again contacted to obtaininformation on maintenance costs andESP performance. The maintenancecosts for this unit had increased from240 man-hours peryear to an average of913 manhours per year. Thesemaintenance costs are higher than theestimate used by the Agency. If it couldbe shown that all of these costs areattributable to the NSPS, theincremental C/E of the NSPS is $200-$300 per ton, which is still reasonable.However, as noted, it is not clear thatthe increased maintenance costs are infact due to the NSPS. The data indicated

that after 10 years of operation, the unitwas still capable of achieving the NSPSlevel. It is true, as the one commenterpointed out, that test data indicate thatat times the unit has had emissionsabove the NSPS level. It must be pointedout that this unit is not subject to theNSPS and is only required to achieve aState regulation which is doubletheNSPS level. Therefore, this unit ismaintained to achieve the State level asopposed to the NSPS level. It is theAgency's judgment that this unit couldconsistently achieve the NSPS if thefrequency of maintenance wereincreased. The Agency's judgment issupported by the data supplied by onecommenter which shows theperformance of an ESP which is notsubject to the NSPS but which is subjectto a State standard about 25 percentlower than the NSPS. This latter unit hasbeen operating for 10 years and hasconsistently achieved the NSPS levels.

The Agency's cost estimates do notinclude the cost of major rebuilds aswas suggested by the commenters. TheESP's were widely used in the kraft pulpindustry for recovery of processchemicals prior to establishment ofNSPS and none of the information whichhas been reviewed indicates that majorrebuilds are needed more frequentlybecause of NSPS for PM. As a result ofNSPS, new ESP's are designed withmore plate area and additionalmaintenance costs for such items asreplacement of broken wires would bepossible. However, the need for majorrebuilds, to repair corrosion damage, forexample, is most likely attributable toprocess parameters, such as the flue gastemperature, and not related to thesizing of the EPS's. Since the NSPS doesnot affect the frequency of majorrebuilds, it would be inappropriate toinclude the costs of rebuilds in thecalculation of control costs.

Selection of Emission Limits

Smelt Dissolving Tanks (SDT)

Five different commenters were inagreement with EPA's decision to raisethe TRS standard for SDT. However,they said that the increase should begreater than the one which wasproposed. One commenter said thatpreliminary data from a new millindicated that the proposed level neededto be doubled. In a follow-up letter, thecommenter described the liquids beingused in their scrubbers and noted thatthey planned to try and redirect sulfide-containing recycle streams from the SDTand scrubbers. In a third letter, thecommenter said that efforts to modifytheir piping system to redirect sulfide

.19541

Federal Register / Vol. 51, No. 97 / Tuesday, May 20, 1986 / Rules and Regulations

bearing liquids away from the smeltta9 ks had been successful and that theyhad passed compliance tests. Thus, theywithdrew their request for a higher TRSlimit than that which was proposed.

A second commenter sent two lettersdescribing experience at two of its mills.The commenter said that selection of thescrubbing liquid is the only knownmethod of modifying TRS emissionsassociated with smelt tank vent gases.The commenter has examined the use ofalternative scrubbing liquids and saidthat TRS emissions exceeded thestandard even when fresh water wasused in the scrubbers at one of the mills.They said their best results at the othermill were obtained when both the smelttank scrubber and the lime kiln mudwasher showers were operated on freshwater, which the commenter considersan artificial condition for that particularmill. The commenter submittedadditional continuous monitoring dataand said the new data showedvariations similar to those in previouslysubmitted information.

A third commenter said the proposedTRS level is a move in the rightdirection, but that two of its facilitiescannot meet that level on a consistentbasis. The commenter said that variousscrubbing media had been tried but thatno controllable process or controltechnology operating conditions hadbeen identified which could limit TRSemissions from smelt tank vents. Thiscommenter said its data (from 50 hoursof continuous monitoring) supported aTRS limit well above the proposed level.Two comments by industry tradeassociations supported the first threecommenters' observations andcomments.

Emissions of TRS compounds aregoverned by the concentration ofreduced sulfur compounds either in thesmelt from the recovery furnace or in thewater in the smelt tank. Additional TRSmay be introduced if liquidscontaminated with TRS compounds areintroduced to the scrubbers used forcontrol of PM. There is no means ofcontrolling the introduction of reducedsulfur compounds via the smelt from therecovery furnace. However, theintroduction of additional TRScompounds to the vent gases can beprevented, substantially reduced, by theselection of liquids to be used in thetanks and scrubbers. Preventing theintroduction of TRS-contaminatedliquids to the SDT system is the basis ofBDT, which is, "to use a liquid that islow in sulfides and TRS compounds-such as fresh water or recycled waterfrom the lime mud washer-in the smelttank and particulate control device" (49

FR 2448). The date base used in thereview to revise NSPS for TRS from0.0084 g/kg BLS to.0.016 g/kg BLSincludes two test reports from one millwhich failed to comply with the 0.0084g/kg emission limit. The operators of themill indicated that they had used freshwater in their mud washers and that theweak wash had been used in both thesmelt tank and scrubbers. Use of thesetypes of liquids is considered to be BDTfor reducing TRS emissions. They thenexperimented with various liquids in thescrubber, including fresh water. Since noreasons for the higher TRS emissions.could be identified, and since thesources were applying BDT, theemission limit for TRS emissions wasproposed to be raised to 0.016 g/kg BLSto reflect the results of these compliancetests.

Information supplied by the firstcommenter showed that relatively smallflows of TRS-contaminated recyclestreams were being introduced to theweak wash storage tanks andsubsequently to the SDT's andscrubbers. The operators of the millwere reluctant to remove the recyclestreams because they did not want toincrease either water usage or theamount of wastewater to be treated.When the mill used BDT and removedthe TRS contaminated liquids from thesmelt dissolving system, they did passtests for compliance with the currentTRS standard. After passing the test, thecommenter withdrew his initialcomment that the TRS limit should begreater than 0.016 g/kg BLS.

The data supplied by the secondcommenter for one of their mills showedthat they had been using contaminatedcondensate in their SDT scrubberrecycle system. When the condensatewas replaced with fresh water, TRSemissions began to drop. Later datafrom the same source showed that useof boiler blowdown (which is very lowin residual sulfides) in the systemreduced TRS emissions to NSPS levels.The commenter said that the best resultswere obtained when lime mud shower(which produces the weak wash used inthe SDT) and SDT scrubber wereoperated on fresh water, but that thisrepresents an artificial conditionestablished solely to minimize TRSemissions. They say that operating inthis manner causes an unusually highhydraulic loading on the effluenttreatment system. The artificialcondition described for the plant is whatthe Agency considers to be BDT. Whilethe plant may. not operate this way now,the Agency has concluded that usingfresh water, or other liquids low in TRScompounds, to reduce TRS emissions is

technically feasible and reasonable froma cost standpoint. The Agency continuesto believe that if BDT is implemented,the TRS limit of 0.016 g TRS/kg BLS canbe met.

The EPA disagrees with the secondcommenter's statement that selection ofscrubbing liquid is the only knownmethod of modifying TRS emissionsassociated with SDT vent gases. Themill which they were discussing hadproblems with excess TRS emissionsand began testing different scrubbingliquids. Initially, they had been usingweak white liquor, which is known toremove some polar compounds, such asH2S. Thus, it is not surprising that TRSemissions increased when water, andvarious other liquids were substituted.However, the scrubber was installed forremoval of PM, not TRS. The key pointis that BDT for TRS is aimed atpreventing introduction of TRS to ventgases by the dissolving liquid orscrubbing medium.

Both the second and third commenterssaid that the ranges in their TRSmonitoring data were indicative that theproposed standard cannot be met on aconsistent basis. The third commenterdid not submit enough information forthe Agency to draw any conclusions. Itis noted that the two tanks to whichthey referred are not subject to NSPSand the comment letter Auggested thatwater used in the SDT's was not of thequality required by BDT. The secondcommenter's data showed variation inTRS concentrations for individualsamples, but when the data points wereaveraged, as they would be for acompliance test report, the emissionlevels were below the proposed TRSlimitations.

Two commenters object to relaxingthe existing TRS standard for SDTbecause of one or two failures toachieve compliance. One commentersuggests an alternative of allowingexemptions based upon site-specificstudies and a requirement that BestAvailable Control Technology beemployed.

These suggestions are inconsistentwith the basis of the NSPS. An emissionlimit must be set at such a level that anyfacility which employs BDT can achievethat emission level during a performancetest. A facility which was employingBDT failed two performance tests. Inselecting an emission limit, variability ofavailable test data must be taken intoconsideration. The Agency proposed torevise the TRS standard from 0.0084 g/kg BLS to 0.016 g/kg BLS in order that allfacilities using BDT can meet the TRSstandard.

18542

Federal Register / Vol. 51, No. 97 / Tuesday, May 20, 1986/ Rules and Regulatiuns

Lime Kilns

Five different commenters suggest thecurrent TRS standard for lime kilnsneeds to be revisedto reflect the resultsof continuous monitoring. Onecommenter says the monitoring datafrom two of its NSPS facilities indicatethat the standard needs to be revised toallow for exceedance of the TRS limit 3percent of the reporting time to allow fornormal variations in operatingconditions. The commenter lists fourfactors which can influence TRSemissions from the kiln stack: (1) Kilnfiring conditions; (2] treatment ofnoncondensable gases; (3] source ofwater used at the particulate scrubber;and (4] porosity of the mud at the filter(which controls oxidation of the residualsulfidecontent). This commenter statedthat TRS emissions associated with thefirst three factors are straightforwardand the control options are understood,but that-the control of mud porosity atthe filter is not completely understood.

One commenter stated that thecurrent TRS standard can be met whenthe kiln and associated systems areoperating normally, but that the natureof the process is such that unavoidableirregularities which can affect TRSemissions will occur 10 percent or lessof the total operating time. He says thatshort-term "blips" or "spikes" areadequately reckoned by the averagingtime, but that a 4 percent allowance forexcess emissions appears reasonable-forthose infrequent, medium-term TRSexcursions which are beyond the controlof the operators. The commenter statedhe is unaware of any evidence that theuse of caustic soda (to control excessemissions) is effective and/or costeffective. He also doubts that lime mudoxidation is a cost-effective techniquefor controlling excess TRS emissions.

One commenter has been unable toexplain variations in-data from acertified continuous monitoring system.The commenter stated that 12-houraverages from this particular facilityrange from 2 to 30 ppmv TRS and thecommenter is concerned that it may notbe possible to meet the 8 ppmv limitcontinuously.

One commenter says that as moreTRS monitoring systems come on-line,there will be additional informationwhich will be useful in determiningwhether or not the current standard isappropriate. The commenter suggeststhat EPA should evaluate availablecontinuous monitoring data from limekilns equipped with wet scrubbersbefore making any final decisions on anNSPS.

Many of the comments were promptedby the requirement that lime kilns

subject to the NSPS install and operatecontinuous emission monitors (CEM's)to measure TRS emission by July 20,1984. After considering the comments,the Agency determined that it would beappropriate to obtain additional data.Subsequently, the first 6 months' CEMdata for all 19 lime kilns subject to theNSPS were requested along withassociated operational data and designparameters for the lime kilns and limemud washing systems. The Agency hasreceived additional information for 14 ofthe 19 lime kilns subject to the NSPS. Ofthe 14 submitting data, 3 were judged tobe using BDT and had CEM data whichwere accompanied by informationneeded to ascertain the accuracy of thecertification reports. The data fromthese 3 facilities indicate that the NSPScan be achieved when BDT isimplemented., During the data period, one of thethree mills had only one excessemission and the excursion occurredwhen the addition of caustic wasdiscontinued for testing of the CEM. Asecond mill, which previously achievedthe NSPS TRS limit a high percentage ofthe time through good mud washing andprocess contol, began using caustic inrecent months. The most recent excessemission reports show no excessemissions. The Agency considers thisinformation to be indicative that causticaddition reduces excess TRS emissions.

Approximately half of the remainingdata could not be used in making adecision because either data needed to -determine if the CEM's had beenproperly certified was missing or theinformation provided showed that theCEM's had histories indicative ofmaintenance.problems. The data fromthe rest of the mills were suggestive offailure to follow all of the practiceswhich constitute BDT.

In general, the mills that haveemployed CEM'S on lime kilns for anextended period have been the mostsuccessful in continfially achieving theNSPS. The'EPA believes this shows thatthe ability to reliably operate CEM's anduse the GEM's for process control playsa central role in identifying andpreventing those process variations andupsets that cause excess emissions andthat such ability is learned over time.The learning time is necessary to allowthe owner/operator to identify theprocess variables that are leading to theperiods of excess emissions. Theseprocess variables and their impacts -onperiods of excess emissions will bespecific to each mill. The industrycontinues to believe it is possible insome cases that, even with experienceand the use of BDT, there could continueto be periods of excess emissions.

Although such a possibility may not beruled out, the Agency has not receivedany data which would indicate:that suchis the case. The Agency expects that, asthe operators of these facilities learn touse their CEM's to aid in controllingtheir processes, the periods of excessemissions caused by process upsetsshould be significantly reduced whenBDT is fully implemented.

Industry representatives haveexpressed concern that reported excessemissions may be construed asviolations .of the Clean Air Act.Compliance or noncompliance with theAct is determined by performancetesting. A detailed description of theAgency's intended use of CEM data waspreviously published in the FederalRegister (43 FR 7568). The overall intentof the requirement to continuouslymonitor TRS emissions is to provideenforcement agencies with aninstrument to determine that BDT hasbeen implemented and is beingpracticed.

Two comments were receivedconcerning the lime kiln controlled withan ESP which was described in theProposed Rules. The commentersemphasized the uniqueness of thisparticular facility, at which an ESP wasinstalled to meet local and Stateparticulate limits that are site specific,and that an exemption should begranted for this facility. One commenterrequested that the NSPS TRS limit berevised to require this particular facilityto meet a TRS emission limit of 20 ppmvcorrected to 10 percent oxygen, on a 12-hour basis, and not to be exceeded morethan 2 percent of the time on a quarterlybasis. The commenter also said that thestack gases from the ESP would dispersebetter than those from a venturiscrubber because the gases from theESP are approximately 180* hotter.

The Agency has reviewed informationon the lime kiln which is controlled withan ESP instead of a wet scrubber.Information reviewed by the Agencysuggests that this particular facility cancontrol TRS emissions to NSPS levels bymaking additional improvements inprocess controls and by raising thetemperature of the cold end of the limekiln by .100* F. During the review, thecosts of implementing BDT wereexamined. These costs included thecosts to increase cold end temperatures,and the Agency continues to believethese costs are reasonable.

V. Administrative

The docket is an organized andcomplete file of all the informationconsidered by EPA in the developmentof this rulemaking. The docket is a

18543

Federal Register / Vol. 51, No. 97 / Tuesday, May 20, 1986 / Rules and Regulations

dynamic file, since material is addedthroughout the rulemaking development.The docketing system is intended toallow members of the public andindustries involved to readily identifyand locate documents so that they canintelligently and effectively participatein the rulemaking process. Along withthe statement of basis and purpose ofthe proposed and promulgatedstandards and EPA responses tosignificant comments, the contents ofthe docket will serve as the record incase of judicial review [section307(d)(7)(A)J.

The effective date of this regulation isMay 20, 1986. Section 111 of the CleanAir Act provides that standards ofperformance or revisions thereofbecome effective upon promulgation.

As prescribed by Section 111, thepromulgation of these standards waspreceded by the Administrator'sdetermination (41 FR 42028, datedSeptember 24, 1976) that kraft pulp millscontribute significantly to air pollutionwhich may reasonably be anticipated toendanger public health or welfare. Inaccordance with Section 117 of the Act,publication of these promulgatedstandards was preceded by consultationwith appropriate advisory committees,independent experts, and Federaldepartments and agencies.

This regulation will be reviewed 4years from the date of promulgation asrequired by the Clean Air Act. Thisreview will include an assessment ofsuch factors as the need for integrationwith other programs, the existence ofalternative methods, enforceability,improvements in emission controltechnology, and reporting requirements.The reporting requirements in thisregulation will be reviewed as requiredunder EPA's sunset policy for reportingrequirements in regulations.

Section 317 of the Clean Air Actrequires the Administrator to prepare aneconomic impact assessment for anynew source standard of performancepromulgated under section 111(b) of theAct. An economic impact assessmentwas prepared for this regulation and forother regulatory alternatives. Allaspects of the assessment wereconsidered in the formulation of thestandards to insure that cost wascarefully considered in determiningBDT.

This review was submitted to theOffice of Management and Budget(OMB) for review as required byExecutive Order 12291 (OMB ControlNo. 2060-0021). Any comments fromOMB to EPA and any EPA response tothose comments are available forinspection at EPA's Central DocketSection, West Tower Lobby, Gallery 1,

Waterside Mall, 401 M Street, SW.,Washington, DC 20460.

Under Executive Order 12291, EPA isrequired to judge whether a regulation isa "major rule" and therefore subject tothe requirements of a regulatory impactanalysis (RIA). The Agency hasdetermined that this regulation wouldresult in none of the adverse economiceffects set forth in Section 1 of the Orderas grounds for finding a regulation to bea "major rule." The Agency has,therefore, concluded that this regulationis not a "major rule" under ExecutiveOrder 12291.

The Regulatory Flexibility Act of 1980requires the identification of potentiallyadverse impacts of Federal regulationsupon small business entities. The Actspecifically requires the completion of aRegulatory Flexibility Analysis in thoseinstances where small business impactsare possible. Since it is possible thatsome kraft pulp mills qualify as smallbusinesses, the impacts of the standardson small businesses were considered.None of the four criteria which wouldsignify significant impact were met.Because these standards impose noadverse economic impacts, a RegulatoryFlexibility Analysis has not beenconducted.

Pursuant to the provisions of 5 U.S.C.605(b), I hereby certify that the proposedrule will not have a significant economicimpact on a substantial number of smallentities.

List of Subjects in 40 CFR Part 60

Air pollution control, Reporting andrecordkeeping requirements,Incorporation by reference,Intergovernmental relations, and Paperand paper products industry.

Dated: May 9, 1986.Lee M. Thomas,Administrator.

PART 60-[AMENDED]

40 CFR Part 60 is afihended as follows:1. The authority citation for Part 60

continues to read as follows:Authority: Secs. 101, 111, 114, 301(a), Clean

Air Act as amended (42 U.S.C. 7401, 7411,7414, 7416, 7601).

2. In § 60.280, paragraphs (a) and (b)are revised to read as follows:

§ 60.280 Applicability and designation ofaffected facility.

(a) The provisions of this subpart areapplicable to the following affectedfacilities in kraft pulp mills: Digestersystem, brown stock washer system,multiple-effect evaporator system,recovery furnace, smelt dissolving tank,lime kiln, and condensate strippersystem. In pulp mills where kraft pulping

is combined with neutral sulfitesemichemical pulping, the provisions ofthis subpart are applicable when anyportion of the material charged to anaffected facility is produced by the kraftpulping operation.

(b) Except as noted in§ 60.283(a)(1)(iv), any facility underparagraph (a) of this section thatcommences construction or modificationafter September 24, 1976, is subject tothe requirements of this subpart.

3. In § 60.281, paragraph (e) is revisedto read as follows:

§ 60.281 Definitions.

(e) "Brown stock washer system"means brown stock washers andassociated knotters, vacuum pumps, andfiltrate tanks used to wash the pulpfollowing the digestion system. Diffusionwashers are excluded from thisdefinition,

4. In § 60.283, the introduct6ry text ofparagraph (a)(1) is revised andparagraphs (a)(1)(iv), (a)(1)(v), and (a)(4)are revised to read as indicated below:

§ 60.283 Standard for total reduced sulfur(TRS).

(a)* * a(1) From any digester system, brown

stock washer system, multiple-effectevaporator system, or condensatestripper system any gases which containTRS in excess of 5 ppm by volume on adry basis, corrected to 10 percentoxygen, unless the following conditionsare met:

(iv) It has been demonstrated to theAdministrator's satisfaction by theowner or operator that incinerating theexhaust gases from a new, modified, orreconstructed brown stock washersystem is technologically oreconomically unfeasible. Any exemptsystem will become subject to theprovisions of this subpart if the facilityis changed so that the gases can beincinerated.

(v) The gases from the digestersystem, brown stock washer system, orcondensate stripper system arecontrolled by a means other thancombustion. In this case, this systemshall not discharge any gases to theatmosphere which contain TRS inexcess of 5 ppm by volume on a drybasis, corrected tothe actual oxygencontent of the untreated gas stream.

(4) From any smelt dissolving tankany gases which contain TRS in excess

18544

Federal Register / Vol. 51, No. 97 / Tuesday, May 20, 1986 / Rules and Regulations

of 0.016 g/kg black liquor solids as H2S(0.033 lb/ton black liquor solids as H2 S).* * * * *

5. In § 60.284, both the introductorytext of paragraph (a)(2) and (b)(1) arerevised to read as indicated below andparagraph (c)(4) is added. Additionally,the introductory text of paragraph (d) isrevised, paragraph (d)(3) is revised, andparagraph (d)(3)(ii) is revised and addOMB number at the end of the section toread as follows:

§ 60.284 Monitoring of emissions andoperations.

(a) * * *

(2) Continuous monitoring systems tomonitor and record the concentration ofTRS emissions on a dry basis and thepercent of oxygen by volume on a drybasis in the gases discharged into theatmosphere from any lime kiln, recoveryfurnace, digester system, brown stockwasher system, multiple-effectevapofator system, or condensatestripper system, except where the

provisions of § 60.283(a)(1) (iii) or (iv)apply. These systems shall be locateddownstream of the control device(s) andthe spans of these continuousmonitoring system(s) shall be set:* * * a *

(b) * * *

(1) For any incinerator, a monitoringdevice which measures and records thecombustion temperature at the point ofincineration of effluent gases, which areemitted from any digester system,brown stock washer system, multiple-effect evaporator system, black liquoroxidation system, or condensate strippersystem where the provisions of§ 60.283(a)(1)(iii) apply. The monitoringdevice is to be certified by themanufacturer to be accurate within J±_1percent of the temperature beingmeasured.

(c) * * *

(4) Record once per shiftmeasurements obtained from the

continuous monitoring devices installedunder paragraph (b)(2) of this section.

(d) For the purpose of reports requiredunder § 60.7(c), any owner or operatorsubject to the provisions of this subpart-shall report semiannually periods ofexcess emissions as follows:* * * * *

(3) For emissions from any digestersystem, brown stock washer system,multiple-effect evaporator system, or'condensate stripper system periods ofexcess emissions are:

(i) * * *(ii) All periods in excess of 5 mnutes

and their duration during which thecombustion temperature at the point ofincineration is less than 1200 *F, wherethe provisions of § 60.283(a)(1)(iii) apply.

(Reporting and recordkeeping requirementsare approved by the Office of Managementand Budget under Control No. 2060-0021)

[FR Doc. 86-11293 Filed 5-19-86; 8:45 am]BILLING CODE 6560-5O-M

18545