Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed RulesENVIRONMENAL.PROTECiON [.....

ENVIRONMENTAL PROTECTIONAGENCY

40 CFR Part 50

[AD-FRL 2491-51

Proposed Revisions to the National.Ambient Air Quality Standards forParticulate Matter

AGENCY: Environmental ProtectionAgency.ACTION: Proposed rule.SUMMARY: In accordance with sections108 and 109 of the Clean Air Act, EPAhas reviewed and revised the criteriaupon which the existing primary andsecondary particulate matter standardsare based. The revised criteriadocument is being publishedsimultaneously with this notice. Theexisting primary standards forparticulate matter (measured as "totalsuspended particulate matter" or "TSP")are 260 Lg/m 3, averaged over a period of24 hours and not to be exceeded morethan once per year, and 75 /g/m3 annualgeometric mean. The secondarystandard (also measured as TSP) is 150jug/m3, averaged over a period of 24hours, and not to be exceeded more thanonce per year.

As a result of its review and revisionof the health and welfare criteria, EPAproposes the following revisions to theparticulate matter standards:

(1J That TSP as an indicator forparticulate matter be replaced for bothof the primary standards by a newindicator that includes only thoseparticles with an aerodynamic diametersmaller than or equal to a nominal 10micrometers (PM1o);

(2) That the level of the 24-hourprimary standard be changed to a valueto be selected from a range of 150 to 250jig/m 3 and that the current deterministicform of the standard be replaced with astatistical form that permits oneexpected exceedance of the standardlevel per year;

(3) That the level and form of theannual primary standard be changed toa value to be selected from a range of 50to 65 Ag/m s, expressed as an expectedannual arithmetic mean; and

(4) That the current 24-hour secondaryTSP standard be replaced by an annualTSP standard selected from a range of70 to 90 jg/ms, expected annualarithmetic mean.

Because no scientific consensus existson specific levels of the standards, andthe analytical and policy bases formaking these decisions under the statuteare limited and difficult to implement,the Administrator isn not proposingspecific standard levels within the

above ranges. Rather, he is solicitingadditional comment and informationfrom the public to be considered inpromulgating the final regulation, whichwill specify a specific level for each ofthe standards. Given the precautionarynature of the Act, the Administrator isinclined to select the levels of primarystandards from the lower portion of theabove ranges.- A new Federal Reference Method(Appendix J) is proposed to provide formeasurement of PMo in the ambient air.EPA also proposes to add a newAppendix K, which would provideguidance on the statistical nature of theproposed revisions to the standards. Inaddition, certain clarifying changes toAppendix B and Appendix G areproposed. This notice also proposesEPA's int~ntion not to change how"particulate matter" is defined currently-for purposes of the prevention ofsignificant deterioration increments at40 CFR 51.24(c) and 52.21(c).

Related notices published elsewherein today's Federal Register set outproposed revisions to EPA's regulationsconcerning Ambient Air Monitoring.Reference and EquivalentlMethods (40CFR Part 53), and Ambient Air QualitySurveillance (40 CFR Part 58). Proposedrevisions to EPA's regulationsconcerning Requirements forPreparation, Adoption, and Submittal ofImplementation Plans (40 CFR Part 51)with associated guidelines, andApproval and Promulgation ofImplementation Plans (40 CFR Part 52)will be published later. Following thepublication of these notices, the Agencywill announce a supplementary reviewperiod for the limited purpose of takingcomments on the implications, if any, ofthe proposed Parts 51 and 52implementation requirements andguidelines for the Part 50 standardsproposed today.DATES: EPA will hold a public hearingon this notice and the related noticeswithin 45 days. The time and place willbe announced in a subsequent FederalRegister notice. Written comments onthis proposal, including anysupplementary and rebuttal informationsubmitted pursuant to section 307(d)(5)of the Clean Air Act, must be receivedby June 18, 1984.ADDRESSES: Submit all comments(duplicate copies are preferred) exceptthose relating to Prevention ofSignificant Deterioration increments(Parts 51 and 52) to: Central DocketSection (A-130), EnvironmentalProtection Agency, Attn: Docket No. A-82-37, 401 M Street SW., Washington,D.C. 20460. Comments on Prevention ofSignificant Deterioration increments

should be sent to the same address,Attn: Docket No. A-83-48. Dockets No.A-82-37 and No. A-83-48 are located inthe Central Docket Section of the U.S.Environmental Protection Agency, WestTower Lobby, Gallery 1, 401 M StreetSW., Washington, D.C. The docket maybe inspected between 8:00 a.m. and 4:00p.m. on weekdays, and a reasonable feemay be charged for copying.

Availability of Related InformationThe revised criteria document, AirQuality Criteria for Particulate Matterand Sulfur Oxides (three volumes, EPA-600/8-82-029a-c, December, 1982;Volume I NTIS #PB-84-156785, $19.00,Volume II NTIS #PB-84-156793, $43.00Volume III NTIS #PB-84-156801, $47.00;complete set #PB-84-156777, $9,00;microfiche $4.50 for each volume) andthe final'revised staff paper, Review ofthe National Ambient Air QualityStandards for Particulate Matter:Assessment of Scientific and TechnicalInformation-OAQPS Staff Paper (EPA-450/5-82-001, January, 1982; NTIS #PB-177874, $21.00 paper copy and $4.00microfiche) are available from: U.S.Department of Commerce, NationalTechnical Information Service, 5265 PortRoyal Road, Springfield, Virginia 22161.A limited number of copies of otherdocuments generated in connection withthis standard review, such as the controltechniques document, can be obtainedfrom: U.S. Environmental ProtectionAgency Library (MD-35), ResearchTriangle Park, N.C. 27711, telephone(919) 541-2777 (FTS 629-2777).FOR FURTHER INFORMATION CONTACT.Mr. John Haines, Strategies and AirStandards Division (MD-12), U.S.Environmental Protection Agency,Research Triangle Park, N.C. 27711,telephone (919) 541-5531 (FTS 629-5531),SUPPLEMENTARY INFORMATION:

INTRODUCTION

The Clean Air Act requires EPA to set,and periodically reexamine, "nationalambient air quality standards" forwidespread pollutants. These standardsconsist of "primary" standards designedto protect public health and "secondary"standards designed to protect publicwelfare. The statute requires primarystandards to be set low enough toprotect public health with an "adequatemargin of safety." Once these standardshave been set, states must submit"implementation plans" that containcontrol measures needed to attain thestandards within specific statutorydeadlines. These legislativerequirements are discussed more fully Insubsequent portions of this preamble.

10408

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

In 1971, EPA established primary andsecondary ambient air quality standardsfor particulate matter. It has nowreviewed those standards and thespecific criteria on which they arebased, as required by the Clean Air Act,and is today proposing appropriatechanges.

This proposal has been preceded byan exhaustive review of all availablescientific information on the health andwalfare effects of airborne particulatematter. As -detailed below in the"Background" section, this review hastaken over three years. It has includedthree public meetings of EPA's Clean AirScientific Advisory Committee(CASAC,* five other formallyannounced public meetings, numerousinformal meetings, and the review ofwritten comments received throughoutthis process. Besides the CASACscientists, the review has involved alarge number of EPA staff, consultants,and external reviewers. In a number ofareas, this wide-ranging discussion hasled to significant agreement on a courseof action. In one crucial area, however-the numerical stringency of the 24-hourand annual standards-the scientificand technical review has only producedrelatively broad ranges of numbers fromwhich the standard levels should bechosen.

The Administrator believes that, giventhe present design of the statute, theselection of a single air quality standardfrom each of the ranges of standardsthat have been recommended to himpresents an extraordinarily difficultregulatory problem, one for which theexisting legislative decision criteria maywell be inadequate.

The review and assessment ofscientific information by Agency andoutside experts was intensive; it wasnot, however, intended to result inrecommendations of any single level(s)of airborne particulate matter thought tobe stringent enough to-meet thestatutory test of protecting public healthwith an "adequate margin of safety.."Indeed, that review has revealed ahighly limited data base-particularlywhere quantitative studies areconcerned-and a wide range of viewsamong qualifiedprofessionals about theexact pollution levels at which health'effects are likely to occur. The setting ofan "adequate margin of safety" belowthese levels calls for a furtherjudgment-in an area for which thescientific data base is even more sparse

*CASAC is a standing committee of scientificandengineers external to the Federal governmentestablished under section 109 of the Clean Air Actto advise the Administrator on the scientific basisfor ambient air quality standards.

and uncertain. No "scientific" approachfor selecting afiy single recommendedstandard seems possible against thisbackground. Instead, the EPA staff hasidentified-with CASAC concurrence--"ranges of interest" to aid theAdministrator in choosing levels forboth the 24-hour and annual standards.In each case, the staff and CASAC haveconcluded that a standard at the upperend of the range would provide little orno margin of safety, and that standardsat lower levels within the range wouldprovide correspondingly greater marginsof safety. This preamble follows staffrecommendations and focuses on thelower levels being considered. Underthe statute, the task remaining is todecide at which level within each ofthese refined ranges the margin of safetyshould be considered "adequate." In theend it will be up to the Administrator tomake this judgment by picking a singlenumber from each range.

The final judgment on standardstriggers the process by which thestandards are met and maintained-aprocess that requires potentially majorexpenditures for compliance within alimited time period. Followingpromulgation of the standards, thestates have nine months to design andsubmit "implementation plans" toachieve the health based standardswithin three years. If, for some areas,attainment in this time period iseconomically, socially, ortechnologically infeasible, theAdministrator's options for adjusting thedealines are quite limited.

Despite'the significant consequencesthat may flow from the establishment ofstandards, the statute as presentlyinterpreted severly restricts the factorsand analytical tools the Administratormay use to help pick these standards.The courts appear to have ruled thateconomic and technological feasibilityhave no function in deciding on anambient standard. Given the scope ofthoes statements, there is some doubtwhether the Administrator may evenconsider the practical problems ofimplementation to guide his choice.Public health appears to be the solecriterion.

Yet long and expert review of publichealth issues has to date revealed noscientific method of assessing exactlywhat level of standards public healthrequires. The scientific review indicatessubstantial uncertainties concerning thehealth risks associated with lower levelsof particulate matter. Assessing theserisks is made even more complex by thefact that the composition of particulatematter and associated air pollutants canvary significantly from city to city.

Given the difficulty of this choice,EPA invites public comment on thegeneral policy questions that it raises. If,indeed, only public health factors maybe considered, what particularanalytical approaches or methodologiesmay the Administrator apply to makehis ultimate choice in a principled way?Is there room-at least scientific opinionis as lacking in definitive answers as itis here-to consider other, non-scientificfactors in making the major social policyjudgment of picking a precise numberfrom a range of scientifically justifiedvalues? If so, what other factors shouldbe considered, and in what manner?

BACKGROUND

Legislative Requirements Affecting ThisProposal

The Standards

Two sections of the Clean Air Actgovern the establishment and revision ofnational ambient air quality standards(NAAQS). Section 103 (42 U.S.C. 7408)directs the Administrator to identifypollutants which may reasonably be

'anticipated to endanger public health orwelfare and to issue air quality criteriafor them. These air quality criteria are toreflect the latest scientific informationuseful in indicating the kind and extentsof all identifible effects on public healthor welfare that may be expected fromthe presence of a pollutant in theambient air.

Section 109(a) (42 U.S.C. 7409) directsthe Administrator to propose andpromulgate "primary" and "secondary"NAAQS for pollutants identified undersection 108. Section 109(b](1] defines aprimary standard as one the attainmentand maintenance of which in thejudgment of the Administrator, based onthe criteria and allowing for anadequate margin of safety, is required toprotect the public health. The secondarystandard, as defined in section 109(b)(2),must specify a level of air quality theattainment and maintenance of which inthe judgment of the Administrator,based on the criteria, is requisite toprotect the public welfare from anyknown or anticipated adverse effectsassociated with the presence of thepollutant in the ambient air. Welfareeffects are defined in section 302(h) (42U.S.C. 7602(h)) to include effects onsoils, water, crops, vegetation, man-made materials, animals, wildlife,weather, visibility, climate, damage toand deterioration of property, hazards totransportation. and effects on economicvalues and on personal comfort andwell-being.

The courts have held that therequirement for an adequate margin of

104t09

Federal Register / Vol. 49, NO. 55 / Tiieidav. March 20.' 1984 / Prnon;nd Rules

safety for primary standards wasintended to address uncertaintiesassociated with inconclusive scientificand technical information available atthe time of standard setting. It was alsointended to provide a reasonable degreeof protection against hazards thatresearch has not yet identified. LeadIndustries Association v. EPA, 647 F.2d1130, 1154 (D.C. Cir. 1980), cert. denied,101 S. Ct. 621 (1980); AmericanPetroleum Institute v. Castle, 665 F.2d1176, 1177 (D.C. Cir. 1981), cert. denied,102 S. Ct. 1737 (1982). Theseuncertainties in the availableinformation and about unidentifiedhuman health effects are bothcomponents of the risk associated withpollution at levels below those at whichhuman health effects can be said tooccur with reasonable scientificcertainty. Thus, in providing anadequate margin of safety, theAdministrator is regulating not only toprevent pollution levels that have beendemonstrated to be harmful, but also toprevent lower pollutant levels that hefinds pose an unacceptable risk of harm,even if that risk is not preciselyidentified as to nature or degree. Inweighing such risks for selecting amargin of safety, EPA has consideredsuch factors as the nature and severityof the health effects involved, the size ofthe sensitive population(s) at risk, andthe kind and degree of the uncertaintiesthat must be addressed. Given that the"margin of safety" requirement bydefinition only comes into play whereno conclusive showing of harm exists,such factors, which involve unknown oronly partially quantified risks, havetheir inherenf limits as guides to action.The selection of any particular approachto providing an adequate margin ofsafety is a policy choice left specificallyto the Administrator's judgment. LeadIndustries Association v. EPA, supra,647 F.2d at 1161-62.

The courts, however, have set strictlimits to the factors EPA may considerin establishing a margin of safety. Tworecent judicial decisions state that theeconomic and technological feasibilityof attaining ambient standards are notto be considered in setting them, even inthe context of a margin of safety. LeadIndustries Association v. EPA, suprq,647 F.2d at 1148-51; American PetroleumInstitute v. Castle, supra, 665 F.2d at1185, 1190. Such factors may, however,be considered to a degree in thedevelopment of State plans toimplement the standards.

Section 109(d) of the Act (42 U.S.C.7409(d)] requires periodic review and, ifappropriate, revision of existing criteriaand standards. The process by which

EPA has reviewed the original criteriaand standards for particdlate matterunder section 109(d) is described in alater section of this notice.Related Control Requirements

States are primarily responsible forassuring attainment and maintenance ofambient air quality standards, once EPAhas established them. Under section 110of the Act (42 U.S.C. 7410), States are tosubmit, for EPA approval, StateImplementation Plans (SIPs) thatprovide for the attainment andmaintenance of such standards thr6ughcontrol programs directed to sources ofthe pollutants involved. Other federalprograms provide for nationwidereductions in emissions of these andother air pollutants through the FederalMotor Vehicle Control Program, whichinvolves controls for automobile, truck,bus, motorcycle, and aircraft emissionsunder Title II of the Act (42 U.S.C. 7501to 7534), and through the development ofNew Source Performance Standards andNational Emission Standards forHazardous Air Pollutants for variouscategories of stationary sources undersection 111 (42 U.S.C. 7411) and section112 (42 U.S.C. 7412).

.Particulate Matter and ExistingStandards for TSP

"Particulate matter" is the genericterm for a broad class of chemically andphysically diverse substances that existas discrete particles (liquid droplets or.solids) over a wide range of sizes.Particles originate from a variety ofstationary and mobile sources. Theymay be emitted directly or formed in theatmosphere by transformations ofgaseous emissions such as sulfur oxides,nitrogen oxides, and volatile organicsubstances. The major chemical andphysical properties of particulate mattervary greatly with time, region,meteorology and source category, thuscomplicating the assessment of healthand welfare effects as related to variousindicators of particulate pollution. Thecharacteristics, origins, concentrationsand potential effects of particulatematter are discussed in more detail inthe staff paper (SP; EPA, 1982a) and inthe revised criteria document (CD; EPA,1982b). The executive summary of thestaff paper is xeprinted in Addendum Ito this notice.

On April 30, 1971, EPA promulgatedprimary and secondary NAAQS forparticulate matter under section 109 ofthe Clean Air Act (36 FR 8186). Thereference method for measuringattainment of-these standards is the"high-volume" sampler (40 CFR Part 50,Appendix B), which effectively collectsparticulate matter up to a nominal size

of 25 to 45 micrometers (Mim) (so-called"total suspended particulate," or "TSP"),Thus, TSP is the current indicator for theparticulate matter standards. Theexisting primary standards forparticulate matter (measured as TSP)are 260 /g/m 3, averaged over a period of24 hours and not to be exceeded morethan once per year, and 75 Lg/m3 annualgeometric mean. The secondarystandard (measured as TSP) is 150 pg/m3

, averaged over a period of 24 hours,and not to be exceeded more than onceper year. The scientific and technicalbases for these standards are containedin the original criteria document, AirQuality Criteria for Particulate Matter(DHEW, 1969).Development of Revised Air QualityCriteria for Particulate Matter

In 1976, as a result of internal agencyreview and the recommendations of~acommittee of EPA's Science AdvisoryBoard, EPA decided to revise theexisting criteria document for.particulate matter. Because of competingpriorities regarding revision of othercriteria documents and the need tocomplete additional research onparticulate matter, the process wasscheduled to commence in 1979. Withthe endorsement of the Clean AirScientific Advisory Committee (CASAC)of EPA's Science Advisory Board, EPAdecided to review and revise the criteriadocument for particulate matterconcurrently with that for sulfur oxidesand to proddlce a combined particulatematter/sulfur oxides (PM/SO) criteriadocument.

On October 2,1979 (44 FR 56731), EPAannounced that it was In the process ofrevising the original criteria documentfor particulate matter and reviewing theexisting air quality standards forpossible revisions in accordance withsection 109(d)(1) of the Clean Air Act.

In developing the revised criteriadocument, EPA has provided a numberof opportunities for review and commentby organizations and individuals outsidethe Agency. Three drafts of the revisedparticulate matter/sulfur oxides criteriadocument,-prepared by EPA'sEnvironmental Criteria and AssessmentOffice (ECAO), have been madeavailable for external review (45 FR24913; 46 FR 9747; 46 FR 53210). EPA hasreceived and considered numerous andoften extensive comments on each ofthese drafts. CASAC has held threepublic meetings (August 20-22, 1980; July7-9, 1981; November 16-18, 1981) toreview successive drafts of thedocument. These meetings were open tothe public and were attended by manyindividuals and representatives of

F I

10410

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Pro'po'sed Rules

organizations who provided criticalreviews and new information forconsideration. Based on CASACrecommendations made after the firstreview meeting, five additional publicmeetings were held at which EPA, itsconsulting authors and reviewers, andother scientifically and technicallyqualified experts selected by EPAdiscussed the various chapters of thedraft document and suggested ways ofresolving outstanding issues (45 FR74047; 45 FR 78224; 45 FR 76790; 45 FR80350; 46 FR 1775).

The comments received on thesuccessive drafts of the revised criteriadocument have been considered in thefinal document, issued simultaneouslywith this proposal. A summary of EPA'sresponses to the comments on the threeexternal review drafts of the documentshas been placed in the public docket(Docket No. A-82-37). Transcripts of thethree CASAC meetings are also in thedocket. In accordance with itsestablished procedures, CASACprepared a "closure" memorandum tothe Administrator indicating itssatisfaction with the final draft(December, 1981] of the criteriadocument and outlining key issues andrecommendations. The closurememorandum, dated January 29, 1982,stated that the EPA office that preparedthis document was "responsive toCommittee advice as well as tocomments provided by the generalpublic * * . The closure memorandumfurther states that the criteria document"fulfills the requirements set forth insection 108 of the Clean Air Act, whichrequires that the criteria document 'shallaccurately reflect the latest scientificknowledge useful in indicating thelkindand extent of all identifiable effects onpublic health or welfare' from sulfuroxides and particulates in the ambientair." Following closure, minor technicaland editorial refinements were made tothe criteria document for printing (EPA,1982b). The CASAC closurememorandum on the criteria documentis reprinted in its entirety as AddendumH to this notice.

A number of scientific and technicalissues were raised during the publicreview process. With respect to theparticulate matter portions of thecriteria document, the major issuesincluded the relationship among variousmeasures of particulate matter airquality, the implications of particledeposition and other studies forselecting a particulate matter indicator,and the development and application ofcriteria for deciding whichepidemiological studies are mostappropriate for use in revising air

quality standards. A summary of theseand other major scientific issues, as wellas CASAC's conclusions, is included inthe closure memorandum on the criteriadocument (Addendum II).Revew of the Standards: Developmentof Staff Paper

In the Spring'of 1981, EPA's Office ofAir Quality Planning and Standards(OAQPS) prepared the first draft of astaff paper, Review of the NationalAmbient Air Quality Standards forParticulate Matter (see Addendum I).This draft staff paper evaluated andinterpreted the available scientific andtechnical information most revelant tothe review of the air quality standardsfor particulate matter and presentedstaff recommendations on alternativeapproaches to revising the standards,based on the then-existent draft of therevised criteria document. This and asecond draft of the paper were reviewedat two CASAC meetings (July 7-9,1981;November 16, 1981). Numerous writtenand oral comments were received on thedrafts from CASAC, representatives oforganizations, individual scientists andother interested members of the public.A summary of major revisions inresponse to comments on the first draftis contained in an October 31,1981 letterto CASAC (Padgett, 1981). Following thesecond CASAC meeting, the staff madefurther revisions in response tocomments and prepared an executivesummary that was reviewed by CASACmembers before preparation of theclosure memorandum on the staff paper.In January, 1982, EPA released the finalOAQPS staff paper (EPA, 1982a), whichreflects the various suggestions made byCASAC and members of the public. TheJanuary 29, 1982, CASAC closurememorandum states that the staff paper"has been modified in accordance withrecommendations made by CASAC." isconsistent with the criteria document.and provides the Admirdstrator "withthe kind and amount of technicalguidance that will be needed to makeappropriate revisions to the standard."

A number of major issues were raizedduring the public review process. Themore important issues are outlinedbelow.

1. Substintial discussion concernedthe maximum size of particles (orparticle size fraction) to be used inmeasuring particulate matter forregulatory purposes. Some groupsfavored retaining TSP and others calledfor alternative size-specific standardswith nominal "size cuts" ("D-o", seelater discussion) of 15 jim, 10jn, 5-7im, and 2.5 im. After CASAC closure

on the staff paper and criteria document,comments were received from one group

favoring a so-called "Do" of 10 ym(approximately equivalent to a nominalsize cut [D-.o] of 6 gm).

2. Much attention was focused on thedevelopment of numerical "ranges ofinterest" for selecting the level ofalternative particulate matter standardsand on which studies were mostappropriate for use in standard setting.Significant criticisms were received onthe major epidemiological studies ofparticulate matter exposureshighlighting their limitations for use instandard setting. In a number ofcomments, specific suggestions forstandards were made.

3. With respect to secondarystandards, most attention focused on thebasis for a fine (<2.5 gm) particlestandard related to visibility protection.

These and other major issues arediscussed more fully in the executivesummary of the staff paper (AddendumI) and in later sections of this notice.CASAC's discussion of these issues andits recommendations are contained inthe closure memorandum on the staffpaper (Addendum III).

Rulemaking Docket

EPA established a standard reviewdocket for the particulate matterstandard revision in July, 1979. With thisproposal. EPA is establishing arulemaking docket (Docket No. A--82-37)as required by section 307(d) of theClean Air Act. The most relevantportions of the standard review docket(Docket No. A-79-29) and of a separatedocket established for criteria documentrevision (Docket No. ECAO-CD-79-1]have been incorporated in thisrulemaking docket. The balance of thestandard review and criteria revisiondockets vill continue to be available forpublic reference.

RATIONALE FOR THE PRIMARYSTANDARDS

In selecting primary standards forparticulate matter, the Administratormust specify: (1) The particle sizefraction that should be used as anindicator of particulate pollution; (2) theappropriate averaging times and formsof the standards; and (3] the numericallevel(s) for the standards. Based on theassessment of relevant scientific andtechnical information in the criteriadocument, the staff paper (hereafter"SP") outlines a number of key factorsto be considered when making decisionsin each of these areas (SP, Section VIU.Evaluation of the margin of safetyafforded by a given particulate matterstandard should include considerationof these specifications collectively,rather than focusing on any one. Both

10411

Federal Register / Vol. 49, No. 55 / Tiisda , M'arch' 20, [94 / Probosed "Ruiles

the staff and CASAC maderecommendations to focus considerationon a discrete range of policy options ineach of these areas. In most respects,the Administrator has adopted therecommendations and supportingreasons contained in the staff paper andthe CASAC closure memorandum.Rather than reiterating thosediscussions at length, the followingdiscussion of the proposed standardsfocuses primarily on thoseconsiderations that were mostinfluential in the Administrator'sselection of a particular option, or thatdiffer in some respect fromconsiderations that influenced the staffand/or CASAC recommendations.

Since CASAC closure on the criteriadocument and staff paper in January,1982, a number of studies on the healtheffects of particulate matter haveappeared in the scientific literature.Examples that have been placed in therulemaking docket include Proctor andSwift, 1982; Ostro, 1983; Mazumdar andSussman, 1983; Mazumdar et al., 1982;Vena, 1983; Perry et al., 1983; Baxter etal., 1983; Avol et al., 1983; and Dockeryet al., 1982. Although none of thesestudies has been used as a basis for thisproposal, some of them could be ofImportance in a final decision. Thepublic is invited to comment on theImplications of these or any other recentstudies for the standards. Afterconclusion of the public commentperiod, but before preparation of thepromulgation notice, the Agency willprepare a document identifying anddiscussing the significance of any suchstudies it considers useful for the finaldecision and submit the document forCASAC and public review.Pollutant Indicator

The Administrator concurs with thestaff conclusions that (1) a.separategeneral particulate matter standard (asopposed to a combination standard forparticulate matter and SO 2) remains areasonable public health policy choice,and (2] given current scientificknowledge and uncertainties, a size-specific (rather than chemical-specific)indicator should be used. The currentindicator (TSP) is size-specific, but hasbeen widely criticized because it directscontrol efforts toward-larger particlesthat can dominate measured mass, butare of less concern to health thansmaller particles. In assessing theinformation in the criteria document, thestaff reached several conclusionssummarized below (see SP, pp. 71-75):

(1) Health risks posed by inhaledparticles are influenced both by thepenetration and deposition of particlesin the various regions of the respiratory

tract, and by the biological iesponses tothese deposited materials.

(2) The risks of adverse health effectsassociatedwith deposition of ambientfine and coarse*particles in the thorax(tracheobronchial and alveolar regionsof the respiratory tract) are markedlygreater than for deposition in theextrathoracic (head) region. Maximumparticle penetration to the thoracicregion occurs during oronasal or mouthbreathing.

(3) The risks of adverse health effectsfrom extrathoracic deposition of generalambient particulate matter aresufficiently low that particles depositingonly in that region can safely beexcluded from the standard indicator.

(4) The size-specific indicator forprimary standards should representthose particles capable of penetrating tothe thoracic region, including both thetracheobronchial and alveolar regions.

Considering these conclusions in lightof data on air quality composition,respiratory tract deposition and healtheffects the need to provide protectionfor sensitive individuals who maybreathe by mouth and/or oronasally,and the similar convention on particlespenetrating the thoracic region recentlyadopted by the International StandardsOrganization (ISO, 1981), the staffrecoinmended that the size-specific -indicator include particles less than orequal to a nominal 10 Am "cut point."*This indicator ensures that the full rangeof particles penetrating to the sensitivealveolar region is included, and followstracheobronchial penetration patterns ina somewhat conservative fashion. Itplaces substantially greater emphasis oncontrolling smaller particles than does aTSP indicator, but does not completelyexclude larger particles from all control.These and other factors considered inrecommending a 10 Am cut-point areoutlined in the staff paper (SP, pp. 75-79).

The Administrator accepts therecommendations of the staff andCASAC and their underlying rationaleand proposes to replace TSP as theparticle indicator for the primarystandards with a new indicator that-includes only those particles less than anominal 10 Am. This indicator is referred

* Th4 more precise term is 509 cut point or 50%diameter (D-,.. This is the aerodynamic particlediameter for which the efficiency of particlecollection is 50%. Larger particles are collected withsubstantially lower efficiency and smaller particleswith greater (up to 1007) efficiency. In practicalusage, acceptable ambient samplers with this cutpoint provide a reliable estimate of the total mass ofsuspended particulate matter of aerodynamic sizeless than or equal to 10 Im. See additionaldiscussion regarding the proposed FederalReference Method below and in the accompanying,notice proposing revision of 40 CFR Part 53.

to as "thoracic particles" (TP) in thestaff paper. For more general use indefining the standards, the regulatedpollutant has been termed PMjo.

In proposing a PMjo indicator, theAdministrator also invites publiccomment on certain informationpublished and submitted after CASACclosure on the criteria document andstaff paper. The American MiningCongress (AMC)*bas sponsored andsubmitted a new analysis (AMC, 1982)of particle deposition in the respiratorytract. The analysis, which was recentlypublished as a preliminarycommunication (Swift and Proctor,1982), suggests that the data used torepresent particle deposition in thecriteria document and staff paperoverstate particle penetration to thethoracic regions of the respiratory tractbecause the experiments used artificialinterventions (mouthpieces and noseclips) that do not simulate naturaloronasal breathing. Swift and Proctoralso attempt to quantify the extent of theoverestimation by developing simulatedparticle deposition curves for oronasalbreathing. Based on theie simulations,AMC has recommended a particle sizeindicator that collects no particlesgreater than 10 /m (Do-10 nm), with acut point (D2o) that EPA interprets asbeing approximately 6 Am.

The Swift and Proctor analysis isrelevant to the final decision on a sizeindicator and should be considered-during the public comment period. TheAdministrator is not, however,proposing the AMC recommendations,in part because the supporting analysiswas only recently published as apreliminary communication and was notconsidered by CASAC, and in partbecause of reservations associated withthe recommendations themselves. Thelikelihood that the data used to derivePMto overstate thoracic deposition wasrecognized in a qualitative sense byCASAC (cf. July, 1981 transcript, p. 581)and presented as one reason forrecommending 10 Am as anappropriately conservative particleindicator (cf. July, 1981 transcript, p.584). Hence, the revised staff paperspecifically reflected this argument withrespect to mounthpiece results in itsrecommendations favoring 10 Am over15 jm as the cutpoint (SP, pp. 76-77).The assumptions used by Swift andProctor (1982), on the other hand, mayresult in underestimating thoracicparticle deposition, at least in somecases; this would reduce any margin ofsafety associated with an indicatorderived from these data. The Swift andProctor analysis itself suggests thatapproximatley 10 to 20% of 10 Am

I . w f A10412

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

particles could penetrate to the thoracicregion, rather than the 0% penetrationimplied by the AMC recommendationfor a2 DD" of 10 gm. If the Swift andProctor analysis were used indetermining the cut-point for the finalstandard, this penetration would have tcbe taken into account.

An additional factor to be consideredin the final decision is that a shift to anindicator other than PMo wouldnecessitate an adjustment in the level ofa given standard to account for thereduced amount of particles collected. Ifthe AMC proposal were selectedfollowing public comment, the levels ofthe standards would probably beadjusted downward by a factorreflecting the best estimate of the ratioof the mass of particles-less than 6 imto-the mass less than 10 ;Lm. Based onavailable interpolations, that ratio isestimated to be approximately 0.8 (Pace,1982).

While the Administrator proposes aPMo indicator at this time, the public isinvited to comment on theappropriateness of using the AMCrecommendations and supportinganalysis to develop a different indicatorfor the primary standards and on theadjustment of the numerical level of thestandard that would be appropriate if adifferent indicator were selected afterconsideration of comments on this issue,

Averaging Tune and Form of the.Standards

The Administrator concurs with staffand CASAC recommendations forretaining both 24-hour and annualprimary standards for particulatematter. A single averaging time wouldnot appear to provide adequateprotection against potential effects fromboth long- and short-term exposureswithout being unduly restrictive. Theform for both 24-hour and annualstandards is discussed below.

1. In accordance with staffrecommendations, the Administratorproposes that the 24-hour standard bestated in a statistical form, rather than:the current deterministic form. Whenused with an appropriate standard levelthe statistical form can, provideimproved health protection that is lesssensitive to changes in samplingfrequency than the deterministic formand also can offer a more stable targetfor control programs. Recognition of thelimitations of the deterministic form hasled EPA to promulgate or proposestatistical forms-for the ozone andcarbon monoxide standards (44 FR 8202.45 FR 55066].

The proposed interpretation of thestatistical form of the particulate matterstandard is detailed in Appendix K of

the proposed regulation. As presentedthere, the standard would be attainedwhen the expected number ofexceedances of the 24-hour standardlevel is no more than one per year.Generally, the determination would be

I based on three consecutive years ofdata. A difficulty in applying the singleexceedance statistical form toparticulate matter arises from thecurrent practice of limiting the samplingfrequency for particulate matter(typically only one 24-hour sample iscollected every six days). This leads toan increased chance of misclassifyingareas as non-attainment. The proposedapproach for addressing this issue ispresented in the accompanying proposalfor 40 CFR Part 58.

An alternative form of the standardconsidered during this review was topermit multiple expected exceedancesof the standard level. Analyses of airquality data indicate that a multipleexceedance form can provide evengreater stability for control programsand reduce the possibility of incorrectlyclassifying areas as attainment ornonattainment because of unusual orinfrequent meteorological conditions. Ifa multiple exceedance form were to beused, the level of the standard would beestablished at a lower numerical valuethan for a single exceedance standard toensure comparable health protection, inaccordance with CASAC'srecommendation. Conceptualapproaches for considering theinteraction between standard level andnumber of exceedances are outlined inthe staff paper (SP, pp. 81-83). The staffpaper analysis does not, however,provide a complete comparison of therelative health protection, stability, andstringency of control afforded by asingle as compared to a multipleexceedance standard. A morecomprehensive analysis of these factorsis underway and will be placed in thedocket when completed. Pendingcompletion of this analysis, the greatercontrol of peak values available throughthe single exceedance form is preferred.

While the Administrator does notpropose a multiple exceedance form forthe 24-hour particulate matter standardat this time, the public is invited tocomment on the advisability of multipleexpected exceedances for determiningattainment of the 24-hour standard andon the adjustment of the nume rical levelof the standard that would beappropriate if a multiple exceedanceform were adopted after considerationof comments on this issue.

2. The Administrator proposes tochange the form of the annual standardfrom the current annual geometric meanto a statistical form expressed as an

expected annual arithmetic mean. Theexpected arithmetic mean form is moredirectly related to the available healtheffects information than is the currentform of the standard. The change to anarithmetic mean was recommended bythe staff and CASAC on this basis. Theproposed interpretation of the statisticalform of the standard is detailed inAppendix K to the proposed regulation.

Under the proposed statistical form.the standard would be expressed as anexpected annual arithmetic averagedetermined by averaging the annualarithmetic averages from threesuccessive years of data. The current -

deterministic form of the standard doesnot adequately take into account therandom nature of meteorologicalvariations. In general, annual meanparticulate matter concentrations willvary from one year to the next, even ifemissions remain constant, due to therandom nature of meteorologicalconditions that affect the formation anddispersion of particles in theatmosphere. This limitation means thatcompliance with the standard and,consequently, emission controlrequirements, may be determined on thebasis of a year with unusually adverseweather conditions. The problem of yearto year variability, is. however, muchless significant for annual averageconcentration standards than for 24-hour standards.

In proposing the statistical form, EPAhas considered the relative protectionprovided by a given standard level withthis form as compared to thedeterministic form, and has assessed therelationship between statistical anddeterministic forms of an annualstandard (Frank, 1982). Based on thisanalysis, the level of an annualstatistical PM2 o standard that would, onaverage, provide protection equivalentto a qtandard with a deterministic form.would be a factor of approximately 0.9times the level of that deterministicstandard.

3. The proposed interpretation of thestatistical forms for both 24-hour andannual standards in Appendix K isconceptually similar to that proposed inAppendix I of the Carbon MonoxideStandard (45 FR 55066]. Howeverspecific adjustments have been made tothe computations necessary foranalyzing particulate matter data toaccount for the different averagingperiods of the proposed standards andfor non-scheduled sampling days asduring episode periods.

The proposed appendix also specifies.criteria for determining attainment ofthe proposed standards when less than3 years of representative data are

10.413

Federal Register / Vol. 49. No. 55 / Tmsdav. March 2 f 1Q5 / Prn noe P,, I,-, -- * ~ .' r. , a..sp..0L.t& A..Aflo

available. The proposed criteria specify:(1) That two years of datarepresentative of "normal" conditionswould be sufficient to perform thecalculation in order to show attainmentof the annual standard; (2) that twoyears of representative data would besufficient to perform the calculations forthe 24-hour'standard if the monitorsamples every day and achieves anannual average data capture of 50percent; and (3) that one year ofrepresentative data will be sufficient forboth annual and 24-hour standards if themonitor samples every day and achieves75 percent capture. In proposing thesecriteria, the Agency sought to minimizethe likelihood of misclassifying areas.due to incomplete data. Although datanot meeting the criteria could also beused, such exceptions would have to beapproved by the RegionalAdministrator. Comments arespecifically requested on these criteria.

Provisions to minimize the influenceof unusual events and trends in thecomputation of exceedances are alsoproposed. These are directdd at: (1) Rareand unusual events that cannot becontrolled through the StateImplementation Plan process; and (2)situations in which trends in emissionsand air quality are evident. The Agencyis currently developing additionalguidance on data requirements andtreatment of both trends and-unusualevents. When completed, theseguidelines will be made available forpublic comment.

The computational formulas for the24-hour standard do not account for anyseasonal differences in data capture andpollutant concentrations. Althoughnormally small, in some cases theycould affect the calculation of expectedexceedances. Therefore, EPA requestscomments on whether computationalformulas that would address this .question should be added to AppendixK.

The proposed attainment test for the24-hour standard would compute anestimate for the expected annualexceedance rate and then compare thisestimate to the allowable exceedancerate of once per year. An alternative

,approach would recognize the statisticalvariability associated with theestimated annual exceedance rate anddevelop a tolerance interval so that asite would not be classified as non-attainment unless the estimated annualexceedance rate exceeds the allowableexceedance rate by more than thistolerance interval. The magnitude of thistolerance interval could be determinedby developing a statistical model toaccount for the underlying variability of

the data and could incorporate factorssuch as autocorrelation. From astatistical viewpoint, this type ofapproach would have the advantage ofcontrolling the "Type I" error, in thiscase the probability of misclassifying asite as non-attainment when it actuallyis attainment. However, by 'incorporating this tolerance intervalfewer sites may, depending on thesignificance level chosen, be classifiedas non-attainment than with theattainment test proposed in this notice.The tolerance interval approach mightalso be applied in the other direction, sothat a site would not be considered tohave demonstrated attainment unlessthe estimated exceedance rate plus thetolerance limit is less than the allowableexceedance rate. Therefore, thisalternative approach may, depending onthe significance level chosen, make itmore difficult for a site to be classifiedas either attainment or non-attainment.The concept of accounting for theunderlying variability of the data is alsoapplicable to the annual standard inwhich a tolerance interval could bedetermined for the annual mean. EPAsolicits comments on such approachesfor this and other ambient air qualitystandards, particularly with respect tothe advisability of their use, thepractical aspects of developing thesetolerance intervals, an appropriatechoice of a significance level used todevelop the tolerance intervals, and howthey might affect ongoing programs.Level of the Standards

The staff paper and CASACrecommendations set forth a frameworkfor considering the levels for theproposed particulate matter standardsto ensure that they protect public healthwith an adequate margin of safety. Thediscussion that folllows on the levels ofthe standard relies heavily on thatframework and on the supportingmaterial in the staff paper and closurememorandum. The essential steps in thisframework are summarized below.

1. Assessment of the moirequantitative epidemiological studies ofparticulate matter. The criteriadocument identifies a small number ofcommunity epidemiological studies thatare useful in developing quantitativeconclusions regarding concentrations atwhich particulate.matter is likely toproduce health effects. The staff usedthese "quantitative" studies to examineconcentration-response relationships forthe various effects observed in thesensitive populations studied anddeveloped numerical "ranges ofinterest" for possible PMo standards.

A number of uncertainties associatedwith use of these studies must be

considered in selecting an appropriatemargin of safety. As discussed in thestaff paper and the criteria document,epidemiological studies are generallysubject to inherent difficulties involvingconfounding variables and somewhatlimited sensitivity. Moreover, most ofthe quantitative studies were conductedin times and places where pollutantcomposition may have variedconsiderably from current U.S.atmospheres. Most also have usedBritish Smoke* or TSP as particleindicators. None of the publishedstudies used the proposed PMoindicator. Thus, assumptions must beused to express the various results Incommon [PM10 ) units (SP, pp. 96-100).

2. Evaluation of additional margin ofsafety considerations. The criteriadocument identifies a substantial bodyof scientific literature that, while notproviding reliable concentration-response relationships for communityexposure, does provide importantqualitative insights into the health risksassociated with human exposure toparticles. The staff assessed thisliterature, including both quantitativeand qualitative epid6miological studies,controlled human exposure experiments,and animal toxicological studies toidentify additional factors anduncertainties that should be consideredin selecting the most appropriate margin6f safety (SP, pp. 1O-101; 107-111).

3. Selection of the levels that might beconsidered to provide an adequatemargin of safety for the sensitivepopulations of concern. The originalintent of the margin of safetyrequirement was to direct theAdministrator to set an air qualitystandard at some level below thepollution level at which adverse healtheffects have been found or might beanticipated to occur in sensitive groups.Experience with this requirement hasshown that the scientific data are oftenso inconclusive that it is difficult toidentify with confidence the lowest levelat which an adverse effect is "likely" tooccur. Even if such a level can beidentified, available data may suggestthat the effect is also "possible" at lowerlevels, or that other effects (not yetadequately studied) may occur at suchlevels. Thus, the Administrator must stilldecide how far below the "effectslikely" level a standard must be set toprovide a margin of safety that is"adequate."

British Smoke (BS) Is a pseudo-mass Indicatorrelated to small particle (size less than a nominal 4.5jim) darkness. This particulate matter Indicator waswidely used In British and other European studies.See the more detailed treatment of BS in the criteriadocument (CD, pp. 1-88 to 1-90 and 14-0 to 14-11),

10414

Federal Register I Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

Assessments of risks can help guidethis decision, but in the end cannotsubstitute for informed judgment Forexample, if $he health effect detected atlow pollution levels is sevefe, a-greatermargin of safety would be called forthen if it were less troubling. However,given the basic fact that determiningwhat constitutes ai adequate margin ofsafety always come into play belowlhepollution levels at which there isconclusive evidence of health effects.decision-making will necessarily involvevalue-judgments made under asubstantial degree of uncertainty. Thatis particularly true in a case such as thepresent one, where the health evidencesuggests that both the severity of anypotential harm from particulate pollutionand the probability that this harm will infact occur, decrease steadily, but do notnecessarily vanish, as we move to lowerand lower pollution levels.

For these reasons, EPA staff, with- CASAC concurrence, recommended a

range of potential standards for theAdministrator's consideration. Anystandard selected from these ranges can

be said to provide some margin ofsafety. Because of the substantialuncertainties in the available effectsinformation, the complex character ofparticulate matter exposures, and theimportance of the decision, theAdministrator wishes to solicit thefullest possible participation andcomment by the public before decidingwhich standard levels should beadopted in the final standard. He is,therefore, proposing ranges from whichthe final standard is, in each case, to beselected. The rationales for the proposedranges of levels for 24-hour and annualstandards are discussed below.

24-Hour Standard.

The staff assessment of the short-termepidemiological data is summarized inTable 1; particulate matter levels areexpressed in both the original and PMounits. Based on these more quantitativestudies, the staff has distinguishedbetween concentration ranges where, inits judgment, the likelihood of effectsoccurring in sensitive populations is high

.and levels where the data indicate that

such effects may be possible, but areless likely. Therefore. the "effects likely"row in Table I denotes concentrationranges derived from the criteriadocument at or above which thereappears greatest certainty that theeffects listed would occur. While theseeffects are much less likely to occur atlevels below the lower end of thecombined "effects likely" range, thedata do not provide evidence of clearthresholds in exposed populations.Instead, they suggest a continuum ofresponse for a given number of exposedindividuals with both the likelihood(risk) of any effects occurring and theextent (incidence and severity) of anypotential effect decreasing withconcentration. Thus, effects may be"possible" at levels below those listed inthe "effects likely" row, but. because theevidence is less clear, the nature andextent of risks at lower levels are muchmore uncertain. Followifig CASACrecommendations, the staff used thecombined range listed in the "effectspossible" row as a starting point fordeveloping alternative standards.

TABLE 1.-STAFF ASSESSMENT OF SHORT-TERM EPIOEMIOLOGICAL STUDIES (AFTER TABLE 6-2, SP)

Meac.r--d erch orr.ko I : a.n E±,a t-rt PFM-.Effects/study

Va!ry nortarty In London'I AT~.a-;n of trcn tiAZ Ccttd rar~o 4Crcbcs.d =ace'

Effects lkety 500 to 1000 250' to &0"79- _ 2,0 to 30 to 60.Effects leIS*150 to 500 M.2I 1 0 1.. to W0

'Devicatos inl sdforauptyrfore leoweer bounndtrtsc ofsi range.Lodn~r'cs i1~-9ad 4Cre2.2wr~r rm 9572 otjwnamwr

charactenized by igh smoke from ost. combust;lon errfssons t.d h''ih 50•

with lrc J~t op f.L atb! r~rtd Er ,.. 354 1 Ct ci, 15 31. L'r"a .,r ci . 1531).'Exaenatio of sym ptoms reported by bronch'tis in London. Stud 'e' cond,_.cte_ f rom tho n'tj.09. to tho c~ily 197's (t.oett'f"ci ci, 1910). _ . .'Conso n assumes that1 for London smoke condidion, 5S < FM, < TSP. Procis con'."'-" t o am at pot h ~o ,c.'r 'o--.. 0 . (0kS=FL.r,) e mo:rOOO S.Cr15e tT1_5f Cf

safety and msused to estimate the lower bound in th ffects posslotrango. Tho uppcr bom (PM fTSP=E+ 1C-3 o~m .,' I-c ff. PfMwc'= s,dr ,- d to p,"cv - ears.e5•of levels where effects are most Ocely.

'Indicares levels used for upper or lower bouned of range.

The "range of interest" derived fromthe-staff analysis is, therefore, 150 to 350tg/m s as PMo, 24-hour average with no.rnre than one expected exceedance peryear. Under the conditions prevailingduring the London studies, which werecharacterized by high SO2. levels andsmoke from open coal fires, PIoconcentrations near 350 ig/m3 representlevels at which effects are consideredlikely in the sensitive populationsstudied. CASAC concluded that,considering the uncertainties intranslating these results to current U.S.conditions and the seriousness of thepotential health effects, the upper end ofthe original range of interest in the staffpaper contains little or no margin ofsafety and should not be considered asan appropriate standard.

As indicated in Table 1, the study ofLawther et al. (1970) provides the lowestparticulate matterlevel at which healtheffects are judged to be likely by thecriteria document. The effects observed

in this study (related to aggravation ofbronchitis) are of concern both becauseof their immediate impact and becauseof the potential.for inducing longer-termdeterioration of health status in asignificant sensitive group. There wereapproximately 6.5 million bronchitics inthe U.S. in 1970 (DHEW, 1973). Based onthe uncertain conversion betweensmoke and PM1Iooutlined in Table 1, thelowest "effects likely" level derivedfrom the Lawther study (250 jig/mW asBS) should be in the range of 2.0 to 350pg/m3, in PMo units.

Based on using this study alone, aPM1 o standard of 250 ;tg/m5 wouldcontain some margin of safety, even forthe sensitive bronchitics studied,because it incorporates the lower BritishSmoke/PMo conversion factor andbecause of differences betweenexposure conditions in the British studyand current U.S. air quality (SP, pp. 100-101). Because bronchitics are identifiedas a group particularly sensitive to

particulate pollution, a standard of 250jigim" (as PMo) also would provide asubstantial margin of safety for otherless sensitive groups in the population.This concentration is also a factor oftwo below the levels (500 to 750 pgfm5

as BS) where the criteria documentindicates excess mortality begins to beconsidered likely (CD, Table 14-7). Theportions of the population at greatestrisk of premature mortality associatedwith particulate matter exposuresinclude the elderly and persons withpre-existing respiratory or cardiac1

disease. Although the extent of lifeshortening (days. weeks, or years)cannot be specified, the seriousness ofthis effect strongly justifies a margin ofsafety for it (below the "effects likely"levels) that is larger than that warrantedfor the effects on bronchitics.

Taken alone, then. this informationwould tend to support the choice of astandard level of 250 ug/m3 as providingadequate health protection. Several

10-1215

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Piorosed Rules

additional factors. however, suggest theneed for considering a larger margin ofsafety than that provided by a standardwith a level of 250 pig/m 3. These include:

(1) The staff assessment of Londonmortality studies suggests risks ofpremature mortality to sensitiveindividuals at concentrations lower thanthose at which such effects areconsidered likely. Although the risks toindividuals may be small at 250 gg/m s

and below, the number of peopleexposed to lower concentrations insubstantially larger than the numberexposed to higher levels (e.g., Biller,1983). The increased number of sensitiveIndividuals exposed increases the riskthat effects will occur in the totalpopulation exposed.

(2) Information from qualitativestudies assessed in the staff paper (SP,pp. 101-103) suggests risks for sensitivegroups (children and asthmatics) and ofpotential effects (morbidity in adults)not demonstrated in the morequantitative epidemiological literature.Thig qualitative studies do not provideclear information on effects levels, butdo justify consideration for standard-setting purposes of these particulatematter effects that have not beensufficiently investigated.

(3) Differences in composition ofparticles and gases among U.S. citiesand between conditions in the U.S. andthose in London at the time quantitativestudies were conducted add to the '-

complexity of assessing the riskassociated with particulate matterexposures in the U.S.

These factors suggest the need toconsider alternative standard levels thatmight extend from 250 tg/ml down tothe lower bound of the staff range ofinterest (150 pg/m 3 or even below. Inevaluating them, the Administrator ismindful of the uncertain and largelyqualitative nature of the effectsinformation when applied to assessinghealth risks of particulate matter incontemporary U.S. atmospheres.Although the CASAC has concurredwith EPA's assessments of the keystudies conducted in London, several ofthe early British investigators havecontinued to express substantial

disagreement with the interpretations inthe criteria document and staff paper(Lawther. 1982; Holland et al., 1983). Adiversity of opinion among scientists isto be expected when evaluating whatconstitutes an adequate margin of safetybelow consensus "likely" effects levels,but the range of uncertainty isparticularly large in the case ofparticulafe matter.

Because of the substantialuncertainties in the scientific andtechnical information available for usein assessing health risks below 250 Jg/m3 , the complexities associated withapplying available effects informationuniformly to a variety of exposureconditions in different geographicalareas, and the narrow range of factorsthat apparently may be reviewed inmaking this important public policydecision, the Administrator finds itdifficult to choose a single standardlevel from a range of 150 to 250 ttg/m .

Given the relatively low health risks toindividuals at levels below 250 gg/m 3,

the Administrator feels that ifconsideration could be given togeographic variability in such factors asthe number of people exposed, thenature and composition of pollutionexposures and the costs and difficultiesin attaining the standareds, differentlevels-spanning the entire range of 150pg/m s, to 250 ,g/m 3-- might beconsidered appropriate for differentareas. The present Act, however,appears to proclude consideration ofcosts and feasibility in setting NAAQSand focuses the basis for a decision onnational public health protection on"worst case" exposure situations.

Given the precautionary nature ofsection 109 and the factors the Actpermits hini to consider, theAdministrator is inclined to select a 24-hour standard from the lower portion ofhe above range. Because of the wideuncertainties and the significance of thisdecision, however, he believes it isimportant to air the issues anduncertainties fully and to encouragebroad public participation and commentbefore choosing a specific level for thestandard. Therefore, the Administratorproposes to select a final standard level

from a range of values between andincluding 150 to 250 pg/m3, and solicitspublic comment on the standard levelthat provides an adequate margin ofsafety given the risk of effects suggestedby the available scientific information.

In proposing this range for the 24-hourstandard, the Administrator also invitespublic comment on an EPA analysis,developed subsequent to CASACclosure on the criteria document andstaff paper, that is of some relevance tothe final decision on the level of thatstandard. The analysis (Morgenstem,1982), which has been placedin therulemaking docket, is a furtherevaluation of the London mortality dataand examines the issue of whether cleareffects thresholds can be identified. Theanalysis adds to the evidence thatwould suggest some possibility of effectseven at levels below 75 1tg/msBS (75 to175 t.g/m 3 as PM lo).

Although the analysis is relevant tothe final decision on a standard leveland should be considered during thepublic comment period, theAdministrator did not consider it indeveloping the proposed standard rangebecause it is an unpublished staffanalysis that has not been consideredby CASAC and because of possiblelimitations identified in initial publiccomment on it (Young, 1982). Thesecomments have been placed in therulemaking docket.

Annual Standard

The staff assessment of importantlong-term epidemiological data issummarized in Table 2. Long-termepidemiological studies are subject toadditional confounding variables thatreduce their sensitivity and makeinterpretation more difficult than forshort-term studies. The "effects likely"levels are derived from the criteriadocument, but again, no clear thresholdscan be identified in exposed populationsfor all indicators of health effects,Effects may occur at lower levels, butthe evidence is inconclusive and effectsare difficult to detect in the availableepidemiological studies.

TABLE 2.-STAFF ASSESSMENT OF LONG-TERM EPIDEMIOLOGICAL STUDIES (AFTER TABLE 6-3, SP)

Measured 8S levels (as ..gI Measured TSP levels (jlg/mr) 5 "ialontEfecslluyml)--increased respirator PaM, 10evelsdotsadisease reduced y Icesed rratory iease Increased cr Cornd ranged (flsm)in children cn reduction In leer ralung function in adults

2 symptoms itsr'nge

Effects likely .............. .......... 230 to 300 BS...........................*230t. 180. ................ ...................................... >180 .... . ............. 90 to 110,Effe:ts possible- .............................<230 BS. ..... . 130 to 180' ..................-. 60 to 150 (110') .......... 110 to 10 .................... 55 to 10,No rignificant effects noted ................ ... ................ 9o' to 130 .. .. . .... ... ............ - . ...... So0 l0...........400 55.Study conducted in 1963-65 In Sheffield, England (Lunn et al.. 1967). BS levels (as yg/rm) uncertain., Studies conducted In 1961-73 in Berlin, NH. Major source in community was pu'.p meit Effects level (180 pg/mr'~bassd on 2-month avarago. Effects on lung function were relatilvel small'Ferns et al. 1973, 1976).

C?

1 ........................

10416

Federal Register / Vol. 49, No. 55. / Tuesday, March 20, 1984 1 Proposed Rules

.Study conducted in 1973 (Bouhuys et al, 1978). Exposums refloct 1965-73 data in Ason!. CT. Medsin vasJo =sd cs an cstze. Esz-cn± _PtMvo Stu$ l.o c2 ucxs en tufunctI. bu some suggestion of effects on respty 'toms.

4 Conversion based on estimated ratio of PMS for cutrent (13M-81) U.S. atmosphE-s Tt rao ci tratd IL use n t Wtoff po cr ra 'cd tc- we .e a. 0-5 to 0.6. 1c--0Jmbers were used as lower and upper bounds for estimatang PMo eqJAalents from TSP wstcs Mae receM ana-' tho ftmacd ra.o trf bo aa ram as 045 to 0.5 (Pare. ISM

SIndcates levels used for upper or ower bound of range.

Based on the staff assessment, the"range of interest" for examiningpotential PMo standards was 55 to 110/.g/m3, annual arithmetic mean. Becausethe original studies measured TSP, someuncertainty exists in deriving precisePMo levels associated with possibleeffects.* Moreover, the upper end of thisrange overlaps the somewhat uncertain"effects levels" derived from thesestudies. CASAC felt that, due to theseuncertainties, the upper end of the range(110 jig/ml} may not include any marginof safety, and should not be consideredas an appropriate standard alternative.

The lowest "effects likely" levelidentified in the assessment summarizedin Table 2 is 90 jg/m3 as PM10, althougheffects are possible at lowerconcentrations. The effects of mostconcern relate to the possibility of long-term deterioration of the respiratorysystem in exposed populations, thepotential for which is indicated by lungfunction (mechanical pulmonary]changes and increased incidence ofrespiratory disease. One set of studies(Ferris et al., 1973, i976) provides someevidence for a "no observed effects"level at or below 60 to 65 Lg/m3 (130 Lg/Ina as TSP] while another, essentiallynegative study (Bouhuys et al., 1978],suggests some possibility ofsymptomatic responses at long-termmedian levels at or below about 50 to 55pg/m3 as PM. It is not clear whetherthese symptomatic responses, -whichwere unaccompanied by lung functionchanges, represent adverse healtheffects.

- A PMo standard of 60 to 65 jg/m3

would provide some margin of safetybased on the studies of Ferris and co-workers, but would leave some smallremaining risk of symptomaticresponpes. Because of associateduncertainties [SP, pp. 104-110) as well asthe limited scope and number of theselong-term quantitative studies, it isparticularly important to examine theresults of qualitative data from anumber of epidemiological, animal, andair quality studies when evaluating whatconstitutes an adequate margin of safety

*As noted in Table 2, the original staff paperassessment assumed a PMo/TSP ratio of 0.5 to 0.6.The discussion that follows also takes into accountthe results of more recent analyses suggesting thatthe median ratio may be as low as 0.45 to 0.5 (Pace.193). Applying this more recent information, thelower bound of the range of interest is reduced to 50pg/rm. Because this level represents a multi-yearaverage in the original study, no adjustment isneeded when considering this level as an expected

nal'mean.

for an annual standard. These studiesjustify concern for serious effects notdirectly evaluated in the studies listed inTable 2. Such effects include damage tolung tissues contributing to chronicrespiratory disease, cancer, andpremature mortality (SP. pp. 109-111).Substantial segments of the populationmay be susceptible to one or more ofthese effects (SP, p. 46). The availablescientific data do not suggest majorrisks for these effects categories atcurrent ambient particle levels in mostU.S. areas. Nevertheless, the risk thatboth fine and coarse particles mayproduce these responses supports theneed to limit long-term levels of PMlo fora variety of ambient aerosolcompositions.

Although the qualitative data do notprovide evidence for major risks ofthese effects categories at currentannual particulate matter levels in mostU.S. cities, the Administrator believesthat the weight of the evidence, theseriousness of the potential effects, thelarge population at risk, and the recentinformation on converting TSP values toPMo levels warrant caution in settingthe standard, and suggest considerationof standards in a range of 50 to 65 ILg/in. Given the precautionary nature ofsection 109 and the factors the Actpermits him to consider, theAdministrator favors a standard in thelower portion of this range. For reasonsarticulated in the discussion of the 24-hour standard, the Administrator hasnot selected a single level, but proposesthat the level of the annual standard beselected from a range of 50 to 65 g/mr(as PMo), expected annual arithmeticmean. Public comment is solicited onwhat standard within this range thatprovides an adequate margin of safetyagainst the risk of effects suggested bythe available scientific information.

RATIONALE FOR THE SECONDARYSTANDARDSIntroduction

Where secondary standards areconcerned, the question of consideringthe costs and difficulties of attainmentstands on a somewhat different footing.The Clean Air Act's legislative historycontains a number of statementsemphasizing that health must be theexclusive basis for setting primarystandards, but no parallel languageconcerning secondary standards. To seta secondary standard that literallyeliminated all welfare effects from air

pollution could lead to very extremecontrol requirements. Indeed. sincesome regions are naturally dustier thanothers, and some have better naturalvisibility than others, a literal readingcould compel those naturally dusty orlow-visibility regions to clean up evenbeyond their natural background levels.Congress appeared to recognize some ofthese difficulties in that the Act allvsthe Administrator to use judgment indetermining what constitutes an adversewelfare effect and the Act does notrequire the same stringent timetable forthe ,secondary standards as for theprimary standards.

If such a literal interpretation of thestatute were truly intended, there wouldbe no need for separate Clean Air Actprograms to protect visibility or guardagainst significant deterioration. YetCongress included these programs inthe1977 amendments to the Clean Air Act,and gave as one of its reasons that thesecondary standards did not protectagainst all welfare effects, and thatadditional measures were thereforeneeded. See 1977 House Report pp. 204-05. Congress gave no indication that itexpected revisions to the secondarystandards that would change thissituation.

Congress therefore seems to haveassumed that the secondary standardswould not literally protect against allwelfare effects. But if the secondarystandards are therefore not to be setbased on a literal reading of the statute,that raises once again the question whatfactors may be considered in settingthem.

In this regard, it is worth noting thatstriking anomalies arise in attempting toset welfare-based air quality standardsaccording to some welfare effects butnot others. For example, a tighter airquality standard might lead to areduction in soiling and nuisance.However, such a standard could causean increase in the price of electricitythus reducing the demand for airconditioning and increasing thediscomfort factor. Both soiling andnuisance and personal well-being andcomfort are welfare effects but in thecase cited, a literal interpretation of theAct would allow one to be used but notthe other in setting the secondarystandards.

When the Agency first considered thisissue in 1972, a staff memorandum(Schwartz, 1972) from the GeneralCounsel's office argued that, though the

10n.7

Federal Register / Vol. 49, No. 55 / Tuesday, March'20, 1984 / Proposed Rules

better statutory reading was that costsand attainment problems could not beconsidered in setting secondarystandards, the opposite conclusion couldbe supported by forceful arguments. Asindicated previously, the D.C. Circuithas addressed the issue twice in thecontext of primary standards. LeadIndustries Association v. EPA, supra,647 F.2d at 1148-51; American PetroleumInstitute v. Castle, supra, 665 F.2d at1-185, 1190. Neither case, however,involved a serious challenge tosecondary standards on cost grounds.Although the opinions contain languagethat can be taken as hearing on thesecondary standards issue, it isuncertain whether the court would havearrived at this conclusion in a casepresenting issues of general welfareunder secondary standards rather thanpublic health under primary standards.

The Agency has not revised itsposition, based on the 1972 legalmemorandum, that the better legal viewis that attainment costs should not beconsidered in setting secondarystandards. However, the Agency has nothad occasion to face the issue directlysince passage of the 1977 amendments.The issue remains a troubling one, worthfurther examination, particularly sincethis proposal sets forward a secondarystandard that could have controlimpacts beyond those due to theprimary standard. The Agency thereforeinvites comment on this issue.

The criteria document and staff paperexamined the effects of particulatematter on such aspects of public welfareas visibility and climate, man-madematerials, vegetation, and personalcomfort and well-being. Each isdiscussed in some detail in the criteriadocument and staff paper. The followingdiscilssion of the rationale for thesecondary standards focuses primarilyon the considerations that were mostinfluential in the Administrator'sselection of a particular option, or thatdiffer in some respects from or expandupon considerations that influenced thestaff and/or CASAC recommendations.

Soiling and NuisanceAt high enough concentrations, both

large and small particles may soilhousehold and other surfaces, orotherwise become a nuisance. Botheffects result in increased cost anddecreased enjoyment'of the environment(SP, p. 140). Efforts to control particulatematter in U.S. cities from 1970 to 1978are estimated to have producedsubstantial economic benefits becauseof reduced soiling and nuisance (CD, p.1-51). Based on these factors, the staffpaper recommends consideration ofsoiling of materials and nuisance

generated by dust and other particles indeciding upon a secondary standard (SP,p. 141).

In considering secondary standard(s)for particulate matter, the Administratorfirst determined whether the pollutantindicator (PMto), averaging times andform, and range of levels of theproposed primary standards wouldprovide adequate protection against theknown or anticipated adverse welfareeffects associated with soiling and •nuisance. The decision with respect toeach of these areas is discussed below.Pollutant Indicator

Although both large and smallparticles can contribute to soiling andnuisance, the available scientific datado not provide quantitative informationon the relative importance of varioussize fractions. The proposed indicatorfor the primary standards, PMo, couldalso be considered as a useful indicatorfor a secondary standard based onsoiling and nuisance because (1) smallparticles in the less than 10 jim sizerange are more likely to penetrateindoors and soil vertical surfaces (SP,pp. 136-137) and (2) due to thecharacteristic size distribution andorigin of particles in the atmosphere (SP,pp. 14-19), control of particles less than10 gim would also limit theconcentration of large (coarse mode)particles, to some extent.. Nevertheless, the criteria document

and everyday experience make it clearthat particles larger than 10 jim,including visible dust and large soot,can soil horizontal and other surfacesoutdoors and present a nuisance. APM1o indicator would exclude the 50percent or more of suspendedparticulate -matter mass that is largerthan 10 gim, and the ratio of PMo tolarge particle levels can varysubstantially from city to city.Therefore, PMo would be an incompleteindicator for large particles and its usecould result in a relatively large degreeof variability in welfare protectionamong cities with-differing PM~o/TSPratios.

The current indicator for thesecondary standard, TSP, incorporates alarger-portion of the particles that cancontribute to soiling and nuisance.Although TSP does not include some ofthe largest particles of concern nearestsources and the efficiency with which itcollects such particles can vary withwind speed, control requirements forTSP would result in reductions in theselarger particles as well. Most of theavailable information on such effects isrelated to TSP measurements,facilitating evaluation of alternativelevels without the additional uncertainty

added by converting to a differentindicator. Given the lack of datapermitting a clear distinction amongparticle size ranges with respect tosoiling and nuisance, the more Inclusivenature of TSP, and the use of TSP Inavailable effects studies, theAdministrator proposes to retain TSP asthe indicator for a secondary standarddesigned to protect against soiling andnuisance.

While proposing to retain a TSPsecondary standard, the Administratorsolicits public comment on anotheroption that is under consideration, thatof making the secondary standardequivalent in all respects to theproposed primary standards for PM 0.As discussed above, both TSP and PMocould be useful indicators for thoseparticles responsible for soiling andnuisance. Based on the availablescientific data, the choice between thetwo is difficult. Depending on the exactlevels of primary standards chosen, thecombined requirements for meeting both24-hour and annual primary standardsfor PM1o might be considered adequateto protect against possible adversewelfare effects related to soiling andnuisance from all relevant particle sizes,This approach would also simplifymonitoring and reporting requirementsbecause information hind controlstrategies would be required for onlyone pollutant.

Averaging Time and Form

Soiling and nuisance may result fromboth short-term dust episodes andlonger-term accumulations of particles(SP, pp. 136-138). Most studies thatindicate a relationship betweenparticulate matter and these effects,however, have been based on annualaverage levels (SP, pp. 138-139).Moreover, strategies to prevent longer-term impacts appear to be moreeffective for protecting public welfarethan those attempting to controltemporary localized problems, such as afew days of windblown dust or pollen,Based on these considerations, anannual averaging time would appear tobe more appropriate than a 24-hourstandard. Therefore, the Administratorproposes to replace the current 24-hoursecondary TSP standard with an annualstandard.

For reasons presented in thediscussion of proposed annual primarystandards and for consistency inaveraging and reporting betweenprimary and secondary standards, theAdministrator proposes an arithmeticmean and statistical form for the annualsecondary standard. The proposedinterpretation of this form Is Identical to

10418

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

that for the primary particulate matterstandard and is detailed in Appendix Kof the proposed regulation.

Level

The available data base providescompelling evidence that elevated levelsof particulate matter can produceadverse welfare effects, but provideslittle quantitative information onconcentration-effects relationships.Physical damage and economic studiestend to show no obvious welfare effects"thresholds" for soiling. With time,particulate matter may accumulate onsurfaces even at low concentrations. Atvery low concentrations, the amounts ofparticulate matter may be virtuallyinvisible to the human eye or be soslight as to be ignored by most people(Carey, 1959; Hancock et al., 1976). Up toa point, the buildup of particles onsurfaces may not be generally regardedas a social problem because it isremoved by rain or routine cleaning andmaintenance before substantialaccumulation can occur. Thus, thecritical judgment for selecting astandard level is to determine aparticulate matter concentration atwhich the soiling effect is significantenough .hat it should be regarded as an"adverse" effect under section 109(b)(2)of the Act.

The available information suggeststhat the public makes a distinctionbetween concentrations at whichparticulate pollution is noticeable andhigher levels at which it is considered anuisance. A study of the response of apanel of human subjects to dust onsurfaces concluded that the level ofdustiness that is found to beobjectionable is higher than the levelthat can be perceived or discriminated(Hancock et al., 1976). No uniqueadverse particulate matter levels were,however, derived from this study. Amore direct study of perception of airpollution as a nuisance (CD. pp. 9-67)suggests that the subjects began toconsider air pollution a nuisance inareas where annual levels were at orsomewhat above the level of the currentannual primary TSP standard. Opinionsexpressed by some-CASAC members(November, 1981 transcript, pp. 63-66)supported this suggestion. Thecommittee as a whole did not, however,make any recommendations regarding arange of interest for a secondary TSPstandard.

Several studies of economic effectsassociated with soiling by particulatematter are discussed in the criteriadocument. These studies suggest thepossibility of substantial economicbenefits in moving from TSPconcentrations equivalent to the current

primary annual standard to levels aslow as the current annual guide forattaining the secondary standard (CD,pp. 10-5 to 10-69). The criteriadocument points out the tentative andlargely qualitative nature of thesestudies, but uscs them to provide acrude estimate of the magnitude anddirection of benefits in reduced outdoorsoiling associated with decreased TSPlevels in .. S. cities. A 20 percentimprovement in TSP from a startingpoint close to the current primaryannual standard resulted in anestimated national economic benefit of$0.2 to $0.7 billion/yr in 1978 dollars(CD, p. 10-73). As discussed in thecriteria document and the staff paper(SP, p. 139-140). the original studies aretoo uncertain to make this anything buta crude qualitative estimate.

The studies and information discussedabove can be used to suggest a range foralternative annual secondary standards.Based on the study of perception andsuggestions by individual CASACmembers, a TSP standard at or near thecurrent annual primary standard mayprotect public welfare against adverseeffects related to soiling and nuisanceby particles. Even though the availabledata are qualitative and uncertain, theweight of evidence would suggest that itis unwise to reverse the progresstowards increasing benefits associatedwith implementing the current primarystandard. Accounting for the change toarithmetic mean and statistical form, alevel of 80 to 90 ,g/m 3 expected annualarithmetic mean would be consistentwith maintaining welfare protectionequivalent or near to that provided bythe present annual standard. Theselevels form the upper bound of the rangeof interest.

The rough economic estimates in thecriteria document suggest the need toconsider the possibility of morestringent standards, at levelsapproximately equivalent to the currentannual guide for the secondarystandard. Applying appropriateadjustments for form and averagingtime, a standard of about 70 yg/m

3

annual arithmetic mean isapproximately equivalent to the currentannual guide. This level forms the lowerbound for the range of interest.

The range of interest for thesecondary standards developed above isthus 70 to 90 pIg/mn, expected annualarithmetic mean. Because CASAC andthe public have not reviewed orcommented on this range and because ofthe uncertainties in the information, theAdministrator is proposing to select thefinal standard from within this rangeand has not decided upon a single

standard level. Given the largeuncertainties in the availableinformation, the Administrator isinclined to continue the level ofprotection provided by the currentannual TSP standard and select a levelfrom the upper portion of the aboverange. The Administrator solicits publiccomment on the most appropriate level,and in particular seeks the guidance ofknowledgeable State and local airpollution officials with respect to levelsat which the public appears to considerparticulate matter to be a nuisance.

Other Welfare Effects

The other welfare effects ofparticulate matter of principal interestinclude impairment of visibility,potential effects on climate, andcontribution to acidic deposition. Thesepotential effects are most stronglyrelated to regional scale fine particlelevels and any standards and controlswould likely involve regional sulfuroxide emissions (SP, p. 147; Friedlander,1982). The staff and CASAC pointed outthe advantages of recognizing theinterrelated aspects of the known andpotential effects of fine particles onvisibility and climate together with theacidic deposition phenomenon whenconsidering a possible fine particlestandard.

In view of these considerations, theAdministrator has decided to defer adecision on a possible fine particlesecondary standard until it is possible tolink such a standard with a coherent,scientifically based strategy for theserelated regional air quality problems.EPA is continuing to evaluatealternative approaches to address acidicdeposition. In parallel with this effort,the Agency is examining theimplications of acidic deposition controlstrategies on visibility and other airquality values. The results of thisexamination together with otherrelevant information will be used inpreparing an advance notice ofproposed rulemaking soliciting publiccomment regarding a possible fineparticle secondary standard.

The Administrator also concurs withthe staff suggestions that a separatesecondary particle standard is notneeded to prdtect vegetation or toprevent adverse effects on personalcomfort and well-being.

NEED FOR ADDITIONAL RESEARCH

A troublesome aspect of this standarddecision is the large uncertainty inavailable scientific information. In thisregard. CASAC transmitted to theAdministrator its review of the scientificdata base and assessment of research

10419

Federal Register / Vol. 49, No. 55 /Tuesday. March 20. 1984 / Prooosed Rules

needs for particulate matter as well as'several other criteria pollutants(Lippmann, 1983). In this most recentassessment, the committee indicatedthat "the research needs for particulatematter (PM) are substantially greaterthan those for any of the other criteriapollutants discussed in this report" andthat their "review also revealed a highlylimited data base, particularly wherequantitative studies were concerned,and a wide range of views about theeffects of specific constituents of PMand the exposure levels at whichadverse health or welfare effects arelikely to, or may possibly, occur." Whilethe report points out a number ofspecific areas where additional researchis needed on both health and welfareeffects, the main areas include:

1. The physicochemical properties ofPM as they affect health;

2. The nature of health effects of PMand its major constituents;

3. Quantitative relationships betweenPM and health effects in sensitivepopulation groups;

4. Effects on visibility; and5. Soiling and nuisance effects.The Agency is examining its research

program in these areas and will, to theextent resources allow, incorporate theCASAC recommendations in itsresearch planning process.FEDERAL REFERENCE METHOD.Approach

The current appendices to 40 CFR Part50 describe requirements for referencemethods to be used for measuring eachof the pollutants for which an NAAQShas been established. For each .pollutant, a corresponding appendixspecifies either (1) a complete, unique,manual reference method (e.g., TSP.SO 2, Pb) or (2) a measurement principleand calibration procedure applicable toautomated reference methods, whichmust also meet performancerequirements specified in 40 CFR Part 53(e.g., CO, 03. NO2).

Accordingly, a new Appendix J isproposed today to describe therequirements for reference methods forPMo. The proposed Appendix, however,would deviate somewhat from theestablished scheme in that the methoddescribed is considered a manualmethod even though the requirementsresemble the measurement principle andperformance requirements normallyprescribed for automated methods. Thisapproach would provide greaterflexibility, allow the use of currentlyavailable particulate matter samplers.and encourage continuing improvementsand innovative sampler design.

. Many of the concepts and approachesunderlying the proposed PMo referencemethodrequirements have beenestablished on-the basis of extensiveconsultation, assistance, and generalconsensus among numerous particulatematter monitoring experts both withinand outside EPA. In particular, EPAsponsored two workshops (October 1979and November 1980) that broughttogethermany of these experts to helpformulate size-specific monitoringrequirements {Kashdan and Ranade.1982).

The proposal of PMo as an indicatorof particulate matter is discussed in anearlier section of this notice. The PM,0indicator is, in effect, defined b r thesampling requirements, which in turnwere designed to approximate particlepenetration to the thoracic region of thehuman respiratory tract (as developedfrom studies using mouthpieces). Inpractice, however, the mass collected byambient samplers is more dependent onthe cut point than on matching otherparameters of respiratory tractpenetration (Rodes et al., 1981). Thus,samples with a 50 percent cut point of 10prm that meet the size discriminationspecifications in 40 CFR Part 53 willprovide a reliable estimate of the totalmass of suspended particulate matterless than or equal to 10 pim inaerodynamic diameter. The proposedrequirements specify a 50 percent cutpoint of 10_1 j. m and designate totalmass collection tolerances for typicalatmospheric particle distributions.

The proposed reference method forPM,0 is based on discrimination andselection of PMo particles by inertialseparation in a specially shaped inlet,followed by conventional filtration of ameasured volume of sampled air anddetermination of the net weight gain ofthe :filter. The normal sampling periodwould be 24 hours. The sizediscrimination characteristics of thesampler (or sampler inlet) would berequired to meet specifications and betested according to explicit testprocedures prescribed in 40 CFR Part 53.Methods for PMo that meet allrequirements would be designated asPMo reference methods according to themethod identification, which in mostcases would probably be themanufacturer and model of the samtpler.

Wherever feasible, the requirementsin Appendix I are prescribed asfunctional or performance specificationsrather than design specifications toallow maximum flexibility in designingor configuring PM0 samplers. Samplershape, inlet geometry, -filter material andsize, operational flow rate, degree ofautomation, and so forth are all

specified in terms of required function orperformance.

The proposed reference methodprescribes no specific flow-rate for PMiosamplers. High flow-rate samplers, suchas the existing high volume samplersequipped with a size-selective inlet,would be approved if they meet theother performance requirements,Medium- and low-flow samplers wouldalso be allowed, provided enough massis collected to meet the reproducibilityspecification. PM1o samplers would berequired to minimize measurementerrors resulting from filter handlingoperations such as conditioning,weighing, and installation in andremoval from samplers. The alkalinityspecification would disallow the use ofexcessively alkaline media that couldlead to possible measurement errorsfrom artifact formation on the filter.Filters meeting this specification shouldshow little or no sulfate artifact but maybe subject to errors due to nitrateartifact. However, for most samplinglocations, PMo mass concentrationerrors due to positive nitrate artifact orloss of nitrate by volatilization orchemical rection are expected to besmall.

The proposed filter mediumspecifications are intended to beminimum requirements whenmeasurement of PMo massconcentration is of primary concern.Users would need to consider additionalfilter medium criteria if other samplingobjectives were to be met.

Reference method samplers must havethe capability of collecting validsamples at ambient concentration levelsconsistent with the proposed NAAQS, A.filter's capability of withstanding highparticle mass loadings under a variety ofenvironmental conditions withoutoverloading (clogging) is an importantconcern. Ideally, the pressure dropacross the filter should be sufficientlylow to minimize the potential foroverloading. To further minimize thispossibility, samplers should have thecapability of maintaining normal flowrates over as wide a range of pressuredrop as is practical.

Sampler manufacturers areencouraged to consider the potential forfilter overloading in designing theirsamplers and in selecting theirrecommended filter media. Samplersthat incorporate filter-changingmechanisms to automatically switchfrom a particle loaded filter to a fresh,filter would be allowed, provided 'thereproducibility of the PMo measurementmet the required specification.

Sampler manufacturers would berequired to provide sampler purchasers

FId lq i p

10420

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rufes

with an operation or instruction manualcontaining detailed procedures forcalibration and operation of thesampler, as well as recommendationsregarding appropriate filter media andtype of analytical balance required formass determinations. Such a manualwould be required by 40 CFR Part 53,which would require submission of themanual as part of the manufacturer'sapplication for a reference methoddetermination.

Technical Change to Appendix GThe high-volume method described in

Appendix B will continue to be used forthe proposed secondary standard and inconjunction with Appendix G("Reference Method for theDetermination of Lead in SuspendedParticulate Matter Collected fromAmbient Air") and for other purposesthat may be specified. Accordingly, EPAproposes to delete reference 10 inAppendix G and to revise section 5.1.1of the Appendix to read as follows:"High Volume Sampler. Use andcalibrate the sampler as described inAppendix B to this Part" The Appendixwould also be revised to specify moredirectly that the high-volume methoddescribed in Appendix B is to be used inconjunction with the reference methodfor lead.PREVENTION OF SIGNIFICANTDETERIORATION

Pursuant to a settlement withpetitioners in ChemicalManufacturersAssociation v. EPA, D.C. Cir. No. 79-1112, EPA agreed to propose revision to

.certain requirements for the preventionof significant deterioration of air quality(PSD), if appropriate, at the time itproposed revisions to the particulatematter standards. Section 163 of the Actlists numerical increments for"particulate mtter" that limit increasesin ambient concentrations of thatpollutant over established baselinevalues. The Act, however, does notdefine "particulate matter." EPA hastraditionally defined this pollutant interms of TSP, because the air qualitystandards have used that indicator. In amajor decision in 1979, the D.C. Circuitindicated-though it was not essentialto the resolution of the issues then-athand-that EPA has discretion to defineparticulate matter differently, providedthat the definition encompassed allparticles having a substantial health orwelfare effect and excluded only thosehaving no such impact. Alabama PowerCompany v. Castle 636 F.2d 323, 370 n.134. Implicit in this view is that theAgency may not change -to a definitionof particulate matter that would excludeparticles that are known to have

adverse impacts on public health orwelfare.

Various industries, especially themining industry, have sought relief fromthe PSD increment based on TSP. In itslitigation settlement with industrypetitioners in Chemical ManufacturersAss'n v. EPA, EPA agreed to proposecertain regulatory changes, but only tothe extent that they could be technicallysupported. In relevant part thesettlement stated that "[w]hen EPAproposes a new size cutoff for purposesof the NAAQS, it shall also propose (a)a new size cutoff for PSD purposes thatwould remain in effect indefinitely (i.e..the 'permanent PSD cutoff') and (b) anditerim size cutoff for PSD purposes thatwould remain in effect until EPA takesfinal action on the permanent PSDcutoff. The interim cutoff will excludeonly those particles which clearlyappear not to pose substantial healthand welfare risks and therefore arehighly likely to be excludedpermanently."'

As discussed in the rationale for theprimary and secondary standards,although th6Agency's review of thedata suggests that particles larger thanten micrometers might be safelyexcluded from the primary standard,such particles can contribute to soilingand nuisance and therefore may havesubstantial welfare effects. TheAdministrator is proposing to retain TSPas the indicator for the secondarystandard in large part because itincludes most of these large particlesand may therefore be a better indicatorof all particles that produce soiling andnuisance. Thus, by the terms of the CMAagreement itself, there appears to be noappropriate cutoff below that for TSP forPSD purposes. and the contemplatedinterim and permanent relief is notavailable. Therefore, EPA proposes notto change how "particulate matter" isdefined for purposes of the PSDincrements. EPA solicits comments onthis proposal.

As discussed above, EPA in responseto comments conceivably might adopt asize cutoff below that for TSP forpurposes of the secondary standard.Even if it aoes that, there is considerablequestion as to whether EPA could thenalso adopt the same cutoff for purposesof the PSD increments. EPA thereforesolicits comment on this possible change

'EPA also agreed to publish certain guidanceregarding postponement of SIP rcvlsions forcorrection of violations of PM increments and forIssuance of basic State new source review permits(i.e.. permits Issued under State regulations meeting

-40 CFR 51.18[a)-(i)} to sources that would cause aviolation of a PM increment. That guidance wasincluded in EPA's proposed rulemaking of August25.1983.48 FR 38742.

to the increments and in particular onthe merits of the position that EPA hassome discretion to define particulatematter for PSD purposes.

EPA has established a separatedocket for the review of the definition ofparticulate matter for purposes of theprevention of significant deteriorationincrements. Comments on this issueshould be sent to this docket (DocketNo. A-83-48) and not to the rulemakingdockets established for the proposednational ambient air quality standards.REGULATORY ANDENVIRONMENTAL IMPACTS

Regulatory Impact Analysis

Under Executive Order 12291, EPAmust judge whether a regulation is a"major" regulation for which aRegulatory Impact Analysis (RIA] isrequired. The Agency has judged theparticulate matter NAAQS proposal tobe a major action, and has prepared adraft RIA based on informationdeveloped by several EPA contractors(inter alia. Argonne, 1983; Mathtech,1983). It includes estimates of costs,benefits, and net benefits associatedwith alternative standards. The draftanalysis, entitled Regulatory ImpactAnalysis of the National Ambient AirQuality Standards for ParticulateMatter-Draft (EPA. 1983), is availablefrom the address given above (seeAvailability bf Related Informationsection). A final RIA will be issued atthe time of promulgation.

Neither the draft RIA nor thecontractor reports have been consideredin issuing this proposal. TheAdministrator has not seen thesedocuments nor has he been briefed ontheir contents. As had been noted,several recent judicial decisions makeclear that the economic andtechnological feasibility of attainingambient standards are not to beconsidered in setting them, althoughsuch factors may be considered to adegree in the development of State plansto implement the standards. The Agencyis currently considering the generic issueof the role. if any, of benefits analysis,or parts thereof, in.setting ambientstandards. If the Agency concludes thatinformation from benefits analyses maybe legally and technically relevant tostandard setting, Agency staff wilreview the benefits information in theparticulate matter RIA and contractorreports to determine whether or not anyportions of if appear to be relevant inthis rulemaking. Becuase the approachesused and the results have not beensubject to extensive peer relview, EPAwould then submit the portions thought

1049-1

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

to be relevant to CASAC and the publicfor comment on their scientific andtechnical adequacy and whether theyshould be considered in the finaldecision on the particulate matterstandards.

The draft RIA has been submitted tothe Office of Management and Budget(OMB) for review under ExecutiveOrder 12291. Any comments from OMBand any EPA responses to thosecomments are available for publicinspection at EPA's Central DocketSection (Docket No. A-82-37), WestTower Lobby, Gallery 1, Waterside Mall,401 M Street, S.W., Washington, D.C.Impact on Reporting Requirements

This proposed rule does not containany information collection requirementssubject to OMB review under thePaperwork Reduction Act of 1980 U.S.C.3501 et seq.

Impact on Small EntitiesUnder the Regulatory Flexibility Act, 5

U.S.C. 600 et seq., the Agency mustprepare a regulatory flexibility analysisassessing the impact of any proposed orfinal rule on small entities. Under 5U.S.C. 605(b) this requirement may bewaived if the Agency certifies that therule will not have a significant economiceffect on a substantial number of smallentities. Small entities include smallbusinesses, small not-for-profitenterprises, and governmental entitieswith jurisdiction overpopulations of lessthan 50,000. EPA has made an effort toassess the potential impacts on smallentity groups as part of the economicimpact analysis in Section V.F. of theRIA (EPA, 1983). The preliminaryassessment based on selected industriesdoes not suggest that the proposedrevisions will significantly affect asubstantial number of small entities forthe group of potentially affectedindustries. Forxreasons outlined below,however, this analysis is limited anddoes not permit definitive findings withrespect to all potentially affected smallentities.

The major analytical difficulty resultsfrom the extremely large number ofpotentially affected industries (over 280at the 4-digit SIC code level). This makesit impractical at present to gather andanalyze all of the information on plantsize and ownership that would benecessary to perform detailed economicimpacts analyses on each small entitygroup. In an attempt to develop some ofthe information necessary for aregulatory flexibility analysis, ascreening analysis was performed inSection V.D. of the RIA (EPA, 1983). Thisanalysis selected 16 of the most affectedindustres (out of a total of 280

industries] for further examination. Forthe 16 industries as a group, thepercentage of potentially affected plants(both large and small) was less than 20percent, which would suggest'that thepercentage of smaU entities affectedwithin the 280 industries would not besubstantial (ie., not greater than 20percent). Nevertheless, this limitedanalysis does not permit a cleardetermination of whether significantimpacts would be incurred by a ,substantial number of small entities.

Additionally, after promulgation ofnational ambient air quality standards,the control measures necessary to attainand maintain them are developed by therespective states as part of their stateimplementation plans. In selecting suchmeasures, the states have considerablediscretion so long as the mix of controlsselected is adequate to attain andmaintain -the ambient standards.Whether a particular standard wouldhave a significant effect on a substantialnumber of small entities then depends tosome extent on how the states wouldchoose to implement it. For thesereasons, any assessment performed byEPA at this time is necessarilysomewhat speculative. Moreover,although the proposed standards mayimpact some small entities when theyare implemented by the states, itappears that this factor, like othereconomic and technological feasibilityfactors, cannot affect the Agency'sdecision.

OTHER REVIEWSThis proposed rule was submitted to

the Office of Management and Budget(OMB) for review. Any comments fromOMB and any EPA responses to thesecomments are available for public

, inspection at EPA's Central DocketSection (Docket No. A-82-37), WestTower Lobby, Gallery I, Waterside Mall,401 M Street. S.W., Washington, D.C.List of Subjects in 40 CFR Part 50

Air pollution control, Carbonmonoxide, Ozone, Sulfur oxides,Particulate matter, Nitrogen dioxide,Lead.

Dated: March 8. 1984.William D. Ruckelshaus,Administrator.

ReferencesAMC [American Mining Congress] (1982).

American Mining Congress Position Paperon Particle Size Issue. American MiningCongress, Washington, D.C. June 17,1982.Docket #A-79-29, II-D-81a.

Argonne (1983). Costs and Air QualityImpacts of Alternative National AmbientAir Quality Standards for ParticulateMatter, Technical Support Document,prepared for U.S. EPA, Ambient Standards

Branch, Research Triangle Park, N.C.,January 1983. Docket #A-79-29, I-A-5,

Avol, E.L., W.S. Linn, D.A. Shamoo, T.G.Venet and J.D. Hackney (1983). Acuterespiratory effects of Los Angeles smog incontinuously exercising adults, J. Air Pollut.Contrl. Assoc. 33:1055-1060,

Baxter, P.J.. R. Ing, H. Falk, and B. Plikaytls(1983). Mount St. Helens eruptions: Theacute respiratory effects of volcanic ash Ina North American community. Arch.Environ. Health 38:138-143.

Biller, W.F. (1983). Estimated PopulationExposures to PM1o for AlternativeStandards. EPA Contract No. 68-02-3600.Prepared for Strategies and Air StandardsDivision. Office of Air Quality Planningand Standards, Research Triangle Park,N.C. Docket #A-79-29, Il-A-7.

Bouhuys, A., G.J. Beck, and J.B. Schoenberg(1978). Do present levels of air pollutionoutdoors affect respiratory health? Nature276:466-471

Carey. W.F. (1959). Atmospheric deposits inBritain: A study of dinginess. Int. J. Air Poll.2:1-26.

CASAC [Clean Air Scientific AdvisoryCommittee] (1981a). Transcript ofproceedings. EPA Environmental ResearchCenter, Research Triangle Park, N.C.Docket #A-79-29, II-G-1.

CASAC (1981b). Transcript of Proceedings.Springfield, Virginia. Docket #A-79-29, I-G-2.

DHEW [U.S. Department of Health,Education, and Welfare] (1969). Air QualityCriteria for Particulate Matter. U.S,Government Printing Office, Washington,D.C. AP-49.

DHEW (1973). Prevalence of Selected ChronloRespiratory Conditions, United States-1970. DHEW Publication No. (HRA) 74-1511. Series 10, Number 84, Rockville, MD.

Dockery D.W., J.H. Ware, B.G. Ferris, Jr., F.E,Speizer, N.R. Cook, and S.M. Herman(1982). Change in pulmonary function Inchildren associated with air pollutionepisodes. J. Air Pollut. Contri. Assoc.32:937-942.

EPA [U.S. Environmental Protection Agency](1982a). Review of the National AmbientAir Quality Standards for ParticulateMatter:. Assessment of Scientific andTechnical Information-OAQPS StaffPaper. Office of Air Quality Planning andStandards, Research Triangle Park, N.C,27711.

EPA [1982b). Air Quality Criteria forParticulate Matter and Sulfur Oxides.Environmental Criteria and AssessmentOffice, Research Triangle Park, N.C. EPA-600/8-82-029a-c.

EPA (19083]. Regulatory Impact Analysis onthe National Ambient Air QualityStandards for Particulate Matter. Office ofAir Quality Planning and Standards,Research Triangle Park, N.C.

Ferris, B.G., Jr.. H. Chen, S. Pulco, and R.L.Hi,Murphy, Jr. (1978). Chronic non-specificrespiratory disease in Berlin, NewHampshire, 1967-1973. A further follow-upstudy. Am. Rev. Respir. Dis. 113:475-485.

Ferris, B.G., Jr.. I.T.T. Higgins, M.W. HIgginsand J.M. Peters (1973). Chronic non-specificrespiratory disease in Berlin, New

10422

Federal Re~ister I Vol. 49. No. 55 I Tuesday, March 20. 1984 / Proposed Rules102

Hampshire. 1981-1957. A follow-up study.Am- Rey.Respir. Dis.107".O-122.

Frank, N.H., US. EPA. Monitoring and DataAnalysis Division (1982). Expected Annual

- Mean Form of the Annual ParticulateMatter Standard. Technical Memorandumot John Bachmann, Ambient StandardsBranch, Research Triangle Park N.CDocket #A-79-29. H-B-8.

Frank. N.H. (1983). Update on the Differencebetween Arithmetic and Geometric Meansfor TSP. Technical Memorandum to JohnBachmann, Ambient Standards Branch.Research Triangle Park. N.C. Docket #A-79-29. 11-B-20.

Friedlander, S.K. (1982). CASAC Review andClosure of the Staff Paper for ParticulateMatter. Memorandum to Anne M. Gorsuch.January 1982. Appendix E in Docket itemA-79-29, I1-A-3.

Hancock. R.P., NA. Esmen, and C.P; Furber(1976). Visual response to dustiness. J. AirPollution Control Association 26:54-57.

Holland. W.W., R.E. Waller, and A.V. Swan(1983). Letter to Lester Grant. U.S. EPA.Environmental Criteria Assessment Office.July 14,1983. Docket #A-79--29. II-D-105.

ISO [International Standards Organization](1981). Size definitions for particlesampling. Am. Ind. Hyg. Assoc. J. 42:64-68a.

Kashdan, EIL, and M.B. Ranade (1932).Workshops on the Federal ReferenceMethod or Determination of InhalableParticles (1979-1980). US. EPAEnvironmental Monitoring SystemsLaboratory. ResearchTriangle Park. EPA-600/4-82-083. NTIS #PB83-107458.

Lawther, P.T.. RE. Waler, and.M. Henderson(1970). Air pollution and exacerbations ofbronchitis. Thorax 25:525-539.

Lawther. P.J. (1982). Letter to John Bachmann.U.S. EPA, Ambient Standards Branch.February 26,1982. Docket #A-79-29, ll-D-76.

Lippmann, M. (1983). Letter from MortonLippmann. CASAC Chairman. to EPAAdministrator William D. Ruckelshaustransmitting CASAC report: ResearchNeeds Assessment for Setting NationalAmbient Air Quality Standards. December30,1983.

Lunn J.E., J. Knowelden, and A.J. Handyside(1987). Patterns of respiratory illness inSheffield infant school children. Br. J. Prey.Soc. Med. 21:7-16.

Martin, A.E., and W.H. Bradley (1960).Mortality, fog and atmosphere pollution-an investigation during the winter of 1958-1959. Mon. Bull. Minist. Health Lab. Serv.19:56-73.

Mathtech, {1983). Benefit and Net BenefitAnalysis of Alternative National AmbientAir Quality Standards for ParticulateMatter, Volumes 1-5 Prepared forEconomics Analysis Branch. Strategies andAir Standards Division, Office of AirQuality Planning and Standards, U.S. EPA,Research Triangle Park. N.C., Mathtech,Inc., Princeton, N.J. Docket # A-79-29, II-A-&

Mazumdar, S., H. Schimmel, and L Higgins(1981). Daily mortality, smoke and SO. inLondon. England 1959-1972. Proceedings ofthe Proposed SO and Particulate Standard

- Specialty Conference. Air Pollution ControlAssociation. Atlanta, Georgia.

Mazumdar, S. H. Schimmel. and LT. Higgins(1982). Relation of daily mortality to airpollution: an analysis of 14 London winters.1958/59-1971172. Arch. Environ. Health37:213-220.

Mazumdar, S.. and V. Sussman (1983).Relationships of air pollution to health:results from the Pittsburgh study. Arch.Environ. Health 38:17-24.

Morgenstern. R.D. US. EPA. Office of PolicyAnalysis (1982). Further Reanalysis ofLondon Winters. Transmittal Memorandum

-to Sheldon Meyers. U.S. EPA. Office of AirQuality Planning and Standards. ResearchTriangle Park, N.C. Docket # A-79-29. lI-B-li.

Ostro. B.D. (1983). The effects of air pollutionon work loss and mortality. J. Environ.Econ. and Mangt. (In Press).

Pace, T.G.. U.S. EPA. Monitoring and DataAnalysis Division (1983). The Use of TSPData to Estimate PM1o Concentrations.Technical Memorandum to John Bachmann.Ambient Standards Branch. ResearchTriangle Park. N.C. Docket # A-79-29. Il-B-19.

Pace, T.G., (1982). Theoretical Approximationof DCP,/D PMW PM Ratio. TechnicalMemorandum to the file. Docket # A-79-29, 11-B-10.

Padgett J. US. EPA, Strategies and AirStandards Division (1981). Letter to Dr.Sheldon Friedlander, Chairman, Clean AirScientific Advisory Committee. October 30.1981. Docket # A-79-29, 11-C-3.

Perry, GS.. H. Chai, D.W. Dickey, R.H ilones.R.A. Kinsman. C.G. Morrik. S.L Spector.and P.C. Weizer (1983). Effect of particulateair pollution on asthmatics. Am. J. PublicHealth 73:50-56.

Rodes. C.E., K.A. Rahme, and J. Purdue(1981). Particle Collection Criteria for 10Micron Samplers. U.S. EnvironmentalProtection Agency EnvironmentalMonitoring Systems Laboratory. Re searchTriangle Park. N.C. Docket # A-79-9. It-A-4.

Schwartz. JH-L, U.S. EPA. Office of theGeneral Counsel (1972). Secondary AirQualitry Standard-Setting: Considerationsof Cost. Staff Memorandum to John R.Quarles Jr., Assistant Administrator forEnforcement and General Counsel, through.-Allen G. Kirk. IL, Deputy General Counsel.September 7.1972. Docket # A-79-29, II-B-25.

Sw-iift, D.L, and D.F. Proctor (1982). Humanrespiratory deposition of particles duringoronasal breathing. Atmos Environ.16:2279-2282.

Vena, J.E. (1983). Lung cancer Incidence andair pollution In Erie County, New York.Arch. Environ. Health 38:229-230.

Ware. J., LA. Thlbodeau. F.E. Speizer. S.Colome, and B.G. Ferris. Jr. (1931).Assessment of the health effects ofatmospheric sulfur oxides and particulatematter evidence from observationalstudies. Environ. Health Persp. 41.255-276.

Young. E.F., Jr., American Iron and SteelInstitute (1982). Letter to Kathleen Bennett.EPA Assistant Administrator, Office of Air.Noise, and Radiation. and Joseph A.Cannon. EPA Associate Administrator.Office of Planning and Management.Washington. D.C. Docket * A-79-29, II-D-87.

Addendum I-Executive Summary-Review of the National Ambient AirQuality Standards for Particulate Matter.Assessment of Scientific and TechnicalInformnation-OAQPS Staff Paper (EPA.1902)EXECUTIVE SUMMARY

This paper evaluates and interpretsthe available scientific and technicalinformation that the EPA staff believesis most relevant to the review of primary(health) and secondary [welfareNational Ambient Air Quality Standards(NAAQS) for particulate matter andpresents staff recommendations onalternative approaches to revising thestandards. Review of the NAAQS is aperiodic process instituted to ensure thescientific adequacy of air qualitystandards and is required by Section 109of the 1977 Clean Air Act Amendments.The assessment in this staff paper isintended to help bridge the gap betweenthe scientific review contained in theEPA criteria document "Air Quality -Criteria for Particulate Matter andSulfur Oxides" and the judgmentsrequired of the Administrator in settingambient standards for particulatematter. The staff paper is, therefore, animportant element in the standardsreview process and provides anopportunity for public comment onproposed staff recommendations beforethey are presented to the Administrator.

Particulate matter represents a broadclass of chemically and physicallydiverse substances that exist as discreteparticles (liquid droplets or solids]ranging in size from molecular clustersof about 0.005 micrometers (pim tocoarse dusts on the order of 100 pm.Particles originate from a variety ofstationary and mobile sources and maybe emitted directly or formed in theatmosphere by transformations ofgaseous emissions such as SO. Themajor chemical and physical propertiesof particulate matter vary greatly withtime, region, meteorology and sourcecategory, complicating the assessmentof health and welfare effects as relatedto various indicators of particulatepollution. Typical particle distributionsreveal differences in origin andcomposition for fine particles (<2.5 pm]and coarse particles (>2.5 pm]. Thereference method for the currentstandards for particulate matter is the"hi volume" sampler which collectsparticulate matter of particle sizes up to25-45 pnm (so called "Total SuspendedParticulate" or TSP].

At elevated concentrations,particulate matter can adversely affecthuman health, visibility, climate.materials, economic values, personal

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules10423

Federal Register / Vol. 49, No. 55 [ Tuesday, March 20, 1984 / Proposed Rules

comfort and well-being, and vegetation.Components of particulate matter (e.g.,sulfuric acid) also contribute to aciddeposition. Typical long-term averagelevels of TSP range from 20-40 pg/m 3 inrural areas to over 150 jig/m3 in themost polluted urban industrial areas.Maximum 24-hour TSP concentrationsexceed 500 jig/rai3 . Long-term fineparticle (<2.5 jim) levels range from 2 to5 Ag/m3 in isolated arid western areasto 20-50 tig/m3 in the rural East. Thehighest annual fine particle levels, onthe order of 50 11g/m 3, occur in the mostpolluted urban industrial areas.

Primary StandardsThe staff has reviewed scientific and

technical information on the known andpotential health effects of particulatematter cited in the criteria document.The information includes studies ofrespiratory tract deposition of particles,studies of mechanisms of toxicity,effects of high exposures to variousparticulate substances in controlledhuman and animal studies,epidemiological studies, and air qualityinformation. Based on this review, thestaff derives the following conclusions.

(1) The mechanisms by which inhaledparticles may pose health risks involve(a) penetration into and deposition ofparticles in the various regions of therespiratory tract, and (b) the biologicalresponses to the deposited materials.

(2) The risks of adverse effectsassociated with deposition of ambientfine and coarse particles in the thorax(tracheobronchial and alvelor regions ofthe respiratory tract) are markedlygreater than for deposition in theextrathoracic (head) region. Maximumparticle penetration to the thoracicregions occurs during oronasal or mouthbreathing.

(3) The major effects categories ofconcern associated with high exposuresto particulate matter include: (a) Effectson respiratory mechanics andsymptoms, (b) aggraivation of existingrespiratory and cardiovascular disease,(c) effects on clearance and other hostdefense mechanisms, (d) morphologicalalterations, (a) carcinogenesis, and (f)mortality.

(4) The major subgroups of thepopulation that appear likely to be most

sensitive to the effects of particulatematter include: (a) Individuals withchronic obstructive pulmonary orcardiovascular disease, (b) individualswith influenza,*(c) asthmatics, (d) theelderly, (e) children, (f) smokers, and (g)mouth or oronasal breathers.

(5) Although controlled animal andhuman studies, and qualitativeepidemiological results can provideimportant insights into the health risksfrom particles, the most usefulconcentration-response informationcomes from a limited set of communityepidemiological studies conducted overthe last 25 years in Great Britain and theUnited States.

Based on the scientific and technicalreviews as well as policyconsiderations, the staff makes thefollows recommendations with respectto primary particulate matter standards.

(1) Despite the variability in thecomposition of ambient particles withtime and space, the available datasuggest that reductions in ambientparticulate matter in Great Britain andthe U.S. have benefited public healthand reduced the need for separatecontrol programs for many of the moreinnately toxic components of particulatematter. Elevated particulate matterexposures in current U.S. settings mostfrequently occur without concomitanthigh SO2 levels. Considering theseobservations, a separate generalparticulate matter standard remains areasonable .public health policy choice.

(2) The current TSP standard directscontrol efforts towards particles oflower risk to health because of itsinclusion of larger particles which candominate the measured massconcentration, but which are depositedonly in the extrathoracic region. A newparticle indicator representing thoseparticles capable of penetrating thethoracic regions (thoracic particles, TP)is recommended. Protection of sensitiveindividuals breathing oronasally or bymouth, sampler reliability, and theconvention recently adopted by theInternational Standards Organization(ISO) suggest that the particle size rangeinclude those particles less than anominal 10 jim (Dso). Samplerperformance criteria should be relatedto respiratory tract deposition data.

Prototype samplers meeting thesecriteria are being field tested; reliablecommercially available models mustawait test results.

(3) Both short-term (24-hour) andannual arithmetic mean standards arerecommended. The short-term standardshould be expressed in statistical formwith the decision on the allowablenumber of exceedances made Inconjunction with establishing a level forthe Standard.

(4) Selecting a level for a particulatestandard with an adequate margin ofsafety will involve a number ofuncertainties in addition to thoseinvolved in making judgments on healthrisks associated with specificsubstances such as CO or SOz.Quantitative assessments must be basedon limited epidemiological studiesconducted in times and places whorepollutant composition may have variedconsiderably from current U.S.atmospheres. Epidemiological studiesare generally subject to a number ofinherent difficulties involvingconfounding variables and somewhatlimited sensitivity. Most studies haveused British smoke (a pseudo massindicator related to small particle (<4.5pum) darkness) or TSP as particleindicators. None of the publishedstudies have used the recommended TP(<10 jim) indicator. Thus, appropriateassumptions must be used to expressavailable results in common units.

The staff assessment of short-termepidemiological data is summarized inTable 1; levels are expressed in both theoriginal and TP units. The "effectslikely" row denotes concentrationranges derived from the criteriadocument at or above which thereappears greatest certainty that effectswould occur. The data do not, however,show evidence of clear population.thrfsholds but suggest a continuum ofresponse with both the risk of effectsoccurring and the magnitude of anypotential effect decreasing withconcentration. Thus, effects may bepossible at levels below those listed Inthe "effects likely" row, but theevidence and risks at lower levels aremuch less certain.

TABLE 1.-STAFF ASSESSMENT OF SHORT-TERM EPIDEMIOLOGICAL STUDIES

Measured British smoke levels (as p g/m3 )

Ewvalent TPEffects/study Daily mortality in London I Aggravation of bronchitis 2 Corrined range ( g) lo}" s

Effects likely ........................ . .................. 500 to 1000 ..................... 250* to 500"............................ 250 to 500 .............. ........ . 350 to £A0,Effects possile .. ..... ................... .............. .._ 150" to 5 0. .. .. 25 " . . . .. ... .. . 150 to 250 ....... ...................... .. .............. 15 to 3 0.

' Deviations in daily mortality from mean levels examined in 3 studies encompassing individual London winters of 1958-59 and 14 aggregate winters from 1958-72. Eafty winlora weesdominated by h:gh smoke dominated by coal combuston emisskons and high SO. with freauent fogs.

II m I

10424

Fedbial Register / Vol. 49, No. '55 / Tuesday, March 20,'1984 / Proposed Rules

2 Exartintion of symatoms reported by brondftics in London. StU-,s o.ct d from thie xr'J-1 B!s t ft0 1970& .zConversbon assumes tmat or London snaoe contfions, BS4 TP < TSP. Pre s0 co,"r.=nZm we rmt ro- rhet ow .r .d - - cn (M. TP uwncratas some M.,rg3n of

safety ad used to est mate me lower bound in tih efecta posble za:,. The upper bcwd (TP - TSP - 83 + 1C0 pgfn 3) Ey o 'c --nZ s TP Ie'es and t used to ;roidestimates of levets w1tere effects are most ilkety.

*ndcates levels used for upper or lawer bound of range.

Based on this staff assessment, therange of 24-hour TP levels of interest are150 to 350pg/ms. Under the conditionsprevailing during the London studies, theupper end of therange represents levelsat-which effects are likely in thesensitive populations studied. Given theuncertainties in translating these resultsto U.S. conditions and the seriousness ofthe potential health effects, the upperend of the above range contains noidentifiable margin of safety and shouldnot be considered as an appropriatestandard alternative. The uncertaintiesand the nature of the potential effectsare important margin-of-safetyconsiderations. Neither the studiessummarized above nor more qualitativestudies of effects in other sensitivepopulation groups (e.g.. asthmatics.children), or effects in controlled human

or animal studies provide scientificsupport for health risks of consequencebelow 150 rg/m. These qualitative dataas well as factors such as aerosolcomposition and exposurecharacteristics should also beconsidered in evaluating margins ofsafety associated with alternativestandards in the range of150 S /M tosomething below 350 pg/m 3.

The staff assessment of importantlong-term epidemiological data issummarized in Table 2. Long-termepidemiological studies are subject toadditional confounding variables thatreduce their sensitivity and makeinterpretation more difficult. The"effects likely" levels are derived fromthe criteria document, but again, noclear population thresholds exist for alleffects indicators. Some risk of effects

are possible at lower levels, but theseare uncertain and difficult to detect inthese studies.

Based on this staff assessment therange of annualTP levels of interest are55 to 110 ILg/ma. The upper end of thisrange overlaps the somewhat uncertain"effects levels" derived from thesestudies. Due to these uncertainties, theupper end of the range (110 Ag/mq maynot include any margin of safety, andshould not be considered as anappropriate standard alternative. Thelower end (55 pg/mgJ represents a levelwhere some risk of symptomatic effectsmight remain but no detectabledifferences in plumonary function ormarked increases in respiratorydiseases are expected. Increases insymptomatic effects at the lower levelsare uncertain and small in comparisonto baseline rates.

TABLE 2.--STAFF ASSESMETr OF LoNa-TERM EPIDE.:OLOGWCAL STUDIES

Measured as levels as g, E,,e-nt TP1q3)-tnceased rospectory W13xr ~ ~ leesa~rfses. rfed lng function I fxntaed tct.p.r crjred

In chrdren ka funtr-cfS in s&s Vt

Effects lkel 2to Soo B008 ISO* >10 so tble.0Effectso ob <220 130 to 18"E tO 0 "1) 10tbl"E S5 to 1o.No 2gntfin effects noted. 602 S 130 80 to110 40 to 55.

SStudy conducted in 1963-65 in Sheffield, En(_an ES levels (uglml) uncetaln.,Su4as conducted in 1961-70 in Beffin. NH. Maor source In co=;Mmurt putp mMt Elffects el (1 n ;Lfml) based on 2 ffzvli, 2re. Effects en kMs rxrcffcn were reL-+irr'/ sraL

3Study conducted in 1973. Expoues relect 1965-73 data inAsr;o fTMedn ntr rt rs a le t u r b ce enof effects on resp=ratory syptoms.

f 4Coe based on stuated ratio of TPTSP for current (196341) US. atmosphres. l. raio rin;es betwoen ai.1 0.5 to M Thr--- rIs wL used as lowr =o4 upprcr bottaf aforestimat TP equtvasents from TSP vaues.

* Inmcates levels used for upper or lower bound of ran,.

When evaluating margins of safety foran annual standard, it is particularlyimportant to examine the results ofqualitative data from a number ofepidemiological, animal and air qualitystudies. These suggest concern foreffects not directly evaluated in thestudies listed in Table 2. Such effectsinclude damage to lung tissuescontributing to chronic respiratorydisease, cancer, and prematuremortality. The available scientific datado not suggest major risks for theseeffects categories at current ambientparticle levels in most U.S. areas.Nevertheless, the risk that both fine andcoarse particles may produce theseresponses supports the need to limit-long-term levels of TP for a variety ofaerosol compositions.

Because of different form, averagingprocedure and size range, precisecomparisions between the above rangeof TP standards and the current primary

TSP standards are not possible.' Thelower bounds, taken together, result Instandards roughly equivalent instringency to the current standards. Ingeneral, the rest of the ranges representincreasing degrees of relaxation ascompared with the current standards. Atthe lower concentrations in the ranges,much of the relaxation would resultbecause only smaller particle sizeswould be collected. Thus, a city whereexceedance of the TSP standard waslargely dominated by coarse mode dust(with substantial mass of particlesgreater than 10 pin) would be less likelyto violate a comparable TP standardthan would an area where exceedance

*By applying observed TlPITSP ratios nd otherfactors. crude comparisons can be made. Thecurrent annual TSP standard (75 pBg/ranSeoretr[cmean) Is roughly equivalent to an arithmetic meanof SO pg/r a s 7TP. The numerical value o the 24 hrTSP standard (2W pg/mr} Is roughly equivalent to140 pg/m3TP. but this does not account fordifferences between the deterministic (currentstandard) and recommended statistical form.

of the TSP standard was dominated byparticles smaller than 10 am. At higherconcentrations in the above ranges,standards would permit increased levelsfor TP as well as for larger particles.

Secondary Standards

The staff examined information in thecriteria document relevant to the reviewof the secondary standards. Categoriesof welfare effects examined includevisibility and climate, man-madematerials, vegetation, and personalcomfort and well-being. Major staffconclusions and recommendations aresummarized below.

(1)(a) Impairment of visibility by fineparticles over urban to multistateregions clearly affects public welfare.Fine particles or major constituentsthereof also are implicated in climaticeffects, materials damage, soiling, andacid deposition. Neither the currentsecondary TSP standard nor therecommended ranges of TP standards

- 10425

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

will protect visibility in an effectivemanner. The staff, therefore,recommends consideration of a fineparticle secondary standard, basedprimarily on the relatively well-definedquantitative relationships between finemass and visibility.

(b] If a fine particle standaid isselected, a seasonal (calendar quarter)averaging time could provide astatistically stable target and yetachieve most short or long-termvisibility goals. Consideration should begiven to specifying a spatial average ofthree or more monitors placed atdistances on the order of 16-50 kun.

(c) Despite the fact that the public isconcerned about visibility and is willingto pay something for clean air,quantitative bases for evaluatingvisibility goals have not beenestablished. Therefore, the level of anystandard must be based on the judgmentof the Administrator after considerationof aesthetics and transportation, as wellas non-visibility related effects. Thestaff recommends that any nationalstandards focus on welfare effectsassociated with multistate easternregional (and western urban) haze. Suchstandards would not of themselvesprotect sensitive scenic areas of theWest, but these areas are directly andindirectly addressed by other provisionsof the Clean Air Act.

(d) Empirical ranges for standards canbe derived from approximate estimatesof eastern natural background andcurrent summertime fine particle levels.The range thus derived is 8-25 Ig/m3,

seasonal and spatial average. The upperportion of the range would'tend tomaintain the status quo in the East.Current summertime visual range inmuch of the East is about 9-15 miles.Because the lower portion of the rangeapproaches natural background levels,standards set at the lower levels wouldbe, in all practicality, unattainable inmost of the eastern U.S.'Estimatedsummertime visibility under easternnatural background conditions is on theorder of 3 to 5 times greater than undercurrent conditions..

(e) Because regional fine particles inthe East appear to be influenced moststrongly by sulfates, adoption of a fineparticle standard would trigger asubstantial departure from currentapproaches to particle control strategies.The evidence suggests that multistatecontrol of regional sulfur oxideemissions might be needed to reducefine particle levels. Thus, fine particle/visibility-climate effects are linked toacid deposition, and these problemswould likely be ameliorated by similarcontrol strategies. Addressing thesewelfare effects with a common standard

or control strategy is likely to be moreefficient than establishing separatecontrol approaches for each.Appropriate scientifically based targetsand control strategies for aciddeposition are not yet available.- (2) Although poetntial effects onclimate support the consideration of afine particle standard, quantitativerelationships are not well enoughdeveloped to provide the principal basisfor selecting the level of the standard.

(3) Consideration should be given tosoiling and nuisance effects indetermining whether a secondarystandard for TP or for TSP or some otherlarge particle indicator is desirable tosupplement the primary health andsecondary fine particle standards. Theavailable data base on such effects is,however, largely qualitative. Therefore,the basis for selecting a particular levelfor a.secondary TP or TSP standard is amatter of judgment.

(4) While chemically active fine modeand hygroscopic coarse mode particleshave been qualitatively associated withmaterials damage, the available data donot clearly suggest major effects ofparticles on materials for concentrationsat or below the ranges recommended forthe primary health and secondaryvisibility standards. Therefore, asecondary standard based solely onmaterials damage is not recommended.

(5) The staff concludes that asecondary particle standard is notneeded to protect vegetation.

(6) The acid deposition issue will notbe addressed directly in the review ofthe particulate matter standards.

Addendum II-Casac Review andClosure of the Criteria Document forSulfur Oxides/Particulate MatterJanuary 29, 1982.

Subject: CASAC Review and Closure ofthe Criteria Document for SulfurOxides/Particulate Matter

From: Sheldon K. Friendlander,Chairman, Clean Air ScientificAdvisory Committee (CASAC)

To: Anne M. Gorsuch, AdministratorOn November 16,1981, the Clean Air

Scientific Advisory Committee of theScience Advisory Board completed itsthird review of the air quality criteriadocument for sulfur oxides/particulatematters (SO./PM). The Committee noteswith satisfaction the improvementsmade in the quality of the documentduring the course of previous CASACreviews on August 20-22, 1980 and July7-9, 1981. The staff of the EnvironmentalCriteria and Assessment Office, directedby Dr. Lester Grant, have provenresponsive to Committee advice as wellas to comments provided by the general

public, and deserve to be commendedfor the high quality of the document.

The purpose in writing you is tosummarize the Committee's majorconclusions to assist you in reviewingthe scientific data and associatedstudies relevant to the establishment ofrevised ambient air quality standardsfor sulfur dioxide and particulatematters as required by law. This letterfurther advises you of the Committee'sconclusion that the criteria documentfulfills the requirements set forth InSection 108 of the Clean Air Act asamended, which requires that thedocument "shall accurately reflect thelatest scientific knowledge useful Inindicating the kind and extent of allindentifiable effects on public health orwelfare" from suflur oxides andparticulates in the ambient air.

The Committee is preparing aseparate letter to you oummarizing theconclusions of its reviews of the DraftStaff Paper for Particulate Matter. Inaddition, CASAC will prepare a similarreport on the Draft Staff Paper for SulfurOxides once that document becomesavailable and its review is completed.

MAJOR SCIENTIFIC ISSUES ANDCASAC CONCLUSIONS IN THE SOX/PM CRITERIA DOCUMENTREVIEW

Chapter 1: Executive Summary

In general, the revised draft ExecutiveSummary critically synthesizes the key.points of information discussed at lengthin the individual chapters. Itsconclusions and interpretations ofscientific data, studies, and issues areconsistent with those presented in eachchapter. Relationships among Individualchapters are clearly definedredundancies that do appear arereasonable given the complexity of thesubject..

The quality of the Executive Summarywould be further improved if morespecific statements and/or tables wereadded to clarify certain Importantinterrelationships. These Include thedifferences in chemical compositionassociated with each of the severalsignificant size ranges of particulatematter; and the health effects associatedwith the respiratory tract depositionpatterns of particulate matter in theseveral size ranges and differentchemical compositions. Quantitativehealth effects information useful Indefining specific concerntrations orranges of concentrations of size-specificand/or chemical specific PM associatedwith the occurrence of health effectsshould also be highlighted. In view ofevidence that total thoracic(tracheobronchial and alveolar) particle

10426

Fiederal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

deposition-is of public health concern, it- would also be helpful to include a -

discussion of the likely equivalenciesamong British Smokeshade (BS). TotalSuspended Particles (TSP), and sizeselective particle aerometricmeasurements that would sample crindex atmospheric concentrations of

-those sized particles identified withtracheobronchial or alveolar deposition.

Chapter 2: Physical and ChemicalProperties of Sgx/PM

This chapter is well written andaddresses the important issues relevantto a criteria document. It presents agood summary of current knowledge ofthe factors affecting the physics andchemistry of sulfur dioxide and thepathways and kinetics of itstransformation into sulfuric acid. It alsoprovides a good summary of particlecharacteristics, dynamics, andhygroscopic growth.

Chapter 3: Techniques for the Collectionand Analysis of SOx/PM

The revised chapter provides anexcellent summary of the measurementof sulfur oxides and particulates.Especially important is the discussion ofthe capabilities of the variousmeasurement techniques and the profileof pollutants in the ambient air whichthese measurements yield. The chaptercorrectly notes that British Smoke (BS),Coefficient of Haze (COHS], and TotalSuspended Particulate (TSP)measurements do not adequately reflectkey physical or chemical properties ofparticulate matter in the contemporaryambient air. Precise interconversionamong units of BS, COHS, and TSP isnot possible. In the context of aparticulate standard, British Smoke isapplicable only to a "sooty" smokeaerosol. It may not be a valid healtheffects indicator for the aerosolcompositions observed in recentsummertime episodes in the UnitedStates and Europe. Thus, it is unlikelythat BS can provide a sensitive index ofhazard for today's air pollution.

Chapter 4: Sources and EmissionsBoth natural and man-made sources

emit sulfur dioxide and particulatematter into the ambient air. Given thelimitations of our ability to derivereliable estimates from both types ofsources, the criteria document presentsan adequate discussion of currentknowledge.

Chapter 5: EnvironmentalConcentrations and Exposure

This chapter is largely acceptable inits present form. Most of the commentsand suggestions which were made for

previous drafts have been effectivelyincorporated. The most importantomission from the chapter is informationrelated to chemical composition withrespect to particle size. Abundantinformation of this type is available forsulfates and some trace metals. Giventhe strong dependence of deposiltionrates and light scattering on particlesize, it might have been worthwhile torefer to this literature in Chapter 5 or todirect attention to other documentchapters (e.g., Chapter 2) where suchrelationships are discussed.Chapter 6: Atmospheric Transport.Transformation and Deposition

This chapter is concise, well-written,and effective in communicatinginformation related to the current statusof mathematical models for air pollution.The utility of various models is clearlydiscussed, and the inadequacy ofcurrent models for quantitativeextrapolation is pointed out. Topicswhich had been omitted from theprevious draft of this chapter have beenadded to other chapters withoverlapping content. The chapter is nowacceptable as written.

Chapter 7: Acidic DepositionThe Commitee has recognized the

desirability of incorporating existinginformation on acidic deposition in thepresent criteria document. Chapter 7provides an abbreviated but adequatesummary of the contribution of sulfuroxides and particulates to the formation,transport, and effects of acidicdeposition. The Committee hasconcluded that Chapter 7 is ascientifically adequate summary withthe conditional understanding that EPAis preparing a Critical AssessmentDocument for Acidic Deposition for itsreview that recognizes and incorporatesinformation on causes, effects,.and databases for all of the various pollutantsrelevant to acidic deposition. CASAChas been briefed several times byAgency officials regarding the status ofthis document. The Committee looksforward to the submission of thisintegrated assessment for its criticalreview.Chapter 8: Effects on Vegetation

In response to CASACrecommendations and public comments.this chapter on vegetation effects hasbeen greatly improved compared toearlier drafts reviewed by theCommittee. It now includes a moreconcise and interpretive criticalevaluation of those few key studiesyielding quantitative dose-effect ordose-response information of most usefor criteria development and standard-

setting purposes. It also reasonablyincludes tables in the appendices whichsummarize studies of particulates andsulfur dioxide related vegetation effectsthat are of less utility for criteriadevelopment and standard setting.

The Committee concurs with Chapter8 evaluations which point to the lack ofdose-response data to establishquantitative evidence of deleteriouseffects on vegetation from particulatesat presently encountered U.S. ambientair concentrations. In contrast toparticulates, much clearer evidenceexists by which to define quantitativeexposure-effect relationships for sulfurdioxide effects on vegetation.Laboratory experiments in particularhave demonstrated the greater relativetoxicity to vegetation from high short-term exposures of sulfur dioxide. This isespecially important in view of the factthat ambient air concentrations of sulfurdioxide from point sources oftenfluctuate widely and result in highintermittent short-term exposures ofplants to sulfur dioxide concentrationsagainst a background of longer-term butmuch lower annual average sulfurdioxide levels. Also of much importanceare differences in the relative sensitivityof various plant species to sulfur dioxideexposures. The degree of sensitivitydepends in part on factors such as phaseof growth at time of exposure. ambienttemperature and humidity levels, andplant water content. Among studiesjudged to be most useful for quantitativecriteria development and standardsetting are those of Dreisinger (1965,1967) and Dreisinger and McGovern(1970) which demonstrate visible injuryto white pine (a commercially importantspecies in some U.S. areas) whennatural stands of the tree in southernCanada were exposed for 4.hours to 0.30ppm or for 8 hours to 0.25 ppm sulfurdioxide emitted from a nearby smelter.Roughly similar exposure-effectrelationships were observed in studiesreported by Jones et al. (1974) andMcLaughlin (1981) on the effects ofsulfur dioxide from a southeastern U.S.power plant on a wide variety of naturalspecies in the vicinity of the pointsource. In these studies some crop andgarden species showed visible injuryeffects with 3 hours exposures to 0.6--0.8ppm sulfur dioxide, while certain othercrop species (potato, cotton, corn.peach) did not show visible injury atlevels below 0.8 ppm. In contrast, achamber study by Hill et al. (1974)suggests that plants common to thesouthwestern U.S., with markedly lowermoisture content and under generallylower ambient air humidity levels, maybe able to withstand much higher

10427

Federal Register / Vol. 49, No. 55 / Tuesday, March 20. 1984 / Proposed Rules

ambient sulfur dioxide concentrations(up to 11 ppm for two hours) withoutvisible injury.Chapter 9. Effects on Visibility andClimate

The technical aspects of this difficultproblem are well characterized. Thechapter does a good job of discussingthe physics and public awareness ofvisibility. The relationship between fineparticle mass concentrations andvisibility has been well established. Thecriteria document thus provfdes anexcellent technical basis for Agencydecision-making on these issues.

This chapter adequately discusses thecurrently available scientificinformation concerning the effects ofOarticulate matter and sulfur oxides onman-made materials. This includescritical assessments of available dataconcerning pertinent materials damagefunctions, uncertainties associated withexisting characterizations of suchfunctions, and limitations regardingestimation of monetary costs and/orbenefits associated with the occurrenceor control of such damage.Chapter 11: Respiratory Deposition andBiological Fate of Inhaled Aerosols andSulfur Dioxide

This chapter is very much improvedcompared to earlier drafts reviewed byCASAC and is now a comprehensiveand more informative summary ofexisting knowledge relevant to a criteriadocument. The existing knowledge inthis area is, in many cases, incomplete.For example, a potentially veryimportant factor is the influence of theintegrity of lung epithelial barriers (bothairway and alveolar) on deposition andclearance. To enhance the chapter'scomprehensiveness, this issue should bediscussed more sufficiently in thecriteria document, despite the paucity ofavailable data.Chapter 12: Toxicological Studies

This chapter is quite comprehensiveas it describes essentially alltoxicological studies relevant to acriteria document on sulfur oxides andparticulates. Also, it providescommentary on many studies and thesignificance of their findings to potentialhuman health effects. In addition, thepresentation of the information is morepolished than the previous draft becauseof improved editing.Chapter 13: Controlled Human Studies

This is a chapter which thoroughlydiscusses the published material oncontrolled human experiments. The"scientific criteria for good studiesdiscussed at the beginning of the

chapter cannot be overemphasized.While not all studies meet these criteria,the Committee recognizes that EPA musttake account of the available literatureand believes the studies cited in thechapter have been appropriatelyselected and discussed. Overall thechapter i well-written and directedtoward addressing those questions towhich answers are needed. One of themost important criteria for good humanclinical studies is that they be double-blind. Unfortunately, most of the studiesin the literature were not so performed.This factor is especially significantwhen sensitive population groups, suchas asthmatics, are under study.

The chapter is also improved by thediscussion of exposures administeredthrough the nose and mouth duringcontrolled studies. It appropriately notes.that caution should be used in any

• attempted extrapolation of observedquantitative exposure/effects resultingfrom such protocols, particularly whencompared to results that might beexpected under ambient exposureconditions. The chapter identifiesadditional research results from studiesusing either face mask or open chamberoronasal breathing that would betterresolve this issue, and it discussesexisting studies in a balanced andthorough fashion.Chapter 14: Epidemiological Studies

The current draft of this chapterrepresents considerable change andimprovement over previous draftsreviewed by CASAC. Followingdiscussion with the Committee, EPA hasapplied a set of guidelines for 'decidingwith epidemiological studies are mostappropriate for use in revising ambientair quality standards.

More specific comments on thechapter include the following: (1) Theintegration of Chapter 14 with Chapter 3has advanced the "real world"understanding concerning theapplication of epidemiological methods;(2] the epidemiological studies providingthe most useful quantitativeconcentration/response information forrevising the 24-hour ambient particulatestandard include: Lawther et al, 1958and 1970; Martin and Bradley 1960;Martin 1964; Ware et al, 1981; andMazumdar el al, 1981; (3] theepidemiological studies providing themost useful quantitative concentration/response information for revising theannual ambient particulate standardinclude: Ferris and Anderson 1962; Lunnet al, 1967; Ferris et al, 1971 and 1976;and Bouhuys et al, 1978; and (4) thestudies by Lave and Seskin, 1970, andMendelsohn and Orcutt, 1979 suggest anassociation between chronic exposure

to high concentrations of sulfates andincreases in the level of mortality, butthey do not indicate any threshold orsafe level from such exposures, and theyare not refined enough to provideestimates of the quantitative effect ofsulfate concentrations on mortality.

SUMMARYThe Committee made numerous

comments of an editorial nature. Theseremarks, as well as a more detaileddiscussion of the recommendations andreview provided above, are included inthe transcripts of the three CASACmeetings held to review this document,With the understanding that the advisedchanges will be incorporated in the finalcriteria document, the Committee issatisfied that the air quality criteriadocument for sulfur oxides/particulatematter is scientifically adequate for usein standard setting.

Addendum 111-CASAC Review andClosure of the OAQPS Staff Paper forParticulate MatterJanuary 29, 1982.

Subject: CASAC Review and Closure ofthe OAQPS Staff Paper forParticulate Matter

From: Sheldon K. Friedlander,Chairman, Clean Air ScientificAdvisory Committee

To: Anne M. Gorsuch, AdministratorThe Clean Air Scientific Advisory

Committee (CASAC) recently completedits second and final review of thedocument entitled Review of theNational Ambient Air QualityStandards for Particulate Matter:Assessment of Scientific and TechnicalInformation, OAQPS Staff Paper. TheCommittee notes with satisfaction theimprovements made in the scientificquality and the completeness of the staffpaper. It has been modified inaccordance with the recommendationsma' de by CASAC in July and November1981. This document is also consistent inall significant respects with thescientific evidence presented andinterpreted in the combined criteriadocument for sulfur oxides andparticulate matter. It has organized thedata relevant to the establishment ofparticulate primary and secondaryambient air quality standards in alogical and compelling way, and theCommittee believes that it provides youwith the kind and amount of technicalguidance that will be needed to makeappropriate revisions to the standards.

CASAC has prepared this closurememorandum to inform you morespecifically of its major findings andconclusions concerning the variousscientific issues and studies discussed in

10428

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

the staff paper. In addition, theCommittee's review of the scientificevidence leading to the particulatestandard revision leads to a discussionof its own role in the process for settingthe standard.

CASAC Conclusions andRecommendations on Major ScientificIssues and Studies Associated With theDevelopment of-Revised NAAQS forParticulates

1. Based upon the review of availablescientific evidence, a separate generalparticulate standard remains areasonable public health policy choice.

2. CASAC reaffirms its initialrecommendation on July 1981 toestablish a 10 micrometer cut point for arevised primary particulate standard.This recommendation is based upon arecognition of the periodic, andsometimes frequent, tendency of bothhealthy and Bensitive populations tobreathe through their mouths and/ororonasally: This practice increases theamount of particulate matter that canpenetrate into the thorax because thelarger particles are not filtered in theoronasal passages. Deposition ofparticulates into this region is of specialconcern to those individuals with pre-existing respiratory problems andchildren. In addition, the collection ofparticles of less than 10 micrometerdiameter size more closely resemblesparticles passing into the thoracticregion of the human body than thecollection of larger sized particles.Furthermore, monitors equipped for a 10micrometer cut are less wind dependentand can provide a more accurate profileof the contemporary ambient air thansamplers which measure totalsuspended particles.

CASAC's recommended size cut isalso similar to proposals of otherscientific associations. For example, 88%of the national members of the AirQuality Committee of the InternationalStandards Organization recently votedfor a particulate cut point at 10micrometers for sampling particleswhich can deposit in the lungs.

The CASAC recommendation is basedupon available scientific data. Otherindividuals and groups have discussedthe possibility of establishing a revisedparticulate standard at a size cutconsiderably less than 10 micrometers.However, for the current revision of thestandard, the scientific data morereadily support a 10 micrometer size cut.

3. CASAC reached several majorconclusions concerning the revision ofthe 24-hour and annual phrticulatestandards. At the upper bound of theproposed ranges of 150-350 jpg/m3 forthe 24-hour and 55-110 pg/m 3for the

annual averages, detectable healtheffects occur in the populationsevaluated in the epidemiological studies.Since the upper end of these rangescontain little or no margin of safety, itwould be appropriate to consider lowervalues for revising the 24-hour andannual standards. In addition, the statedranges are based solely on quantitativeevidence reported in epidemiologicalstudies. A final decision on a revisedstandard should also incorporateinformation generated throughcontrolled human, animal toxicology.and from other less quantitativeepidemiological studies discussed in thecriteria document.

There is an absence of a clearlydefinable exposure-responserelationship for particles, as amplydiscussed in the criteria document andthe staff paper. In addition, becauseairborne particles are heterogeneous incomposition, the potential toxic effectsof individual constituents should beconsidered in setting the standard. Thus.compared to margins of safety set forpollutants such as ozone and carbonmonoxide, where exposure-responserelationships are better established andsmall margins of safety are morejustifiable, CASAC believes you shouldconsider a revised standard with awider margin of safety.

4. The Committee reached generalagreement that the annual particulatestandard should consist of an arithmeticmean. It is recommended that the 24-hour standard include a statistical formand that the number of exceedances isset in relation to the revised standardlevel.

5. During the past decade, the linkbetween visibility and rme particle massconcentrations has been convincinglydocumented. Visibility is a sensitiveindicator of accumulated man-madepollutants in the ambient air. The publiccares about visibility and is willing topay something for clean air. However,the quantitative basis for establishing apsychological, economic, transportationor any other welfare cost associatedwith visibility impairment has not beenestablished. In addition, cofitrolsrequired to.achieve a given visibilitystandard are not known due to thecomplexities of pollutant transport andtransformation..

Defining acceptable levels of visibilityis a social/policy judgment as well as ascientific decision, but science canprovide some guidance. The upper endof the 8-25 jIg/ M 3 range for fine particles(those particles with a diameter size ofless than 2.5 micrometers) would tend tomaintain the status quo for the easternUnited States and some western urbanareas, but would permit air quality

degradation for large areas in the westincluding national parks. Also, it ishighly uncertain that the recommendedthoracic particle ranges for the primarystandard will protect visibility. The 8-25jig/m3 range for fine particles suggestedfor visibility protection is a seasonaland spatial average, unlike peak valueswhich will be recommended for theprimary standard.

The strongest case for a visibilityreldted standard is one that linksemissions of nitrogen oxides and sulfurdioxide with the interrelated aspects ofacidic deposition, possibleclimatological effects, and visibility.Each of these three air quality issues isrelated to the fine particles whichoriginate both as primary particulateemissions and as secondary aerosolsfrom atmospheric conversions of sulfurdioxide and nitrogen oxides'emitted asvapors. In terms of a control strategy toprotect public welfare, it may be moreefficient to consider a common standardlinked to fine particles than to establisha separate set of controls for each ofthese problems and pollutants.

6. The Committee's evaluation ofscientific data and studies in the criteriadocument and the staff paper lead it toconclude that there is no scientificjustification for the establishment of aparticulate standard for the specificprotection of vegetation.

7. The Committee discussed whateffect elimination of a Total SuspendedParticulate (TSP) standard would haveoha the environment. The soiling andnuisance aspects of TSP are essentiallylocal air quality problems because suchcoarse particles are not transportedgreat distances. This contrasts withvisibility or oxidant related problemswhich are distinctly issues of long rangepollution transport. Individuals whoserve on the Committee made variousrecommendations regarding retention orelimination of a secondary standard forTSP, but no clear consensus evolved.

The Process for Setling the AmbientParticulate Standard

In its report of September 21,1981,CASAC made several majorrecomxfiendations relating to the processfor setting ambient air standards. TheCommittee is aware that your staff isanalyzing its report and is awaiting aresponse.

A major underlying assumption of theCommittee's recommendations was theneed to make more explicit therelationship between the scientificevidence in the criteria do.ument andthe staff paper and the eventualselection of a numerical level forindividual standards. The Committee

104.29

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

strongly believes in the need to clarifythe standard setting process byidentifying the key studies that willshape the determination of a standard.Intensive evaluation of such studies byCASAC and the public will considerablyincrease your ability to set ascientifically supportable standard.

The Committee is greatly encouragedby your decision to improve the formatand content of OAQPS scientific issuestaff papers. In the Draft Staff Paper forParticulate Matter key studies areidentified and their implications forsetting primary and secondarystandards are discussed. Moreimportantly, the inclusion of numericalranges and their supporting rationaleenabled the Committee and the public tocritically examine the staff's proposeduse of the studies. This led to a markedimprovement in the quality of the publicdialogue concerning the scientific basisfor revising the standard. CASACcommends your effort and recommendsthat all staff papers developed forambient air standards contain numericalranges.

CASAC recognizes that your statutoryresponsibility to set standards requirespublic health policy judgments inaddition to determinations of a strictlyscientific nature. While the Committee iswilling to further advise you on theparticulate standard, we see no need, inview of the already extensive commentsprovided, to review the proposedparticulate standards prior to theirpublication in the Federal Register. Inthis instance, the public comment periodwill provide sufficient opportunity forthe Committee to provide any additionalcomment or review that may benecessary.

PART 50-NATIONAL PRIMARY ANDSECONDARY AMBIENT AIR QUALITYSTANDARDS

For reasons set forth in the preamble,EPA proposes to amend Part 50, Chapter1 of Title 40 of the Code of FederalRegulations as follows:

1. The table of contents for Part 50 isamended by adding new entries forAppendix J and Appendix K as follows:

Appendix I-Reference Method for theDetermination of Particulate Matter asPMio in the Atmosphere

Appendix K-Interpretation of the NationalAmbient Air Quality Standards forParticulate Matter

2. Section 50.6 is revised to read asfollows:

§ 50.6 National primary ambient air qualitystandards for particulate matter.

(a) The level of the national primary24-hour ambient air quality standard forparticulate matter is [value to beselected from range of 150 to 2501micrograms per cubic meter (jzg/mg), 24-hour average concentration. Thestandard is attained when the expectednumber of days per calendar year with a24-hour average concentration above thestandard level is equal to or less thanone, as determined by Appendix K.

(b) The level of the national primaryannual standard for particulate matter is[value to be selected from range of 50 to65] microgramb per cubic meter (pg/mn,annual arithmetic mean. The standard isattained when the expected annualarithmetic mean concentration is lessthan or equal to the standard level, asdetermined by Appendix K.

(c) For purposes of the primarystandards, particulate matter shall bemeasured in the ambient air as PMo(particles with an aerodynamic diameterless than or equal to a nominal 10micrometers] by:

(1] A reference method based onAppendix J and designated inaccordance with Part 53 of this chapter,or

(2) An equivalent method designatedin accordance with Part 53 of thischapter.

3. Section 50.7 is revised to read asfollows:

§50.7 National secondary ambient airquality standard for particulate matter.

(a) The level of the nationalsecondary standard for particulatematter is [value to be selected fromrange of 70 to 90] micrograms per cubicmeter (pg/m1, annual arithmetic mean.The standard is attained when theexpected annual arithmetic meanconcentration is less than or equal to thestandard level, as determined byAppendix K.

(b) For purposes of the secondarystandard, particulate matter shall bemeasured in the ambient air by themethod described in Appendix B of Part50 of this Chapter.

4. In Appendix G, reference 10 isremoved and section 5.1.1 is revised toread as follows:

5.1.1 High-Volume Sampler. Use andcalibrate the sampler as described inAppendix-B to this part.

5. Appendix J is added as follows:

Appendix J-Reference Method for theDetermination of Particulate Matter asPM,, In The Atmosphere

1.0 Applicability.

This method provides for the measurementof the mass concentration of particulatematter with an aerodynamic diameter lessthan or equal to a nominal 10 micrometers(PMo} in ambient air over a 24-hour periodfor purposes of determining attainment andmaintenance of the primary national ambientair quality standards for particulate matterspeacified in § 50.0 of this chapter. Themeasurement process is nondestrucative, andthe PMio sample can be subjected tosubsequent physical and chemical analyses.Quality assurance procedures and guidanceare provided in Part 58, Appendices A and B,of this chapter and in References (1) and (2).

2.0 Principle.2.1 An air sampler draws a measured

quanity of ambient air at a constant flow rateinto a specially shaped inlet where thesuspensed particulate matter Is Inertliallyseparated into one or more size fractionswithin the PMio size range. Each size fractionin the PMo size range is then collected on aseparated filter over the specified samplingperiod. The particle size discriminationcharacteristics (sampling effectiveness and s0percent cutpoint) of the sampler Inlet over thePMo size range are functional specificationsdescribed in Part 53 of this chapter.

2.2 Each filer is weighed (after moisture* equilibration) before and after use to

determine the net weight (mass] gain due tocollected PMo. The total volume of airsampled, corrected to EPA referenceconditions (25°C, 101.3 kPa), Is determinedfrom the measured volumetric flow rate andthe sampling time. The concentration of PMioin the ambient air is computed as the totalmass of collected particles in the PMio sizerange divided by the volume of air sampled,corrected to reference conditions, and Isexpressed in micrograms per standard cubicmeter ({g/std m]. For samples collected attemperatures and pressures significantlydifferent from EPA reference conditions,these corrected concentrations sometimesdiffer substantihlly from actualconcentrations (in micrograms per actualcubic meter), particularly at high elevations.Although not required, the actual PMoconcentration can be calculated from thecorrected concentration, using the actualtemperature and pressure during thesampling period.

2.3 A method based on this principle willbe considered a reference method only if (a)the-associated sampler meets therequirements specified in this appendix andthose in Part 53 of this chapter and, (b) themethod has been designated as a referencemethod in accordance with Part 53 of thischapter.

3.0 Range.The lower limit of the mass concentration

range is limited by the repeatability of filtertare weights, assuming the nominal airsample volume for the sampler. The upperlimit of the concentration range cannot bespecified. For samplers having a filter-changing mechanism, there may be no upperlimit. For samplers that do not have a filter-changing mechanism, the upper limit isdetermined by the point at which the samplerno longer maintains the specified operatingflow rate due to increased pressure drop

bt10430

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

across the loaded filter(s). This limit cannotbe specified because it is a complex andundeterminedLunction of particle sizedistribution and type, humidity, filter type.and perhaps other factors.

4.0 Precision.The reproducibility of PMo samplers must

be 15 percent or better as required by Part 53of this chapter, which prescribes areproducibility test procedure thatdetermines the variation in the PMoconcentration measurements of identicalsamplers under typical sampling conditions.Other specifications are provided in Part 53for the particle size discriminationcharacteristics and the flow rate stability ofthe sampler. Continual assessment of theprecision via collocated samplers is requiredby Part 58 of this chapter for PMho samplersused in certain monitoring netvorks.

5.0 Accuracy.Because the sizes of the particles making

up ambient particulate matter vary over awide range and the concentration of particlesvaries with particle size, it is difficult todefine the absolute accuraacy of PMsamplers. Part 53 of this chapter provides aspecification for the sampling effectiveness ofPM,9 samplers. This sampling effectivenessspecification requires that the expected massconcentration measurement calculated for acandidate PMo sampler, when sampling aspecified typical ambiefit particledistribution, be within J±10 percent of thatcalculated for an ideal sampler-whosesampling effectiveness is explicitly specified.Also, the particle size for 50 perent samplingeffectiveness is required to by 10±1micrometers /jin. Other specificationsrelated toaccuracy apply to flowmeasurement and calibration, filter media.analytical (weighing) procedures, loss ofvolatiles, and artifact and nonsampledparticulate matter. The flow rate accuracy ofPMo samplers used in certain monitoring

- networks is required by Part 58 of thischapter to be assessed periodically via flowrate audits. -

6.0 Potential Sources of Error.6.1 Loss of Volatile Particles. Volatile

particles collected on filters are often lostduring shipment and/or storage of the filtersprior to the postsampling weighing (3).Although.shipment and storage of loadedfilters are sometimes unavoidable, filtersshould be reweighed as soon as practical tominimize these losses.

6.2 Artifact Particulate Matter. Positiveerrors in particle mass measurements mayresult from retention of gaseous species onfilters (4, 5). Such errors include the retentionof sulfur dioxide and nitric acid. Retention ofsulfur dioxide on filters, followed byoxidation to sulfate is referred to as artifactparticulate sulfate formation, a phenomenonwhich increases with increasing filteralkalinity (6). Artifact particulate nitrate,resulting primarily from retention of nitricacid, occurs to varying degrees on many filtertypes, including glass fiber. cellulose ester,and many quartz fiber filters. (5. 7. 8, 9. 10).Filters that meet the alkalinity specification(section 7.2.4) should show little or no Artifactsulfate. Negative artifact is the loss ofcollected particulate matter during samplingby volatilization or-chemical reaction (11).

Loss of true atmospheric particulate nitratehas been observed on Teflon5 filters (8) andinferred for quartz fiber filters (11). Thesignificance of this problem for PMs massmeasurements will vary with location andambient temperature. However. for mostsampling locations. PMzo mass concentrationerrors due to nitrate artifact are expected tobe small.

6.3 NonsampledParliculate AMalter.Particulate matter is sometimes deposited onfilters during periods when the sampler isinoperative (12). Timely installation andretrieval of filters prior to and following thesampling period should help to minimize thisproblem.

6.4 Humidily. The effects of ambienthumidity on the sample are unavoidable. Themoisture conditioning procedure In section9.0 is designed to minimize the effect, ofmoisture on the filter medium.

6.5 FilterHandli&. Careful handlLng offilters betwen presampling and posteamplingweighing is necessary to avoid errcza due todamaged filters orloss of particles from thefilters. Use of a filter cartridge or cassettemay reduce the magnitude of these errors.

6.6 Flow Rate Variation Variations in thesampler's operating flow rate may alter theparticle size discrimination characteristizs ofthe sampler inleL The manitude of this errorwill depend on the sensitivity of the inlet tovariations in flow rate and on the particledistribution in the atmosphere during thesampling period. The use of an automaticflow controller (section 7.1.4) is required tominimize this error.

6.7 Air Volume Deiermiatid: -rrors inthe air volume determination can result fromerrors in the flow rate and/or sampling timemeasurements. The automatic flow controlleralso serves to minimize errors in the overageflow rate determination. The use of anelapsed time meter (sect:on 7.1.5) is requiredto minimize the error in sampling time.

7.0 Apparatus.7.1 Pfo Sampler.7.1.1 The sampler shall be designed to:a. Draw the air sample, via reduced

internal pressure, into the sampler inlet andthrough the filter~s) at a uniform facevelocity.

b. Hold and seal the filter(s) in a horizontalposition so that sample air is dravmdownward through the filter(s).

c. Allow the filters) to be installed andremoved conveniently.

d. Protect the filter(s) and sompler fromprecipitation andprevent insects and otherdebris from being sampled.

e. Minimize leaks that would'cause error Inthe measurement of the air volume passingthrough the filter(s).

f Discharge exhaust air at a sufficientdistance from the sampler inlet to minimizethe sampling of exhaust air.

g. Minimize the collection of dust from thesupporting surface.

7.1.2 The sampler shall operate at acontrolled flow rate specified by its designeror manufacturer, and it shall have an inletsystem that provides particle sizediscrimination characteristics meeting all ofthe applicable performance specificationsprescribed in Part 53 of this chapter. Thesampler inlet shall show no significant wind

direction depend-once. This requirement cangenerally be satisfied by an inlet shape thatis circularly symmetrical about a verticalaxis.

7.1.3 The sampler shall provide a meansto measure the total flow rate during thesampling period.A continuous flow recorderis recommended. The sampler may beequipped with additional flow measurementdevices if it is designed to collect more thanone particle size fraction.

7.1.4 The sampler shall have an automaticflow control device capable of adjusting andmaintaining the sample flow rate within thelimits specified for the sampler inlet overnormal variations in line voltage and filterpressure drop. A convenient means must beprovided to temporarily disable theautomatic flow control device to allowcalibration of the samplers flowmeasurement device.

7.1.5 A timing/control device capable of -starting and stoppingthe sampler shall beused to obtain an elapsed run-time of 24 ±1hr (1.440 .0 min). An elapsed time meter.accurate to within 15 minutes. shall be usedto measure sampling time. This meter isoptional for samplers ,ith continuous flowrecorders if the sampling time measurementobtained by means of the recorder meets the4_15 minute accuracy specification.

7.1.6 The sampler shall have anassociated operation or instruction manual asrequired by § 53A of this chapter and whichincludes either the text or a reproduction ofthis appendix.

7.2 FMters7.2.1 FilterAedium. &o commercially

available filter medium is ideal in all respectsfor all samplers. The useWs goals in samplingdetermine the relative importance of variousfilter evaluation criteria (e.g, cost ease ofhandling, physical and chemicalcharacteristics. etc.) and, consequently.determine the choice among acceptablefilters. Furthermore. certain types of filtersmay not be suitable for use with somesamplers, particularly under heavy loadingconditions (high mass concentrations).because of high or rapid increase in the filterflow resistance that would exceed thecapability of the samplers automatic flowcontroller. However. samplers equipped withautomatic filter-changing mechanisms mayallow use of these types of filters. Thespecifications given below are minimumrequirepients to insure acceptability of thefilter medium for measurement of PMto massconcentrations. Other filter evaluationcriteria should be considered to meetindividual sampling and analysis objectives.

72.2 Collection Efficiency. >99 percent asmeasured by DOP test (ASTN--2986) with 0.3pn particles at the sampler's operating facevelocity.

723 Integrity. ±5 pg/ml (assumingsampler's nominal 24-hour air samplevolume), measured as the concentrationequivalent corresponding to the differencebetween the initial and final weights of thefilter when weighed and handled undersimulated sampling conditions (equilibration.initial weighing. placement on inoperativesampler, removal from sampler. re-equilibration, and final weighting).

10431

Federal Register J Vol. 49, No. 55 f Tuesday, March 20, 1984 / Proposed Rules

7.2.4 Alkalinity. <0.005 milliequivalents/gram of filter as measured by ASTM-D202following at least two months storage atambient temperature and relative humidity.

7.3 Flow Rate Transfer Standard7.3.1 A flow rate transfer standard,

suitable for the flow raie of the sampler andcalibrated against a primary standard that istraceable to NBS, must be used to calibratethe sampler's flow measurement device.

7.3.2 The reproducibility and resolution ofthe transfer standard must be 2 percent orless of the sampler's operating flow rate.

7.3.3 The flow rate transfer standard mustinclude a means to vary the sampler flow rateduring calibration of the sampler's flowmeasurement device.

7.4 Filter Conditioning Environment7.4.1 Temperature range: 15 to 30° C.7.4.2 Temperature control: ±3 y C.7.4.3 Humidity: 45 ±5 percent relative

humidity.7.5 Analytical Balance7.5.1 The analytical balance must be

suitable for weighting the type and size offilters required by the sampler. The range andsensitivity required will depend on the filtertare weight and mass loading. Typically, ananalytical balance with a sensitivity of 0.1 mgis required for high volume samplers (flowrates >0.5 m3/min). Lower volume samplers(flow rates <0.5 m3/min) will require a moresensitive balance.

8.0 CALIBRATION.8.1 General Requirements8.1.1 Calibration of the sampler's flow

measurement device is required to establishtraceability of the flow measurement to aprimary standard. A flow rate transferstandard calibrated against a primary flow orvolume standard shall be used to calibratethe sampler's flow measurement device at thefield site.

8.1.2 The particle size separationcharacteristics of PMo samplers usuallyrequire that specific air velocities bemaintained in the separation system.Therefore, the sampler must be set to operateat and maintain the specified volumetric flowrate, measured under the actual ambientconditions of use ( 0 ). In contrast, the massconcentration of PMo must be computedusing the flow rate based on the standardvolume at EPA reference conditions (Q,).

8.2 Flow Rate Calibration Procedures8.2,1 The calibration procedure given here

is based on flow rates at ambient conditions(QJ and serves to illustrate the stepsinvolved in the calibration process.Alternative procedures based on othermeasures of flow rate (e.g., Qtd1-may be usedprovided the requirements of section 8.1 aremet. Consult the sampler manufacturer'sinstruction manual for specific guidance oncalibration. Reference (13) providesadditional information on the use of thecommonly used measures of flow rate andtheir interrelationships.

8.2.2 Calibrate the flow rate tranferstandard against a primary flow or volumestandard traceable to NBS. Establish acalibration relationship (e.g., and equation orfamily of curves) such that traceability to theprimary standard is accurate over theexpected range of ambient conditions (i.e.,temperatures and pressures) under which the

transfer standard will be used. Recalibratethe transfer standard periodically (minimumof once per year).

8.2.3 Disable the sampler's flow controllerduring calibration of the sampler's flowmeasurement device.

8.2.4 Install a clean filter (or filters] in thesampler. Remov e the sampler inlet andconnect the transfer standard to the samplersuch that the transfer standard accuratelymeasures the sampler's flow rate. Make surethere are no leaks between the transferstandard and the sampler.

8.2.5 Choose three flow rates evenlyspaced over a range of ± 10 percent of thesampler's specified operating flow rate(actual m3/min), and by suitable adjustmentof the sampler flow rate, obtain a calibrationcurve of flow'rate (actual m3/min) versus thesampler's flow indicator reading. Record thebarometric pressure and ambienttemperature. Daily or seasonal temperatureand daily or average pressure corrections forsubsequent flow indicator readings may berequired for certain types of flowmeasurement devices (see Note followingstep 9.8).

8.2.6 Re-enable the flow controller, adjustthe flow rate (actual m3/min] to themanufacturer's specified operating set point,and use the transfer standard to verify thatthe flow rate is correct with a clean filter (orfilters) in place.

8.2.7 Replace the sampler inlet.9.0 Procedure.9.1 The sampler shall be operated in

accordance with the general instructionsgiven here and with the specific instructionsprovided in the sampler manufacturer'sinstruction manual. Note.-This procedureassumes that the sampler's flow ratecalibration was performed using flow rates atambient conditions (Q.).

9.2 , Inspect each filter for pinholes,particles, and other imperfections; establish afilter information record and assign anidentification number to each filter.

9.3 Equilibrate each filter in theconditioning environment for at least 24-hours.

9.4 Following equilibration, weigh eachfilter, and record the presampling weight withthe filter identification number.

9.5 Install a preweighed filter (or filters)in the sampler following the instructionsprovided in the sampler manufacturer'sinstructional manual.

9.6 Turn on the sampler and adjust (ifnecessary) the automatic flow controller tothe manufacturer's specified operating setpoint. Run the sampler for at least 5 minutesto establish run-temperature conditions.Record the flow indicator reading and, ifneeded, the barometric pressure and ambienttemperature. Determine the sampler flow rate(in actual m3/min) using the sampler's flowrate calibration curve.

Note. -No onsite pressure of temperaturemeasurements are necessary if the samplerflow indicator does not require pressure or'temperature corrections or if averagebarometric pressure and seasonal averagetemperature for the site are incorporated intothe sampler calibration (see step 8.2.5). Forindividual pressure and temperaturecorrections, the ambient pressure and

temperature can be obtained by onsitemeasurements or from a nearby weatherstation. Barometric pressure readingsobtained from airports must be stationpressure, not corrected to sea level, and mayneed to be corrected for differences Inelevation between the sampler site and theairport.

9.7 If the sampler flow rate (actual m'/min) is outside the acceptable range specifiedby the sampler manufacturer, check thesampler for leaks and, If necessary, adjust theautomatic flow controller set point. Stop thesampler.

9.8 For samplers without continuous flowrecorders, record the initial flow rate (inactual ml/min) as Q(init.).

9.9 Set the timer to start and stop thesampler at appropriate times. Set the elapsedtime meter to zero.

9.10 Record the sample information (filteridentification number(s), site location oridentification number, sample date, andstarting time).

9.11 Sample for 24 ±1 hours.9.12 For samplers without continuous

flow recorders, as soon as practical followingthe sampling period, run the sampler for 5minutes to again establish run-temperatureconditions. Record the flow indicator readingand, it needed, the barometric pressure andambient temperature. Stop the sampler.Determine the final flow rate (In actual m3/min) using the sampler's flow rate calibrationcurve and record as Q.(, o (see Notefollowing step 9.6). If Q, 0 .,, is outside thesampler manufacturer's specified operatingrange, the sample must be invalidated. rorvalid samples, calculate the average flow rate(in actual m3/min), and recoid as Q.

9.13 For samplers with continuous flowrecorders, examine the flow record. If Q, Isoutside the sampler manufacturer's specifiedoperating range for more than 0 hours of the24-hour sampling period, the sample must beinvalidated. For valid samples, record theaverage flow recorder reading during thesampling period. If needed, estimate theaverage temperature and pressure at the siteduring the sampling period from weatherbureau or other available data. Determine theaverage flow rate (in actual r33/min using thesampler's flow rate calibration curve andrecord as Q. (see NOTE following step 9.0).

9.14 Carefully remove the filter (or filters)from the sampler following the samplermanufacturer's instructions. Touch only theouter edges of the filter.

9.15 Place the filter(s) In a protectiveholder or container (e.g., petri dish, glassineenvelope, or manila folder).

9.16 Record the elapsed time on the filterinformation record and any other factors.such as meteorological conditions,construction activity, fires or dust storms,etc., that might be pertinent to themeasurement. If the sample is known to bedefective, void it at this time.

9.17 Transport the exposed sample filter(or filters) to the filter conditioningenvironment as soon as possible forequilibration and subsequent weighing,

9.18 Equilibrate the exposed filter(s) inthe conditioning environment for 24-hours.

l

10432

Federal Register / Vol. 49, No. 55 1 Tuesday, March 20, 1984 / Proposed Rules

9.19 Immediately after equilibration.reweigh the filter(s) and record the weight(s)with the filter i'dentification number(s).

10.0 Calculations.10.1 Calculate the average flow rate over

the sampling period corrected to EPAreference conditions-as Qj. When thesampler's flow rate calibration and operationis based on flow rat~s at ambient conditions.Qtd is calculated as:

Pb Td

T. Pad

where:6Q,-=average flowrate at EPA reference

conditions. std mni/min;Q.=avbage flow rate at ambient conditions,

mr3 min;Pb=average barometric pressure for thesite

or average barometric pressure duringthe-sampling period. kPa.

T==seasonal average ambient temperaturefor the site or-average ambienttemperature during the sampling period.K;

PS=standard pressure, defined as OL3 kPaT,i=standard temperature, defined as 298 K

102 Calculate the total volume of airsampled as.V=.stQdXtwhere:V=total air sampled in standard volume

-units, std m ,t=sampling time, min.

10.3 Calculate the PMo concentration as:

. (Wr-WAX1O6

PMIo=V

where:PMo=mass concentration of PM1 o, jigistd

Wi, Wi=final and initial weights of filter(s)collecting PMo particles. g;

10 6=conversion of g to jig.11.0 References.(1) Quality Assurance Handbook for Air

Pollution Measurement Systems, Volume I.Principles. EPA-600/9-76-005, U.S.Environmental Protection Agency, ResearchTriangle Park North Carolina 27711,1976.

(21 Quality Assurance Handbook for AirPollution Measurement Systems. Volume ILAmbient Air Specific Methods. EPA-600/4-77-027a, U.S. Environmental ProtectionAgency. Research Triahgle Park, NorthCarolina 27711,1977.

(3) Clement.R. E., and F. W. Karasek.Sample Composition Changes in Samplingand Analysis of Organic Compounds inAerosols. Int J. Environ. Analyt. Chem.. 7:101979.

(4) Lee, R. E., Jr., and J. Wagman. ASampling Anomaly in the Determination ofAtmospheric Sulfate Concentration. Amer.Ind. Hyg. Assoc. J.. 27:266. 1966.

(5) Appel, B. R., S. M. Wall. Y. Tokiwa, aniM. Haik. Interference Effects in SamplingParticulate Nitrate in Ambient Air. Atmos.Environ., 13:319,-1979.

(6) Coutant. R. I.Effect of EnvironmentalVariables on Collection of AtmosphericSulfate. Environ. Sci. Technol. 11:873.1977.

(71 Spicer. C. W.. and P. Schumacher.Interference in Sampling AtmosphericParticulate Nitrate. Atmos. Environ.. 11:873.1977.

(8) AppeL B. R., Y. Tokiwa. and hl. Haik.Sampling of Nitrates in Ambient Air. Atmos.Environ. 15:283.1981.

(9] Spicer. C. W.. and P.,M. Schumacher.Particulate Nitrate: Laboratory and FieldStudies of Major Sampling Interferences.Atmos. Environ.. 13:543. 1979.

(10) Appel, B. R. Private Communication.1982.

(11) Pierson. I. R, W. WLV. Brachaczek. T. J.Korniski. T. J. Truex. and J. W. Butler.Artifact Formation of Sulfate. Nitrate. andHydrogen Ion on Backup Filters: AllerhenyMountain Experiment. J. Air. Pollut. ControlAssoc., 30.30.1980.

(12] Chahal, I-L S., and D. J. Romano. High.Volume Sampling Effect of WindhomeParticulate Matter Deposited During IdlePeriods. J. Air PolluL Control Assoc. 26.8M5.1976.

(13) Smith, F.. P. S. Wohlschlegel. R. S. C.Rogers. and D. 1. Mulligan. Investigation ofFlow Rate Calibratidn Procedures Associatedwith the High Volume Method forDetermination of Suspended Particulates.EPA-600/4-78-047. U.S. EnvironmentalProtection Agency, Research Triangle Park.North Carolina 2MIL V1978.

6. Appendix K is added as follows:

Appendix K-Interpretation of theNational Ambient Air Quality Standardsfor Particulate Matter

1.0 General.This appendix explains the computations

necessary for analyzing particulate matterdata to determine attainment of the 24-hourand annual standards specified in 40 CFR§ § 50.6 and 5o.7. For the primary standards.particulate matter is measured in the ambientair as P ,o (particles with an aerodynamicdiameter less than or equal to a nominal 10micrometers) by a reference method based onAppendix I of this part and designated Inaccordance with Part 53 of this chapter or anequivalentrmethod designated In accordancewith Part 53 of this chapter. For thesecondary standard, particulate matter ismeasured in the ambient air by the methoddescribed in Appendix B of this part.hereafter referred to as Total SuspendedParticulate Matter (TSP).

Several terms used throughout theappendix must be defined. A"daily value"for PMo refers to the 24-hour concentrationof PMo calculated or measured frommidnight to midnight (local time- The term"exceedance" means a daily value that Isabove the level of the 24-hour standard afterrounding to the nearest 10 pglm3 (i.e.. valuesending in or greater than 5 are to be roundedup). The term "average" refers to anarithmetic mean. All particulate matterstandards are expressed in terms of expected

d annual values: expected number ofexceedances per year for the 24-hourstandard and expected annual arithmeticmean for the annual standards. The

"expected annual value" is the numberapproached when the annual values from anincreasing number of years are averaged.assuming no long-term trends in emissions ormeteorological conditions.

Although the discussion in this appendixfocuses on monitored data. the sameprinciples apply to modeling data.

2.0 Attaimentletermation.2.1 24-HourPrimaryStandard.Under40 CFR § 50.6fa) the 24-hour primary

standard is attained vhen the expectednumber of exceedances peryear is less thanor equal to one. In the simplest situation, thisdetermination is to be made by recording thenumber of exceedances at a monitoring sitefor each calendar year and then averaingthem over the past 3 calendar years, todetermine whether the average is less than orequal to one. Situations in which 3 years ofdata are not available and possibleadjustments for unusual events or trends arediscussed in Sections 2.4 and 2.5. Moreover,because of the potential for incomplete dataduring a year. it may also be necessary firstto compute an estimated number ofexceedances for a given year by adjusting theobserved number of exceedances. This isdescribed in Section 3. The expected numberof exceedances is then estimated on the basisof the average of the individual annualestimates for the past 3 years.

The comparison with the allowableexpected exceedance rate of oneperyearismade in terms of a number rounded to thenearest tenth (fractional values equal to orgreater than 0.05 are to be rounded up; e.g.,an exceedance rate of 1.05 would be roundedto 1.1. which is the lowest rate fornonattainment).

2.2 Annuol Pr'mary Standard.Under 40 CFR 50.6(b) the annual primary

standard is attained when the expectedannual arithmetic mean PMIo concentration isless than or equal to the level of the standard.In the simplest case, the arithmetic annualmean PM,0 concentration for each calendaryear would be averaged over the past 3calendaryears to estimate whether this valueis less than or equal to the level of the annualprimary standard. Due to the potential forincomplete data and the possible seasonalityIn the PMro concentrations, the annual meanshall be calculated in terms of the fourquarterly means of PMio concentrationswithin the calendar year. The formulas forcalculating the annual arithmetic mean aregiven in Section 4. The expected annualarithmetic means shall be estimated as thearithmetic mean of the individual annualarithmetic means for the past 3 years.Situations in which 3 years of data are notavailable and other possible adjustments forunusual events or trends are discussed InSections 2.4 and 2.5. Comparisons with thelevel of the annual primary standard aremade in terms of integers (fractional valuesequal to or greater than 0.5 are to be roundedup).

2.3 Annual Secondary Standard.Under 40 CFR 501(a). the annual secondary

standard Is attained when the expectedannual arithmetic mean TSP concentration isless than or equal to the level of the standard.In the simplest case, the arithmetic annual

10433

1434Federal Register I Vol; 49, No. 55 I Tuesday. March 20. 1984 I Pronncwrld Rl l

mean TSP concentrations for each calendaryear would be averaged over the past 3calendar years to estimate whether this valueis less than or equal to the level of the annualsecondary standard. Due to the potential forincomplete data and the possible seasonalityin the TSP concentrations, the annualarithmetic mean shall be calculated in termsof the four quarterly arithmetic means of TSPconcentrations within each calendar year.The formulas for calculating the annualarithmetic mean are given in Section 4. Theexpected annual arithmetic mean shall beestimated as the arithmetic means of theseindividual annual arithmetic means for thepast 3 years. Situations in which 3 years ofdata are not available and other possibleadjustments for unusual events or trends arediscussed in Sections 2.4 and 2.5.Comparisons with the level of the standardare made in terms of integers (fractionalvalues equal to or greater than 0.5 are to berounded up).

2.4 Data Requirements.40 CFR 58.13 specifies the required

minimum frequency of sampling for PMo andTSP. For the purposes of making comparisonswith the particulate matter standards, allNational Air Monitoring Stations (NAMS)and State and Local Air Monitoring Stations(SLAMS) data submitted to EPA inaccordance with the Part 58 requirementsmust be used and in minimum of 12 TSP orPMto samples per quarter are required. For aone in six-day sampling schedule, thisrequirement corresponds to a 75 percent datacapture.

To demonstrate attainment of either theannual or 24-hour primary standard or theannual secondary standard at a particularlocation, the monitoring site must providesufficient data in order to perform therequired calculations of Sections 3 and 4. Theamount of data required varies with thesampling frequency, data capture and thenumber or years of record. In all cases, 3years of m'onitoring data that meet theminimum data completeness criterion of theprevious paragraph and are representative of"normal" conditions (defined below] wouldsuffice. In the event that 3 years ofmonitoring data are not available, then 2years of representative data will be sufficientto perform the calculations in order to showattainment of annual standards and 2 yearsof representative data will also be sufficientto perform the calculations for the 24-hourstandard if the monitor samples every dayand achieves an annual average data captureof 50 percent. If only 1 year of representativedata is available, then 1 year of data will besufficient for both annual and 24-hourstandards if the monitor samples every dayand achieves a 75 percent data capture.Furthermore, the calculations for estimatedexceedances will not be necessary if themonitor samples every day and achieves a 75percent data capture for the first year ofmonitoring; in other words, no more than oneobserved exceedance would demonstrateattainment under this condition. Data notmeeting these criteria may also suffice toshow attainment; however, such exceptionswill have to be approved by the RegionalAdministrator in accordance with establishedguidelines (currently under development).

There are less stringent data requirementsfor showing that a monitor has failed anattainment test and thus has recorded aviolation of the particulate matter standards.A single observed exceedance from I year ofrepresentative data may be sufficient todemonstrate nonattainment of the 24-hourstandard as illustrated in example 1 ofSection 3.1 and an annual mean from anindividual representative year may besufficient to demonstrate nonattainment ofthe annual standards. Although it isnecessary to meet the minimum datacompleteness requirement of 12 TSP or PMosamples per quarter to use the computationalformulas described in Sections 3 and 4, thiscriterion does not apply when there areobvious nonattainment situations. Forexample, when a site fails to meet thecompleteness criteria, nonattainment of the24-hour primary standard can still beestablished on the basis of the observed'annual number of exceedances.Nonattainment of the annual standards canbe demonstrated on the basis of quarterlymean concentrations developed-fromobserved data combined with zerossubstituted for missing values.

Normal conditions, associated withrepresentative air quality data, are defined asusual emission levels and typicalmeteorology and are conditions which wouldbe expected to continue to occur in the future.Departures from normal conditions wouldhave to be substantiated on the basis ofhistorical emissions or meterological dataand may require treatment as described inSection 2.5. The designation of air qualitydata as representative, for judging attainmentor non-attainment, is subject to the review ofthe Regional Administrator.

2.5. Considerations for Unusual Eventsand Trends.

In some cases it is possible for a rare orunsual event to result in a PMo or TSPmeasurement which either is not expected tooccur in the future or cannot be controlledthrough the State Implementation Planprocess. Inclusion of such a value in thecomputation of exceedances or averagescould result in inappropriate estimates oftheir respective expected annual values. Toreduce the .effect of unusual events, morethan 3 years of data may be used if thesedata are representative of normal conditions.If 3 or more years of data are not available tominimize the influence of such events, othertechniques, such as the use of statisticalmodels or the use of historical data could beconsidered. For example the use of historicalmeteorological data may be used to establishthe frequency of occurrence of an unusualevent, so that the event may be discounted otweighted according to the likelihood it willreoccur, subject to the approval of theRegional Administrator in accordance withestablished guidelines (currently underdevelopment).

In cases where long-term trends inemissions and air quality. are evident,mathematical techniques should be appliedto account for the trends to ensure that theexpected annual values are notinappropriately biased by data that are notrepresentative of normal conditions. In thesimplest cas&, if 3 years of data are available

under stable emission conditions, this datashould be used. In the event of a trend orshift in emission patterns, either the mostcurrent representative year(s) could be usedor statistical techniques or models could beused in conjunction with previous years ofdata to adjust for trends, subject to theapproval of the Regional Administrator Inaccordance with established guidelines(currently under development).

3.0. -Computation Formulas for the 24-Hour Standard.

3.1. Estimating Exceedances for a Year.Because of practical conoldertions, a PM1o

value may not be available for each day ofthe year. To account for the possible effect ofincomplete data, an adjustment must bemade to the data collected at a particularmonitoring location to estimate the number ofexceedances in a calendar year. In thisadjustment, the assumption Is made that thefraction of missing values that would haveexceeded the standard level Is Identical tothe fraction of measured values above thislevel. This computation is to be made for allNAMS, SLAMS, and all other sites that arescheduled to monitor consistently throughoutthe entire year and meet the minimum datarequirements of Section 2.4. Because ofpossible seasonal imbalance, this adjustmentis not intended for short-term monitoring. Theestimate of the expected number ofexceedances for the year is equal to theobserved number of exceedances plus anincrement associated with the missing data.

The following formula must be used forthese. computations:[Formula 1]e=v+((v/n) X (N-n)]=v X N/nwheree=the estimated number of exceedances for

the year,v=the observed number of exceedances,N=the number of days in the year, andn=the number of days with PMo data.

This adjustment for incomplete data willnot be necessary for monitoring or modelingdata which results in a complete record, I.e.,365 days per year. Other exceptions of theadjustment for incomplete data are discussedin Section 2.4.

If the sampling schedule changes within acalendar year, formula [1] may be appliedquarterly and the annual number ofexceedances may be estimated as the sum ofthe quarterly estimates.

The estimated number of exceeilances for asingle year must be rounded to one decimalplace (fractional values equal to or greaterthan 0.05 are to be rounded up],

Example 1During the most recent calendar year, 01

out of a possible 365 samples were recorded,with one observed exceedance of theapplicable 24-hour standard. Using formula[1], the estimated number of exceedances forthe year ise=1 X 365/61=6.0

If the estimated exceedances for the twoprevious years were both 0.0 then the,expected number of exceedances iestimated by (V3) X (6.0+0+0) = 2,0.Since 2.0 is greater than the allowable

Federal Re ster / Vol.'49, No. 55 / Tuesday March 20 1984 / Proposed Ruler,10434

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 1 Proposed Rules

number of expected exceedances, thismonitoring site would fail theattainment test.

Example 2The sampling frequency at this monitoring

site is once every two days. In the mostrecent year, 183 days were sampled and oneexceedance was recorded. Using formula [1].the estimated number of exceedances is

e=tX36 sa=2.0In each of the-two previous years theestimated number of exceedances was 0.0.Therefore, the expected number ofexceedances is estimated by ( ) X (2.0 + 0+ 0) = 0.7. Since 0.7 is less than theallowable number of expected exceedances,this monitoring site would pass theattainment tesL

3.2 Adjustments for Non-ScheduledSampling Days.

if a systematic sampling schedule is usedand sampling is performed on days inaddition to the days specified by thesystematic sampling schedule, e.g., episodeconditions, then an adjustment must be madein the formulT for the estimation ofexceedances. This is intended to eliminateany bias in the estimate of the annual numberof exceedances, which could occur if thechance of an exceedance is different betweenscheduled and non-scheduled days, as wouldbe the case with episode sampling.

This approach effectively treats thesystematic sampling schedules as a stratifiedsampling plan. If the period from onescheduled sample until the day preceding thenext scheduled-sample is defined'as nsampling stratum, then there is one stratumfor each scheduled sampling day. An averagenumber of observed exceedances iscomputed for each of these sampling stratum,allowing for differences in exceedance ratesthat may exist during periods of unscheduledsamples, such as episode conditions. Withnonscheduled sampling days, the estimatednumber of exceedances is defined as[Formula 2]

me={Nlm} x I (vjlkj}

j=1

wheree=the estimated number of exceedances for

the year,N=the number of days in the year,ni=the number of strata with samples during

the year,vj= the number of observed exceedances in

stratum j, andkj=the number of actual samples in stratum j.Note that if only one sample value isrecorded in each stratum, then formula [2]reduces to formula [1].

Example 3A monitoring site samples according to a

systematic sampling schedule of once every 6days for a total of 61 samples in a year.

During one 6-day period, potential episodelevels of PMNo were suspected, so 3 additionalsamples were taken. Five of the regularlyscheduled samples were not recorded, so atotal of 59 samples in 56 sampling strata weremeasured. The one 6-day sampling stratumwith 4 samples recorded 2 exceedances. Theremainder of the year with one sample perstratum recorded a single exceedance. Usingformula [2], the estimated number ofexceedances for the year ise= (365/56) X (2/4 +1) =9.8

If formula [1] were used Instead of therevised formula [2], then the estimatednumber of exceedances for the year wouldhave beene=3X365/59=18.6.This computation would have produced anestimate for exceedances that Is roughlytwice as high as that produced by the revisedformula.

4.0 Computational FormulaforAnnualStandards.

4.1 Calculation of the Annual ArithmeticMean.

The annual arithmetic means for PMo andTSP shall be based on the average of thequarterly means for each calendar quarter forthe most recent representative calendar yearsof data at a particular monitoring site. Thefollowing formula Is to be used forcalculation of mean for a calendar quarter.[Formual 3]

a,

i=1

where&1=the quarterly mean concentration for

quarter q. q=12. 2 3. or4,n= the number of samples In the quarter.

andxi=the ith concentration value recorded in

the quarter.The average of quarterly means must be

rounded to the nearest integer (fractionalvalues of 0.5 should be rounded up). Theannual mean is calculated by using thefollowing formula:[Formula 4]

4

x=(1l4) X Z i,

q=2

where:R=the annual mean. andx.= the mean for calendar quarter q.. The use of quarterly averages to computethe annual average will not be necessary formonitoring or modeling data which results ina complete record, i.e., 365 days per year.

Exomple 4Using formula [3]. the quarterly means are

calculated for each calendar quarter. If the

quarterly means are 52.4.75.3,82.1. and 63.2pg/m3. then the annual mean is

*=(V×X(52.4+75.3+821+63.2)=8.3 or 68

4.2 Adjustments forNon-ScheduledSampling Days

An adjustment in the calculation of theannual mean is needed if sampling isperformed on days in addition to the daysspecified by the systematic samplingschedule. Using the notation and rationaledescribed for estimated exceedapces (Section3.2). the quarterly averages would becalculated by usrnd the following formula:[Formula 5]

mqk

&=(llmJx 1 2 (xu/k]j =1i=1

where&=the quarterly mean concentration for

quarter q. q=1. 2, 3. or 4.xu= the Ith concentration value recorded in

stratum 1.kj=the number of actual samples in stratum j.

andm.=the number of strata with data in the

quarter.

If one sample value s recorded in eachstratum, formula [5] reduces to a simplearithmetic average of the observed values asdescribed by formula [31.

Example 5

During one calendar quarter, 9observations were recorded. These samples.;ere distributed among 7 sampling strata.with 3 observations in one stratum. Theconcentrations of the 3 observations in thesingle stratum were 202. 242. and 180 pglm3 .The remaining 6 observed concentrationswere 55. 6. 73. 92.120, and 155 pg/m3.Applying the weighting factors specified informula [5]. the quarterly mean is

&=('h)x[(VJx(22+24Z+180]+55+68+73+92+120+1551=110.1

Although 24-hour measurements arerounded to the nearest 10 jg/m3 fordeterminations of exceedances of the 24-hourstandard, note that these values are treatedas integers for the calculation of means.(42 U.S.C. 7403 and 7409)

IFR Dfl. D,:,-co F5Ld 3-19-eu 8:45 amleUWJ-,l CODE sslo-so-M

40 CFR Part 58

[AD-FRL 2491;6]

Ambient Air Quality Surveillance forParticulate Matter

AGENCY Environmental ProtectionAgency.

ACTION: Proposed rule.

a10435

10436 Federal Register / Vol. 49, No. 55 / Tuesday, Marsh 20, 1984 / Pi'oposed Rules

SUMMARY: EPA proposes to amendprovisions of Part 58 of Chapter I of Title40 of the Code of Federal Regulations(CFR) to account for revisions to theNational Ambient Air Quality Standards(NAAQS) for particulate matter (PM)that are being proposed elsewhere iitoday's Federal Register. Under theproposed revisions to the primaryambient standard, particulate matterwould be measured in the ambient air asPM,o (particles with an aerodynamicdiameter less than or equal to a nominal10 micrometers). Under the proposedrevision to the secondary ambientstandard, particulate matter wouldcontinue to be measured in the ambientair by the method described inAppendix B (40 CFR Part 50) hereinafterreferred to as Total SuspendedParticulate Matter (TSP). Thisnecessitates proposed revisions to Part68 that would establish ambient airquality monitoring requirements forPMo as measured by a new referencemethod being proposed as AppendixJ(40 CFR Part 50) elsewhere in today'sFederal Register or an equivalentmethod. In addition new network designand monitoring siting requirements arebeing proposed for the secondary TSPstandard. The proposed requirementsare comparable to those alreadyestablished for the other pollutants(criteria pollutants) for which NAAQShave been set. These includerequirements for reporting and assuringthe quality of ambient PM1o data, for thedesign of monitoring networks and thesiting of samplers for both TSP and -PMo. Also proposed are revisions to thelead (Pb) and sulfur dioxide (SO 2) lowerlimits for precision estimates containedin Section 4.2 of Appendix A. Finallyseveral clarifying changes are proposedin each of the appendices. Proposedrevisions to EPA's regulationsconcerning Requirements forPreparation, Adoption, and Submittal ofImplemention Plans (40 CFR Part 51)with associated guidelines, will bepublished later. Following thepublication of these notices, the Agencywill announce a supplementary reviewperiod for the limited purpose of takingcomments on the implications of theproposed Part 51 implementationrequirements and guidelines on the Part58 regulations proposed today.DATES: Comments must be submitted onor before June 18, 1984, for the proposalsherein.ADDRESSES: In conjunction with themonitoring requirements for PMo, twodraft guideline documents have beenprepared and are included in Docket No.A-83-13 for public comment. These twodraft documents are entitled:

* Guideline for Particulate EpisodeMonitoring Methods.

* Optimum Network Design and SiteExposure Criteria for Particulate Matter.

These documents are available forinspection and copying at:

* The Central Docket Section.* State Air Programs Branch, U.S.

EPA, Region I, JFK Federal Building,Boston, MA 02203.

e Air Programs Branch, U.S. EPA,Region II, 26 Federal Plaza, New York,NY 10278.

- Air Programs and Energy Branch,U.S. EPA Region III, Curtis Building, 6thand Walnut Streets, Philadelphia, PA19106.

* Air Management Branch, U.S. EPA,Region IV, 345 Courtland Street, NEAtlanta, GA 30365.

* AirPrograms Branch, U.S. EPA,Region V, 230 S. Dearborn Street,Chicago, IL 60604.

e Air Programs Branch, U.S. RegionVI, First International Building, 1201 ElmStreet, Dallas, TX 75270.

* Air Programs Branch, U.S. EPA,Region VII, 324 East 11th Street, KansasCity, MO 64106.

* Air Programs Branch, U.S. EPA,Region VIII, 1860 Lincoln Street, Denver,CO 80295.

• Air Programs Branch, U.S. EPA,Region IX, 215 Fremont Street, SanFrancisco, CAS 94105.

- Air Programs Branch, U.S. EPA,Region X, 1200 6th Avenue, Seattle, WA98101.

Submit comments (duplicate copiesare preferred) to: Central DocketSection, U.S. Environmental ProtectionAgency, Attn: Docket No. A-83-13, 401M Street, S.W., Washington, DC 2(1460.Docket No. A-83-13 is located in theCentral Docket Section of theEnvironiental Protection Agency, WestTower Lobby Gallery I, 401 M St., S.W.Washington, D.C. The docket may beinspected between 8:00 a.m. and 4:00p.m. on week days and a reasonable feemay be charged for copying.FOR FURTHER INFORMATION CONTACT:Neil Berg or Stanley Sleva, Monitoringand Data Analysis Division (MD-14),Office of Air Quality Planning andStandards, Environmental ProtectionAgency, Research Triangle Park, N.C.27711, phone: 919-541-5651 or (FTS) 629-5651.SUPPLEMENTARY INFORMATION:

Background

Elsewhere in today's Federal Register40 CFR Part 50, EPA is proposing revisednational ambient air quality standards(NAAQS) for particulate matter and anew reference method for thedetermination of ambient concentrations

of particulate matter with anaerodynamic diameter less than or equalto a nominal 10 micrometers (PM,o).Corresponding revisions are also beingproposed elsewhere in today's FederalRegister to the regulations in 40 CFRPart 53, Ambient Air Monitoring,Reference and Equivalent Methods, Themethod discussed in Appendix B of 40.CFR Part 50 will continue to be used tomeasure particulate matter in theambient air for purposes of determiningattainment of the proposed revislons tothe secondary standard hereinafterreferred to as Total SuspendedParticulate Matter (TSP). The AppendixB method will also be used inconjunction with Appendix G of Part 50(Reference Method for theDetermination of Lead in SuspendedParticulate Matter Collected fromAmbient Air) as well as for otherpurposes as specified in the proposedrevisions to this Part. Additionalinformation on these proposed actionscan be found in the respective noticesand comments on them should be sent-to the addresses provided therein.

Section 110(aJ(2)(C) of the Clean AirAct requires ambient air qualitymonitoring for purposes of the StateImplementation Plans (SIP's) and forreporting air quality data to EPA.Criteria to be followed when measuringair quality and provisions for daily airpollution index reporting are required by

Section 313 of the Act. To satisfy theserequirements, on May 10, 1979 (44 FR27558), EPA established 40 CFR Part 58which provided detailed requirementsfor air quality monitoring, datareporting, and surveillance for all of thepollutants for which ambient air qualitystandards have been established(criteria pollutants) except lead. OnSeptember 3, 1981 (44 FR 27558), similarrules were promulgated for lead. Theregulations in this notice deal with theambient air quality monitoring, datareporting, and surveillance requirementsassociated with today's proposedrevisions of the particulate matterstandards. Comments sent to Docket No.A-83-13, therefore, should concern onlythe regulations being proposed in thisnotice, and the two previouslymentioned guideline documents,Guideline for Particulate EpisodeMonitoring Methods, and OptimumNetwork Design and Site ExposureCriteria for Particulate Matter.

Proposed Revisions to Part 58--AmbientAir Quality Surveillance

Section 58.1, Definitions.

The revisions proposed today wouldadd definitions of the terms "TSP" (total

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules 4

suspended particulates), "Pb" (lead),and "PM' (particles with anaerodynamic diameter less than or equalto a nominalIQ micrometers). Thedefinition for Pb is added at this timebecause it was inadvertently left o6t inthe promulgation of the Pb monitoringregulations.

Section 58.13, Operating Schedules

The current monitoring regulationsspecify that, "for manual methods, atleast one 24-hour sample (is required)every 6 days except during periods orseasons exempted by the RegionalAdministrator." The revision proposedfor this section would require the Statesto conduct more frequent sampling forPMo by manual methods in order toprovide for more accurate SIP designvalues and more correct attainment/nonattainment determinations. Theserevisions will apply only to the site withexpected maximum doncentration ineach monitoring area. It is recognizedthat the probability of detectingexceedances of the 24-hour standard isdependent upon the frequency ofmonitoring, therefore, more-frequentsampling is required to overcome thedeficiency of thd current monitoringrequirements in detecting exceedancesof the 24-hour standard. While everysixth day monitoring is adequate forestimating annual averageconcentrations, more frequent

-monitoring is considered necessary toassess status with respect to the 24-hourstandard.

The operating schedule proposed forthe measurement of PMo will consist ofa short-term and long-term monitoringplan. The short-term monitoringschedule will be based on a probabilityassigned to areas charadterizing thelikelihood that they are not attaining thePMo standards. The use of PMononattaifrment probabilities is necessarysince most areas of the country do nothave PMo ambient monitoring data.These probabilities will-be developedaccording to "Procedures for EstimatingProbability of Nonattainment of a PMoNAAQS Using Total SuspendedParticulate or Inhalable ParticulateData." OAQPS, U.S. EnvironmentalProtection Agency, Research TrianglePark, N.C. February 1984. Theseprobabilities will be classified into threecategories: low (<20 percent), medium(>20 to <95 percent) and high (>95percent). A low probability, defined as aprobability less than 20 percent, waschosen as a reasonable cutpoint in orderto identify areas which are unlikely tobe nonattainment. Such areas wouldhave less than a one-in-five chance ofbeing nonattainment. A mediumprobability, defined as a probability

greater than or equal to 20 but less than95 percent was chosen to identify areaswhich are more likely to benonattainment and includes areas whichwould be expected to be near thestandards. A high probability, definedas a probability greater than or equal to95 percent was chosen as a reasonablecutpoint to identity areas which aremost likely to be nonattainment; such -

areas would have a less than one-in-twenty chance of being attainment.

Data collection requirements for thefirst year of monitoring (short term) shallbe based on the estimated probabilitiesof not attaining the PMo standards andthe associated importance of additionalPMo ambient data. For the areas with alow probability of not attaining the PM~ostandards, the value of collecting morethan a minimum of PM,o data isrelatively low. Such areas are likely to

'be attainment, and intensified PMo datacollection is not warranted. Due to thesmall chance for being nonattainment,however, a minimum sampling programis still required. Accordingly, a minimumsampling schedule of one in 6 dayswould be required.

Areas with a medium probability ofnot attaining the PM,o standards arelikely to either be near or in exceedanceof the level of the standard. For suchareas, the chance of misclassifying-current attainment status is high,especially with respect to the 24-hourstandard. Moreover, the value ofadditional PMo information is importantif it is found that a State ImplementationPlan (SIP) must be prepared.Consequently, a more intensifiedsampling schedule of once in 2 dayswould be required.

For areas with a high probability ofnot attaining the PMo standards, thevalue of a first year intensified PModata collection is most important. This isbecause these areas are most likely tobe required to develop a SIP.Consequently, everyday sampling for aminimum of 1 year is being proposed forthese areas in order to confirm aprobable nonattainment status, as wellas the degree. Although the specificdetails of EPA policy for PM2ho controlstrategy development will be proposedat a later date with revisions to Part 51"Requirements for Preparation,Adoption, and Submittal ofImplementation Plans" and associatedguidelines, this monitoring requirementwill facilitate the determination ofcorrect air quality status so thatappropriate action can be made in atimely fashion. The short-term strategyalso contains provisions for monitoringto be intensified to everyday at the siteof expected maximum concentration in

medium and low probability areas, ifexceedances of the 24-hour standard aremeasured during the first year ofmonitoring.

The-long-term selective samplingschedule is based on an analysis of theratio of measured PM,. concentrationsto the controlling PM standard.Depending upon the ratio, the samplingfrequency could be either everyday,every other day, or every sixth day. The -

long-term monitoring strategy isdesigned to optimize monitoringresources and maximize informationconcerning attainment status. Similar tothe short-term strategy, the increasedsampling frequency provisions onlyapply to the site with expectedmaximum concentration in eachmonitoring area.

For thoseareas wherein the annualstandard is controlling. 1 in 6 daymonitoring would be required; thisfrequency is adequate for assessingstatus with respect to this standard. Forthose areas wherein the 24-hourstandard is controlling, the requiredminimum sampling frequency, for thecalendar year, will vary according to therelative level of the most currentmaximum concentration site to the levelof the standard. In other words, thesampling requirement applies to the sitewhich drives attainment/nonattainmentstatus for the monitoring area. Theleastfrequent monitoring (1 in 6 days) wouldbe required for those areas wherein themaximum concentration site is clearlyabove the standard {>40% above orclearly below the standard (<267below). For such sites a minimumamount of data collection would beadequate to verify correct attainment/nonattainment status. As the areaapproaches the standard, the monitoringfrequency for the maximumconcentration site would increase sothat the misclassification of correctattainment/nonattainment status can beminimized. If the area is either 10-20percent below or 20-40 percent abovethe 24-hour standard. 1 in 2 daymonitoring would be required. When thearea is close to the standard, i.e. 10percent below to 20 percent above,everyday sampling would be required inorder to maximize the stability of theattainment/nonattaiment classification.Modification to the sampling schedulewill be based on the SLAMS networkannual data review, in which the mostrecent calendar year of air quality datawould generally be used to determinecurrent air quality status. Although themost recent year of measured datawould be considered first in determiningwhich site would operate at the morefrequent sampling schedule.other

10437

10438 Federal Register [ Vol. 49, No. 55 / Tuesday. March 20. 1.q4 I Prnnnoi~r R,,Ic

factors should also be considered. For'example, the most recent three years ofdata might be used to provide stabilityto the network. In addition, anappropriate adjustment for trends couldbe made when a change in air qualitystatus is shown to correspond to acommensurate change in underlyingparticulate matter emissions. Finally,major changes in sources of PMoemissions or in sampling sitecharacteristics could possibly influencethe location of the expected maximumconcentration area PM1 o site. If thelocation of expected maximumconcentration varies annually, EPAsuggests that the recommendedmonitoring schedule be used at morethan one site.

The annual operating costs associatedwith the sampling schedules of thisproposed regulation are approximately 5million dollars for the most stringentstandard described in Part 50. This iscomparable to the operating costs forthe other criteria pollutants. The costsfor the other sampling optionsconsidered are presented in "PM 0Monitoring Costs for Three SamplingOptions," (October 26, 1983). A moredetailed discussion of the rationale forthe selected sampling options iscurrently being developed and will beavailable for public comment as a paperentitled, "Revising the NAAQS forParticulate Matter-A SelectiveSampling Monitoring Strategy."Section 58.20, Air Quality Surveillance:Plan Content

The revisions proposed today wouldrequire the States to have their revisedair quality network descriptions for TSPand their new air quality networkdescriptions for PMo available forsubmission to the appropriate RegionalAdministrators by 6 months aferpromulgation. Since most PMomonitoring stations are expected to bechosen from existing TSP or Pbmonitoring stations, EPA believes it isreasonable to require States to submittheir PMo State and local air monitoring.station (SLAMS) network descriptionswithin 6 months of promulgation ofthese regulations. Although some TSPmonitors may need to be relocated tofulfill the micro/middle scale TSPsecondary requirements, the 6 monthdeadline also applies to SLAMS TSPnetwork descriptions since the new TbPSLAMS network will, in most cases, bea modification of the existing TSPSLAMS network.

As described in the proposedrevisions to section 2.2 of Appendix C,EPA would allow the continued use ofTSP high volume samplers in PMoSLAMS as a substitute for PMo

samplers as long as measured 24-hourTSP concentrations and annual TSPlevels remain below the PMo standards.Such substitute TSP samplers wouldhave to be identified and included in theinitial SLAMS PMo networkdescriptions. Should a TSP samplerrecord TSP levels that exceed the PM 0NAAQS, the substitute TSP samplerwould be considered for replacement bya PMo sampler during the annualSLAMS network review required by§ 58.20(d).

Section 58.23, Monitoring NetworkCompletion

Two dates, 1 year and 2 years afterpromulgation are proposed forcompletion of the PMo SLAMS network.By I year after promulgation, each areawithin the approved SLAMS network forwhich a probability of PMo NAAQSnonattainment is greater than or equalto twenty percent, must have at leastone PMo sampler which is located in thearea of expected maximumconcentration in operation, be sited inaccordance with Appendix E, be locatedas described on the station's SAROADidentification form and meet all of thequality assurance requirementspertinent to PM1o contained in AppendixA. The remaining PMo samplers haveuntil 2 years after promulgation to befully operational and to meet the sitingand quality assurance requirements.This latter date would also beapplicable to the relocated TSPsamplers required by the secondaryNAAQS monitoring requirements.

The SLAMS network design andprobe requirements being proposedtoday (as revisions to Appendices D andE respectively) would be similar to thoserequired for TSP in the currentAppendices D and E; the majordifference would be the addition of amicroscale to the applicable spatialscale for monitoring TSP and PMo.(Definitions of monitoring scale3 ofrepresentativeness and SLAMSmonitoring objectives and spatial scaleare found in section 1 of Appendix D.)Since the expected size of the revisedTSP SLAMS network and the PMoSLAMS network would beapproximately the same or smaller thanthe current TSP SLAMS network,complying with the revised requirementsof Appendices A, D and E should notpose unmanageable resource burdens.Although the Agency does not anticipateany major problem in SLAMS networkdesign, siting, or quality assurance, itdoes recognize that because of thepotential lack of sufficient commerciallyavailable PMo reference or equivalentsamplers the States may havedifficulties in completing their entire

.planned SLAMS PMo network. Becauseof this situation, and based on the timerequired to complete SLAMS for theother criteria pollutants, the Agencyproposes to allow a two year timeperiod for SLAMS TSP and a two yearphased approach for PM10 networkcompletion.

Section 58.26, Annual SLAMS SummaryReport, and§ 58.27, Compliance Datafor Air Quality Data Reporting

These sections, which remainunchanged, specify when the SLAMSannual reporting procedures were tobegin. Under these procedures the datafrom each particulate matter SLAMS arecurrently summarized and submitted Inthe annual report as TSP data. After thenew NAAQS are promulgated, therewould be a requirement to report PMtodata in the annual report. However, asdiscussed more fully in a later part ofthis preamble, the revision proposed toAppendix C of Part 58 would allow highvolume samplers to be used undercertain circumstances as substitutesamplers for PM 0 samplers, and therevisions being proposed to Appendix Fwould provide for reporting of both TSPand PMo data. Accordingly, EPA wouldexpect States to continue reporting TSPdata from each TSP slams until thatstation is taken out of service. Also, aseach PMo SLAMS is put into operation,the PMo data would be included In theannual report as required by § 58.20.There would, therefore, be a gradualtransition in the data reporting processand no revision is needed to § 58.26 or§ 58.27 to establish an initial reportingdate for PM1 o annual summary data.

Section 58.30, NAMS NetworkEstablishment

The revision proposed today woulddesignate 6 months after promulgationas the date by which the National AirMonitoring Station (NAMS) networkportion of each State's SLAMS networkmust be fully described and documentedin a submittal to the Administrator(through the appropriate RegionalOffice). Since the number of PMoNAMS required would be less than thenumber of existing TSP NAMS, EPAbelieves that the design of the PMtoNAMS network can reasonably beaccomplished and submitted within sixmoriths after the promulgation of theseregulations. Also, since the new TSPsecondary NAAQS NAMS network willlikely be smaller than the existing TSPprimary NAAQS NAMS Network, sixmonths is a reasonable time to designand submit a description of the new TSPnetwork as well.

Federal Register I Vol. 4b, No. 55 / Tuesday March 20 1984 / Proposed Rules-10438

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules0

Section'58.34, NAMS NetworkCompletion

The revision proposed today woulddesignate 1 year after promulgation asthe date by which the State must haveall PMo and TSP NAMS in operation.Specifically, each PMo and TSP NAMSwould have to be sited in accordancewith the-criteria in Appendix E, belocated as described in the stationsSAROAD site identification form and beoperating under the quality assurancerequirements of Appendix A. TheAgency believes this shorter period forcompletion of th6 NAMS portion of theSLAMS network is reasonable in viewof the smallernumber of TSP and PMoNAMS versus the entire TSP and PMoSLAMS network. EPA also anticipatesthat an adequate number of PM10reference or equivalent samplers will be

-available within 12 months afterpromulgation of these regulations.

Section 58.35, NAMS Data Submittal

Today's proposed-revisions woulddesignate 90 days after the first quarterof operation as the date by which datacollected during the first quarterlyperiod after PMo NAMISnetwork -completion mustbe reported. Thepurpose of this revision would be toestablish a date for the submission ofthe initial quarterly report of data fromPMo NAMS and the relocated or newlyestablished TSP NAMS. States havingPMo NAMS operating according to, allPart 58 criteria prior to 1 year afterpromulgation would be encouraged tosubmit data from those stations in theearliest NAMS report possible.

Reirfsions to Appendix A-

Appendix A sets forth quality,assurance and quality-assessmentrequirements for ambient air monitoringdata. Revisions to various sections areproposed to include-appropriate dataquality assessment procedures for PMomonitoring. The proposed accuracy andprecision assessment procedures forPM 0 would be very similar to thecurrent requirements for-TSP. Accuracywould be assessed with rotating flowaudits each quarter, and precisionwould be assessed with collocatedsamplers.

A minor change in section 4.2.1(a)would eliminate, from the precisioncalculations, paired measurements fromcollocated samplers where eithermeasurements is below a specifiedlower limit. This would avoid'exaggerated estimates of precision thatoften result from very low.measurements. The specified lower limitfor PMo would be 20 /g/m3, and thelower limits for SO 2 and Pb would be

changed slightly to provide moremeaningful precision estimates: SO.from 40 to 45 pg/m s , and Pb from 0.15 to0.25 ig/m3.

The data assessment report form(Form 1, Figure 1, Back) would berevised to add data blocks from PIModata, and the instructions for the form inAppendix A, Section 5.3 would berevised accordingly.

Revisions to Appendix BAppendix B contains quality

assurance requirements for PSDmonitoring. Appropriate amendments toextend the data quality assessmentrequirements explicitly to PMomonitoring are proposed. As inAppendix A. a new provision forexcluding from the precision caclulationany paired measurements fromcollocated samplers below specifiedlimits would be added to apply to TSP,Pb, and PMo.

Revisions to Appendix CBecause TSP high volume samplers

measure a larger particle size fraction ofsuspended particulate matter than PMjosamplers, EPA believes that after thispromulgation there would be no need torequire the high volume sampler to bereplaced with PMo sampler in theSLAMS network in stations wheremeasured TSP ambient concentrationlevels are below the PMo ambientstandards. State or local agencies wouldbe allowed to continue to operate thehigh volume sampler to demonstratecompliance with ambient PMostandards as long as measured TSPlevels remain below those standards. Assoon as a TSP sampler measures asingle value which is higher than thePMo 24-hour standard or has an annualaverage greater than the PMjo annualstandard, it would be necessary toreplace the high volume samplersdesignated as substitute P, 0o samplerswith PMo samplers. This is because thePMo portion of TSP varies from area toarea and could possibly be close to 100percent of the TSP during air stagnationperiods in some areas. This correctionmust be reflected in the SLAMS annualnetwork review.

This proposed revision to Appendix Cwould be added as section 2.2, and theexisting material in section 2.2 would bedeleted because it involvesrequirements that are now out of date.

In addition to allowing the continueduse of the high volume (TSP) method asa substitute for a PMio sampler incertain areas, EPA believes that there isa strong need to require a limitedamount of TSP and PMjo air qualitysampling (at least one year ofcollocatedsampling) at those existing TSP NAMS

which will be designataed as P.MiNAMS. This requirement would providesupporting data to determine possiblerelationships between TSP and PMo air-quality data so that historical trends andpatterns for ambient particulate mattercan be continued. Because the TSP/Marelationships may vary geographicallyand seasonally, the proposed rule wouldrequire that States continue ta operate aTSP sampler for one year at all PMiaNAMS that were previously TSP NAMS.The year would begin for each station atthe time the PMa sampler in the PMlrNAMS is put iiito operation.

The requirements in section 4.0 forepisode monitoring would be revised torreplace "TSP" with "PMw" and allreferences to the high volume methodwould be changed to reflect the newreference method for PMi sinceepisodes would be based upon PMorather than TSP.

Also, the reference in section 5.0pertaining to selecting TSP episodemonitoring methods would be replacedwith a reference to the document,Guideline for Particulate EpisodeMonitoring Methods, which providesguidance on selecting PMi, episodemonitoring methods. EPA soliciLs -

comments on this guideline document.

Revisions to Appendix DThe revisions to AppendixfD proposed

today would revise sections 2.2 and 3.,1to incorporate changes necessitated bythe revised TSP secondary NAAQS. addnew sections 2.8 and 3.7 dealing withconsiderations necessary in the designof PMo SLAMS and NAMS networksand revise section 3.2 to deletereferences to TSP in the SOC section.Also added are references to guidelineato be used in siting PM10 stations andreferences for calculating PMa NAAQSnonattainment probability.

The revisions to section 7.2 wouldallow micro scale TSP SLAMS to beused to monitor for compliance with therevised TSP secondary NAAQS,

The revisions to section 3.1 wouldrevise the range for the number of TSPNAMS required in urbanx areas.

Section 2.8 is a new section thatdescribes the criteria that would be usedin designing the SLAMS network andspecifies situations where the PMomonitoring requirements could be metby TSP samplers (in accordance withthe proposed revision to Appendix Cj.This section also includes descriptionsof applicable PMo spatial scales.. Section 3.2 specifies SOz designcriteria for NAMS and makescomparisons to TSP monitoring. Becauseof the proposed revisions to the TSPmonitoring criteria it is appropriate to

104f39f

Federal Register / Vol. 49, No. 55 / Tuesday, March 20. 1984 / ProDosed Rules

eliminate comparisons to TSP in the SOsection at this time.

Section 3.7 is a new section. It wouldestablish design criteria for determiningthe number of NAMS based on PMoNAAQS nonattainment probability,source types, and urban area population.Consistent with design criteria for otherpollutants, one of the NAMS would be acategory (a) maximum concentrationstation, and the other a category (b)population exposure station. For PMo,category (a) stations would be micro ormiddle scale and category (b) stationswould be neighborhood scale. As notedIn Appendix C, all PM1o NAMS thatwere previously designated as TSPNAMS would be required to collect bothambient TSP and PMo data for a one-year period starting at the time the PMosampler in the PM1o NAMS is put intooperations.

Revisions to Appendix EToday's proposed revisions to

Appendix E would revise section 2 toinclude revised TSP siting requirements,and add a new section 8 to cover thespecific PMo sampler sitingrequirements.

The proposed changes consist of sitingparameters that would specify spacingdistances from roadways and minimumseparation distances from buildings,trees and other obstructions. Verticaland horizontal sampler distances arealso proposed so that sampling can beconducted representative of thebreathing zone while preventingvandalism to the sampler.

For microscale sites, a height of 2-7meters is proposed while for larger scalesites, a height of 2-15 meters isproposed. A horizontal roadway setbackdistance of 5-15 meters is proposed for amicroscale roadway site and a range ofdistances is proposed for larger scales.EPA welcomes further comments on thissubject, as well as comments on theguideline document, Optimum NetworkDesign and Site Exposure Criteria forParticulate Matter, mentioned earlierunder the heading, AVAILABILITY OFRELATED INFORMATION. The othersiting provisions being proposed areconsistent with those currently requiredfor the other criteria pollutants.Revisions to Appendix F

As a result of the revisions proposedfor the particulate matter standards,revisions would be necessary to thedata reporting requirements ofAppendix F. A new section 2.7 isproposed to be added to Appendix Fthat would specify the ambient PMoinformation that would be required inthe annual air quality data report. Asdiscussed under Appendix C, States

could operate high volume samplers intheir PMo SLAMS network wheremeasured TSP levels were below thePMp ambient standards. This TSP datawould be reported as such in the annualSLAMS report. The annual TSParithmetic mean is required in order toestimate PMo levels in those caseswhere TSP is measured as a substitutefor PMo. Also, the annual TSPartithmetic mean is required in order todetermine compliance with the proposedsecondary TSP standard. Section 2.2,therefore, would be retained to addressTSP reporting but modified to beconsistent with Section 2.7.

New concentration ranges related tothe proposed PM standards areproposed in section 2.7 for compilingPMo data. For consistency, in section2.2, compatible ranges are also proposedfor compiling TSP data. Also, the TSPannual mean would be changed from ageometric mean to an arithmetic mean.In order to calculate the expectednumber of exceedances of the 24-hourPM1o ambient standard (40 CFR Part 50,Appendix K, section 2.2), allexceedances of the standard would bereported as well as the samplingfrequency. The proposal, therefore,contains the requirement to report all.24-hour values exceeding the 24-hour PM1oambient standard, their dates ofoccurence, and the sampling schedule.

According to the procedures ofAppendix K, Part 50, the calculations todetermine attainment/nonattainment ofthe PMo ambient standards also musttake into account episode statistics. Ifepisoie occurrences are not taken intoaccount, an overestimate of the numberof exceedances, as well as the level ofthe annual arithmetic mean could result.

In section 2.2.2, only the ten highestTSP values above the PMo standardsare proposed to be reported. Since onlyTSP data would be available until PM 0samplers are installed, the ten highestvalues would be sufficient for statisticalpurposes.

Revisions to Appendix GThe proposed requirements in

Appendix G for air quality indexreporting are to eliminate provisionsrelated to TSP or TSP x S0 2 and to addprovisions for PMo. A number ofproposed revisions involve removingTSP provisions (including the examplecalculation) in section 7.2, and replacingthem with PM 0. The pollutant standardindex (PSI) function in Figure 2 and thesecond column in Table I are proposedto be renamed "PMo" and revised toreflect breakpoints which coincide withthe proposed episode and significantharm levels for PM 0. This is consistentwith the provisions for other pollutants.

The proposed action also Includesremoving the breakpoints and PSIfunction figure for TSP x SO.

A minor proposed change involves atypographical error in the fifth column ofTable 1. The correct term for the columnis mg/m3 instead of umg/m3.

Impact on Small Entities

The Regulator Flexibility Act requiresthat all federal agencies consider theimpacts of final regulations on smallentities, which are defined to be smallbusiness, small organizations, and smallgovernmental jurisdictions (5 U.S.C. 601et seq.). EPA's consideration pursuant tothis Act indicates that no small entitygroup would be significantly affected inan adverse way by the proposal.Therefore, pursuant to 5 U.S.C. 605(b),the Administrator certifies that thisregulation will not have a significanteconomic impact on a substantialnumber of small entities.

Other Reviews

The regulatory impact of the proposedrevisions to Part 58 is addressed withinthe Regulatory Impact Analysis (RIA)referenced under the Proposed IRevisions to the National Ambient AirQuality Standards for Particulate MatterPublished elsewhere in today's FederalRegister.

The proposed revisions to Part 58were submitted to the Office ofManagement and Budget (OMB) forreview (under Executive Order 12291),This is not a "major" rule under E.O.12291 because it does not meet any ofthe criteria defined in the ExecutiveOrder.

The reporting and recordkeepingprovisions addressed in this notice,however, have been submittedseparately for review by OMB undersection 3504(b) of the PaperworkReduction Act of 1980 U.S.C. 3501 et seq.Any OMB comments and EPA responsesto those comments are available forpublic inspection at EPA's CentralDocket Section (Docket No. A-83-13),West Tower Lobby, Gallery, I,Waterside Mall, 401 M Street SW.,Washington, D.C.

List of Subjects in 40 CFR Part 50Air Pollution Control,

Intergovernmental relations, Reportingand recordkeeping requirements,Pollutant standard index, Ambient airqualitymonitoring network.(Sees. 110, 301(a) and 319. Clean Air Act, 42USC 7410, 7601(a). 7019)

10440i m

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 1 Proposed Rules

Dated: March 8, 1984.William D. Ruckelshaus,Administrator.

PART 58--AMBIENT AIR QUALITYSURVEILLANCE

For the reasons set out in thepreamble, Part 58 of Chapter I of Title 40of the Code of FederalRegulations isproposed to be amended as follows:

1. Section 58.1 is amended by addingnew paragraphs (t), (u), and (v) asfollows:

§ 58.1 Definitions.

(t) "TSP" (total suspendedparticulates) means particulate matteras measured by the method described inAppendix B of Part 50 of this chapter.

(u) "PM1o" means particulate matterwith an aerodynamic diameter less thanor equal to a nominal 10 micrometers asmeasured by a reference method basedon Appendix J of Part 50 of this chapterand designated in accordance with Part53 of this chapter or by an equivalentmethod designated in accordance withPart 53 of this chapter.

(v] "Pb" means lead.

2. Section 58.13 is amended byrevising paragraph (b) and addingparagraph (c] to read as follows:

§ 58.13 Operating Schedule.

(b) For manual methods (excludingPM1o samplers)-at least one 24-hoursample every six days except duringperiods or seasons exempted by theRegional Administrator.

(c) For PMo samplers-a 24-hoursample must be taken -from midnight tomidnight (local time) to ensure nationalconsistecy. The sampling shall beconducted on the following scheduleswhich -are based on either the first yearof PM10 monitoring or a long-termselective PM10 monitoring plan:

(1) First year PM1o monitoring. Thesampling frequency for the first year (12consecutive months) of ambient PMomonitoring shall be based on adescribed area's probability ofnonattainment of the PMo NAAQSusing total suspended particulate data.Procedures to develop theseprobabilities are found in Frank, N. andT. Pace. "Procedures for EstimatingProbability of Nonattainment of a PMoNAAQS Using Total SuspendedParticulate or Inhalable ParticulateData." OAQPS, U.S. EnvironmentalProtection Agency, Research TrianglePark, N.C. September 1983. The mostrecent 3 calendar years of air- qualitydata must be used in this determination.

The probabilities are divided into threecategories: (i) High-greater than orequal to 95 percent probability;, (ii)medium-greater than or equal to 20percent to less than 95 percentprobability, and (iii) low-less than 20percent probability. A described areacould be: An urbanized area; a city ortown and; a rural area. The starting datefor this first year of PMo monitoringmay begin prior to the effective date(promulgation date) of this regulation.

(i) For high probability areas,everyday PMo sampling is required forat least one PMzo site which must belocated in the area of expectedmaximum concentration. The remainderrequire every sixth day sampling.

(ii) For medium probability areas,every other day sampling is required forat least one PMo site which must belocated in the area of expectedmaximum concentration. The remainderrequire every sixth day sampling.

(iii] For low probability areas, aminimum of one in six day sampling isrequired.If a monitoring site in a medium or lowprobability area later records levelsexceeding the short term (24-hour) PMo

NAAQS, as described in Part 50Appendix K. and the monitoringfrequency was less than everyday, theneveryday sampling must be initiated inthe area of expected maximumconcentration no later than 90daysfollowing the end of the calendarquarter in which the exceedanceoccurred and continue for thesubsequent four calendar quarters-

(2) Long term monitoring selectivesampling. After one year of PMomonitoring has been obtained. theminimum monitoring schedule for thesite in the area of expected maximumconcentration shall be based on therelative level of that monitoring siteconcentration with respect to the levelof the controlling standard. For thoseareas in which the short-term (21-hourlstandard is controlling i.e., has thehighest ratio, the selective samplingrequirements are illustrated in Figure 1.The minimum sampling schedule for allother sites in the area would be onceevery six days. For those areas in whfchthe annual standard is the controllingstandard, the minimum samplingschedule for all monitors in the areawould be once every six days.

Every Sixth Day

,1

Every Other Day

Every Day

u.0 U40 1.u J.Z I .J .4

Ratio to Standard

Figure 1. Selective Sampling Requirements

During the annual review of theSLAMS network, the most recent year ofdata must be considered to estimate theair quality status for the controlling airquality standard (24-hour or annual).Statistical models such as analysis ofconcentration frequency distributions asdescribed in "Guideline for theInterpretation of Ozone Air QualityStandards," EPA-450/479-03, U.S.Environmental Protection Agency,Research Triangle Park, N.C., January1979, should be used. Adjustments to the

monitoring schedule must be made onthe basis of the annual review. The sitehaving the concentration in the mostcurrent year must be given firstconsideration when selecting the site formore frequent sampling schedule. Otherfactors such as major change in sourcesof PM1 o emissions or in sampling sitecharacteristics could influence thelocation, of the expected maximumconcentration site. Also, the use of themost recent three years of data might insome cases, be justified in order to

i 104

, I

I,+

0 Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

provide a more representative data basefrom which to estimate current air'quality status and to provide stability tothe network. If the maximumconcentration site based on the mostcurrent year is riot selected for the morefrequent operating schedule,documentation of the justification forselection of an alternate site must besubmitted to the Regional Office forapproval during the annual reviewprocess. It should be noted thatminimum data completeness criteria,number of years of data and samplingfrequency for judging attainment of theNAAQS are discussed in Appendix.K ofPart 50.

§ 58.20 (Amended]3. Paragraph (e) of § 58.20 is amended

by adding "and for PMo and relocatedTSP monitors which must be availableby 6 months after promulgation" after"by December 1, 1981."

§ 58.23 [Amended]4. Section 58.23 is amended by adding

"with the exception of PMo samplerswhose probability of nonattainment ofthe PMo ambient standard is greaterthan or equal to 20 percent which shall'be by 1 year after promulgation and theremaining PMo and relocated TSPsamplers -which shall be by 2 years afterpromulgation" after "January 1, 1983," inthe introductory sentence.

§ 58.30 [Amended]5. Section 58.30 is amended by adding

"and PMo and relocated TSP samplers,which shall be by 6 months afterpromulgation," after "by December 1,1981" in paragraph (a).

§ 58.34 [Amended]6. Section 58.34 is amended by adding

"and PMo and relonted TSP samplers,which shall be by 1 year afterpromulgation" after "by July 1, 1982" inthe introductory sentence.

7. Section 58.35 is amended by addinga new sentence after the last sentence inparagraph (d) as follows:

§ 58.35 NAMS data submittal.

(d) • • * For PMo and relocated ornewly established TSP samplers, thefirst quarterly report will be due 90 daysafter the first quarter of operation.Appendix A-[Amended]

8. In Appendix A, sections 3, 4, and 5are amended as follows:

a. The fourth sentence in section 3.2.1Is changed by replacing the phrase "highvolume" with the phrase "high volumeand PMto."

b. Section 3.2.2 is amended byreplacing the phrase "For TSP" with thephrase "For TSP and PM1o," in thesecond sentence and by addingparagraph (e) to read as follows:

(e) PM 1oMethods. Each calendar quarter,audit the flow rate of at least 25 percent ofthe PMo samplers such that each sampler isaudited at least once per year. If there arefewer than four PM 1o samplers within areporting organization, reaudit one or morerandomly selected samplers so that onesampler is audited each calendar quarter.

Audit the flow rate of the sampler at itsspecified operating flow rate, using a certifiedflow transfer standard (see i'eference 2]. Theflow transfer standard used for the auditmust not be the same one used to calibratethe flow of the sampler being audited,although both transfer standards may bereferenced to the same primary flow ofvolume standard, The difference between theaudit flow measurement and the flowindicated by the sampler's flow indicator isused to calculate accuracy, as described insection 4.2.2.

c. In section 4.2.1, paragraph (a) isrevised to read as follows:

(a) Single Instrument Precision. For thepaired measured obtained as described insection 3.2.1, select all pairs in which bothmeasurements are aboie the concentrdtionsgiven at the end of paragraph (a). For eachselected measurement pair, calculate thepercent difference (d,) using equation 1.where Y, is the concentration of pollutantmeasured by the duplicate sampler and X, isthe concentration measured by the samplerreporting air quality for the site. For each site,calculate the quarterly average percentdifference (dj), equation 2, and the standarddeviation (Sj), equation 3.

At low concentrations, agreement betweenthe measurements of collocated samples,expressed as 95 Percent Probability Limits,may be poor. For this reason a separate count-is made of the occurrence of pollutantmeasurements below specified levels. Countthe number of data pairs from all collocated -sites that indicate Ei measurement from either

of the collocated samplers (see section 3.2.1)below the following limits:

TSP: 20 jag/ma,SO2: 45 pg/m 3.

NO2:30 jg/m3,

Pb: 0.25 psg/m3, andPM10: 20 ,g/m.

Report the counts In columns 20-23 of Form12 (Back).

d. In section 4.2.2, paragraphs (a) and(b) are revised to read as follows:

(a) Single Sampler Accuracy (TSP andPM10). For the flow rate audit described insections 3.2.2(a) and 3.2.2(a), let X, representthe known flow rate and Y, represent theindicated flow rate. Calculate the percentdifference (dl) for each audit, using equation-1.

(b) Accuracy for Reporting Orgaization(TSP andPM o). Using equation 8, calculatethe averages (D) of the individual percentdifferences for all TSP or PMto samplersaudited during the calendar quarter. Computethe standard deviation (S.) of all the percentdifferences for all of the samplers auditedduring the calendar quarter, using equation 9.

Calculate the 95 Percent Probability Limitfor the accuracy of a reporting organization,using equations 6 and 7, and record theselimits on the back of Form I under blocks 40-'51. Note that since the audit Is conducted atonly one level, blocks 40-45 and 52-57 are not

'used. For reporting organizations having fouror fewer TSP or PMo samplers, only oneaudit is required each quarter. For suchreporting organizations, the audit results oftwo consecutive quarters are required tocalculate an average and a standarddeviation, using equations 8 and 9. Therefore,semi-annual (instead of quarterly] reportingof probability limits is required.

e. In the introduction to section 5,"Environmental Monitoring and SupportLaboratory" is changed to"Environmental Monitoring SystemsLaboratory."

f. In section 5.1, delete the phrase",with a copy to EMSL/RTP".

g. In section 5.3, the instructions forForm 1, Back, are revised as follows:

5.3 Instructions for Form 1.

Block No. Oescdpion

Manual Methods (Form 1. Back)9-14............... Pollutant IdenOleGs. (Preceded).15-17.............. Number of Ssmplers Count only those samplers for each poioutant that are osxclatcd with an

approved method and from which monitoring data are reported a3 part of a SLAMS notwork,18-19 ........ Number of Collocated Sites: Number of cites having collocated samplers. The rnmi!mum number la 220-23................ Number of Collocated Data Pairs Below tho imit Count the number of data pars from the colloated

sites where a measurement from either of the collocated samplers (Soo section 3.2.1) Is below thofollowing inits:

TSP 20 pg TSP/mSO.: 45 pg SO/m'NO: 30 pg NO=lm2Pb: 0.25 pg Pb/mlPM: 20 pg PM,./ml

24-26.......... Lower Probablity im/ti Pfec&oxon Block 24 Is eilthor "+" or "-" Blocka 25-20 contaln thepercentage obtained from equation 11.

10442

Federal Resister /-Vol. 49. No. 55 / Tuesday, M~arch 20, 19341 I Proposed Rules

Block No. Descoripton

27-29 - Upper Probab:hay Lknif Prodi: Bi oc 27 is eite -+- or - os 28w-23 Wrr'l fluopercentage obtained from equar on 10. Nola If precision tirt e .vd two 6Vs. e.g. 103% recctas 99.

30-35. PoLutant Identilis: (Precoded).36 Not used.37-39 Number of Aud Count the tota] number of audits performed on tho ontao r*o rk f the r'. z -.

A single audit may consist of several audt Ievl chdks. but count oe-y oro for ea:h ar. Fcrexample, although an audit conducted for NO or SO. consis . of d-rcs at chine di-tcidconcentration levels. count this as one audl. not three.

40-57L Pobab7y Lh.ts Accmc): The lower and upper prob&I'ty lnr for cub ke of thef cv ccr yaudits are entered in blocks 40-57. Audt levels. thcir corrcspandnj raw or co.,rz--.a1 r"-.esand the appropriate bWocks for ttrs information are gren in the foniag tablo

Flow audt Antai1fc auldt

Blocks A udirei TSP Pb PMA: Pb pgftrip 3 S IZO, F31

40-45 1 ...... 10D toS3. 02 to 03.46-51 2 Normal sampler , alrr sam!cr M to 1.003 0.5 to 0.6

fla. flow.52-57 3 Norra sampler 08 to 09.

I_ I row %

'Audit ranges apply only to the Pb reference method. Audi ranges for an ooeL&u nt Pb mc.lod n-ut be ccmp=.o w'hthe specific requirements of the equivalent method.

'Applies to Pb equivalent methods which do not use the hgh vobume sampler.

Block Description

40-42,46-48.52-54-_ Lower FrobasbIy Li.t Accuraec Blocks 40. 46. and 52 we dftr " or "-. Blocks 41-AZ47-48 and 53-54 contain the percentage obtained from equation 7.

43-45,49-51. 55-57- Upper PRbabiy firsts Accrcy: Bocks 43. 49 and 55 aye clher "+" or " Bcks 44-4550-51 and 56-57 contain the perc ntsge obtaned from cquaton 6. Note: I ac--ry Er-sexceed two digits. e.g., 103%. rcport as 99. Rcport as requied, all pol .ats (TSP, Fb. PIM%.S%. and NO.) detencenod by manual ncthod3. Note that only blocks 46-51 ae used for TSPand PM,. NOTE. If on.'y one audi is performd during a ".en quarter for a g-,mn pc.ant. IIsnot possible to calculate probabW1y lim"ts for that quarter. In that ceso. Vicks 4-57 cc k1ftbank lor the fit such quarter and the rnber 001 Is reportod in b 'Xks 37-nO. PRcba:tfrlj¢ U,-3are then computded and reported on a send-annul basis (Le.. alter fte~ netrartes) from fteaudit data obtained durng the two consecutive quters.

58-60 Xwmber of Vad Co-bcatsd Data Pa.sr Enter the total nunbcr of dt Pss from a3 t cc a=tsites.

h. The Data Assessment Report Form (back). Figure 1, is revised to read asfollows:

10443Feder Repister /- Vol. 49, No. 55 / Tuesday, March 20, 1934 / Proposed Rules

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, '1984 1 Proposed Rules

DATA ASSESSMENT REPORT - MANUALMETHODS OMB No, 2000.0003REPORTING Expires 10-31-00

STATE ORGANIZATION YEAR QUARTER SEND COMPLETED FORMTO REGIONAL OFFICE

1 2 34 6- 67 a

NAME OF REPORTING ORGANIZATION

DATE SUBMITTED ---- ORIGINAL REVISION

PRECISION

NO OF NO. OFNO. OF COLLOCATED COLLOCATED

SAMPLERS' SITES SAMPLES < LIMIT

A,7SP 1,11101D -W Ei FTW9-14 15-17 18-19 20-23

S 11141214101 EIi D I]9-14 15-17 18-19 20-23

C. N02 1[14121 [01 1 E1][-"19-14 15-17 18-191 20-23

D Pb j4j112 2jO [ II I9-14 15-17 18-19 20-23

E eM-o EI' [1i- - [F-] [ [!19-14 15-17 18-19 20-23

PROBABILITY LIMITS

LOWER UPPER

24-29LOWER UPPER

24-29LOWER UPPER

W11z 1i24-29

LOWER UPPER

M IIN724-29

LOWER UPPER

FL -N t-24-29

LIMITS APPLICABLE

TO BLOCKS 20-23

TSP. 20pg TSP/m

S02 45/..g SO2/m'

N0 2 : 30pg N0 2 /m'

Pb: 0256pg Pb/rl

PM 1 O: 20pg PM 1Oml

ACCURACY

A. TSP 111111111 0 1[30-35 38

NO OFAUDITS

LWT-37-39

a 502 1114121410111 r_ -_-130-35 38 37-39

C. No 1114121B10121 ] LF30-35 36 37-39

D 11112111218 I L M30-35 38 37-39

E-,0 ,IJOIO2 [! LW30-35 36 37-39

LEVEL I

LOWER UPPER

40-45

LOWER UPPERV[_ I N ll40-45

LOWER UPPER

40-45LOWER UPPER

40.45LOWER UPPER

40 45

PROBABILITY LIMITS

LEVEL 2

LOWER UPPERM_ I FillT45-51

LOWER UPPER

46-51LOWER UPPER

46-51LOWER UPPERM I1N -II

46-51LOWER UPPER

486 51

LEVEL 3

LOWER UPPER

52 57

LOWER UPPER

52-57LOWER UPPER

52-57LOWER UPPER

62 57LOWER UPPER

62 61

'COUNT ONLY REFERENCE OR EQUIVALENT MONITORING METHODS.

FIGURE 1 FORM I IBACK)

Appendix B-[Amended]

9. Appendix B is amended as follows:

a. The heading of paragraph 3.3.1 isrevised to read as follows:

3.3.1 TSP and PM10 Methods. ***

b. The first paragraph of 3.4.1 isrevised to read as follows:

3.4.1 TSP and PMo Methods. Eachsampling quarter, audit the flow rate ofeachsampler at least qnce. Audit the flow at thenormal flow rate, using a certified flowtransfer standard (see reference 2). The flowtransfer standard used for the audit must notbe the same one used to calibrate the flow ofthe sampler being audited, although bothtransfer standards may be referenced to thesame primary flow or volume standard. Thedifference between the audit flow

measurement and the flow indicated by thesampler's flow indicator is used to calculateaccuracy, as described in paragraph 5,2.

c. Section 5.1 is revised to read asfollows:

5.1 Single Instrument Precision for TSP,Pb andPM10. Estimates of precision forambient air quality particulate measurementsare calculated from results obtained from

NO OF VALIDCOLLOCATEDDATA PAIRS

58-60

50-60

50-0

LW.5

ACCURACY

104

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

collocated samplers as described in section3.3. At the end of each sampling quarter,calculate and report a precision probabilityinterval, using weekly results from thecollocated samplers. Directions forcalculations are given below, and directionsfor reporting are given in section 6.

For the paired measurements obtained asdescribed in section 3.3.1 and 3.3.2, select allpairs in which both measurements are above20 ;ng/ml for TSP. 0.25 11g/m 3 for Pb, or 20 gg/ml for PMo. For each selected pair, calculatethe percent difference [d,} using equation 1,where Y, is the concentration of pollutantmeasured by the duplicate sampler, and X1 isthe concentration measured by the samplerrep&rting air quality for the site. Calculate thequarterly average-percent difference (djj, "equation 2; standard deviation (Sj), equation

.3; and upper and lower 95 percent probabilitylimits for precision, equations 6 and 7.Upper 95 Percent Probability .

Limit=dj+1.96Sj2 (6)Lower 95 Percent Probability

Limt=dj-I.9Sj2 (7)

d. In paragraph 5.2, revise the headingto read "Single Instrument Accuracy forTSP andPM& " and replace the phrase"each high volume sampler" with thephrase "each high-volume or PMosampler."

Appendix C--[Amendedl1 10. In Appendix C, sections 2.0, 4.0,

and 5.0 are amended as follows:.a. In section 2.0, paragraphs 2.2.1 and

2.2.2 are removed and paragraph 2.2 isrevised-to read as follows:

2.2 For purposes of showing compliancewith the NAAQS for particulate matter, thehigh volume sampler described in AppendixB of Part 50 of this chapter may be used in aSLAMS as long as the ambient concentrationof particles measured by the high volumesampler is below the PM~o NAAQS.

As soon as the TSP sampler measures asingle value which is higher than the PM~o 24-hour standard or has an annual averagegreater than the PM~o annual standard, itwould be necessary to replace the highvolume sampler designated as a substitutePM1o sampler with a PMlo sampler.

In order to maintain historical continuity ofambient particulate matter trends andpatterns, for PM,0 NAMS that werepreviously TSP NAMS, the TSP high volume'sampler must be concurrently operated withthe PMo sampler for a one-year periodbeginning with the PM~o NAMS start up date.

B. Section 4.0 is revised to read asfollows:

4.1 For short-term measurements ofPM,oduring air pollution episodes (see § 51.152 ofthis chapter) the measurement method mustbe:

4.1.1 Either the "Staggered PMo" methodor the "PMo Sampling Over Short SamplingTimes" method, both of which are based onthe reference method for PMo and aredescribed in reference 1: or

4.1.2 Any other method for measuringPMzo:4.1.2.1 Which has a measurement range or

ranges appropriate to accurately measure airpollution episode concentration of PMo.

4.1.2.2 Which has a sample periodappropriate for short-term PMsomeasurements, and

4.1.2.3. For which a quantitativerelationship to a reference or equivalentmethod for PM,o has been established at theuse site. Procedures for establishing aquantitative site-specific relationship arecontained in reference 1.

4.2 Quality Assurance. PM~o methodsother than the reference method are notcovered under the quality assessmentrequirements of Appendix A. Therefore,States must develop and implement their ownquality assessment procedures for thosemethods allowed under this section 4. Thesequality assessment procedures should besimilar or analogous to those described insection 3 of Appendix A for the PM~oreference method.

c. Section 5.1 is revised to read asfollows:

5.1 Pelton. D.J. Guideline for ParticulateEpisode Monitoring Methods, GEOMEITechnologies, Inc., Rockville, MD. Preparedfor U.S. Environmental Protection Agency,Research Triangle Park. NC. EPA ContractNo. 68-02-3584. (February 1983 Draft.)

Appendix D-[Amended]

11. Appendix D is amended asfollows:

a. In the Table of Contents, sections2.8 and 3.7 are added in the appropriateplaces as follows:

8 Di C S

2.8 PM~o Design Criteria for SLAMS

3.7 PMmo Design Criteria for NAMS

b. In section 2, in the secondparagraph of section 2.2 in the secondsentence, the words "one of four scales"are revised to read "one of five scales",and the word "micro" is inserted in theparenthetical expression immediatelybefore the word "middle." In the thirdparagraph, the word "four" Is replacedby the word "five." Following the thirdparagraph and before the discussion of"middle scale" the following Is inserted:

"- Microscale"-This scale wouldtypify areas such as downtown streetcanyons, traffic corridors, unpavedroads, haul roads, track out dirt fromconstruction sites, dust from storagepiles and fugitive emissions. Because ofthe very steep ambient TSP gradientsresulting from these sources, thedimensions of the microscale for TSPgenerally would not extend beyond 15meters. In the case of roadway sources,the microscale could continue the lengthof the roadway, which may be severalkilometers. Microscale TSP sites should

be located near inhabited buildingswhere property can be expected to beexposed to high ambient particulateconcentrations. Emissions fromstationary sources such as primary andsecondary smelters, power plants, steelmills and other large industrialprocesses may. under certain plumeconditions, likewise result in highground level concentrations at themicroscale. In the latter case, themicroscale would represent an areaimpacted by the plume with dimensionsextending up to approximately 100meters. Data collected at microscalestations provide information forevaluating and developing "hot spot"control measures.

The first 3 sentences of the"- middlescale"-discussion aredeleted and replaced by "the previouslymentioned sources also have an impactat the middlescale level."

In the "o neighborhood scale"-discussion, the remainder of the 4thsentence after the word "conditions" isdeleted and replaced by "to which alarge part of the city is subjected"

c. In section 2, a new section 2.8 isadded as follows:

2. SLAMS Network Design Procedure

2.8 PM10 Desin crteria for SLAMS.As with other pollutants measured in the

SLAMS network, the first step in designingthe PMo network is to collect the necessarybackground information. Various studies 11,12,13.14.15 have documented the majorsource categories of particulate matter andtheir contribution to ambient levels in variouslocations throughout the country. Because thesources for PM,o are similar to those for TSP,the procedures for collecting the necessarybackground information for PIMo are similarto those described in section 2.2 for TotalSuspended Particulates. After completing thefirst step, existing TSP SLAMS or otherparticulate matter stations should beevaluated to determine their potential ascandidates for SLAMS designation. Stationsmeeting one or more of the four basicmonitoring objectives described in section 1of this Appendix must be classified into oneof the five scales of representativeness(micro, middle, neighborhood, urban andregional) if the stations are to becomeSLAMS. In sitting and classifying PMzostations, the procedures described inreference 16 should be used.

If existing TSP samplers meet the qualityassurance requirements of Appendix A, thesiting requirements of Appensix E. and arelocated in areas of suspected maximumconcentrations as discribed in sectioi3 ofAppendix D. and if the TSP Levels are belo,the ambient PMi standards, TSP samplersmay continue to be used as substitutes forPM,o SLAMS samplers under the provisionsof Section 2.2 of Appendix C.

10445

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

The most important spatial scales toeffectively characterize the emissions of PM1ofrom both mobile and stationary sources arethe micro, middle and neighborhood scales.For purposes of estabishing monitoringstations to represent large homogenous areasother than the above scales ofrepresentativeness, urban or regional scalestations would also be'needed.

0 Microscae-This scale would typifyareas such as downtown street canyons andtraffic corridors where the general publicwould be exposed to maximumconcentrations from mobile sources. Becauseof the very steep ambient PMo gradientsresulting from mobile sources, the dimensionsof the microscale for PM1o generally wouldnot extend beyond 15 meters from theroadway, but could continue the length of theroadway which could be several kilometers.Microscale PMo sites should be located nearInhabited buildings or locations where thegeneral public can be expected to be -exposedto the concentration measured. Emissionsfrom stationary sources such as primary andsecondary smelters, power plants, and otherlarge industrial processes may, under certainplume conditions, likewise result in highground level concentrations at themicroscale. In the latter case, the microscalewould represent an area impacted by theplume with dimensions extending up toapproximately 100 meters. Data collected atmicroscale stations provide information forevaluating and developing "hotspot" controlmeasures.

* Middle Scale--Much of the measurementof short-term public exposure to PMo is onthis scale. People moving through downtownareas, or living near major roadways,encounter particles that would be adequatelycharacterized by measurements of thisspatial scale. Thus, measurements of thistype would be appropriate for the evaluationof possible short-term public health effects ofparticulate matter pollution. This scale alsoincludes the characteristic concentrations forother areas with dimensions of a fewhundred meters such as the parking lot andfeeder streets associated with shoppingcenters, stadia, and office buildings. In thecase of PM10, unpaved or seldom sweptparking lots associated with these sourcescould be an important source in addition tothe vehicular emissions themselves.

* NeighborhoodScale-Measurements inthis category would represent conditionsthroughout some reasonably homogeneousurban subregion with dimensions of a fewkilometers and of generally more regularshape than the middle scale. Homogeneityrefers to the PM,o concentrations, as well asthe land use and land surface characteristics.In some cases, a location carefully chosen toprovide neighborhood scale data wouldrepresent not only the immediateneighborhood but also neighborhoods of thesame type in other parts of the city. Stationsof this kind provide good information abouttrends and compliance with standardsbecause they often represent cdnditions inareas where people commonly live and workfor period comparable to those specified inthe NAAQS. This category also includesindustrial and commercial neighborhoods, aswell as residential

Neighborhood scale data could providevaluable information for developing, testing,and revising models that describe the larger-scale concentration patterns, especially those

models relying on spatially smoothedemissions fields for inputs. The neighborhoodscale measurements could also be used forneighborhood comparisons within or betweencities. This is the most likely scale ofmeasurements to meet the needs of planners.

e Urban Scale-This class of measurementwould be made to characterize the PM,,concentration over an entire metropolitanarea. Such measurements would be useful forassessing trends in city-wide air quality, andhence, the effectiveness of large scale airpollution control strategies.

* RegionalScale-These measurementswould characterize conditions over areaswith dimensions of as much as hundreds ofkilometers. As noted earlier, usingrepresentative conditions for an area impliessome degree of homogeneity in that area. Forthis reason, regional scale measurementswould be-most applicable to sparselypopulated areas with reasonably uniformground cover. Data characteristics of thisscale would provide information about largerscale processes of PM.. emissions, losses andtransport.

d. In section 3, the third paragraph isrevised to read as follows:

3. Network Design for National AirMonitoring Stations (NAMS)

Category (a]: stations located in area(s) ofexpected maximum concentrations (generallymicroscale for CO. microscale or middlescale for TSP, Pb and PM0,, neighborhoodscale for SO, and NO, and urban scale for0.

e. In Section 3.1 in the secondsentence, "500,000" is replaced by"1,000,000,,, the word "primary" isreplaced by the word "secondary," andthe number "8" is replaced by thenumber "10" Table 2 is revised aafollows:

In the first sentence after Table 2, thephrase "600 to 700" is replaced by "300to 450."

The rest of the first paragraph afterTable 2 and the second and thirdparagraphs after Table 2 are removedand are replaced by "this range ofmonitors is believed to be sufficient toprovide a national overview withrespect to the welfare effects associatedwith the secondary TSP NAAQS inurban areas of 100,000 population orgreater."

f. In Section 3.2, the phrase "As withTSP monitoring" at the beginning of thefirst paragraph Is removed and thesentence begins with the next word It."-The second, third, fourth and fifthsentences in the second paragraph areremoved and replaced with "Thisnumber of NAMS S02 monitors Issufficient for national trend purposesdue to the low background SOz levels,and the fact that air quality Is verysensitive to SO emission changes.

g. A new Section 3.7 is added as setforth below,

3.7 PM1o, Design Criteria for NAMS.Table 4 indicates the approximate number

of permanent stations required in urban areasto characterize national and regional PMI, airquality trends and geographical patterns. Thenumber orstations In areas where urbanpopulations exceed 1,000,000 must be In therange from 2 to 10 stations, while in lowpopulation urban areas, no more than twostations are required. A range of monitoringstations Is specified In Table 4 becausesources of pollutants and local control effortscan vary from one part of the country toanother and therefore, some flexibility Isallowed in selecting the actual number ofstations in any one locale.

TABLE 2.-TSP NATIONAL AIR MONITORING STATION CRITERIA

CAPPROXIMATE NUMBER OF STATIONS PER AREA]'

Population aeoy High ModOum Lowconcentrationb concentration' doncontationd

> 1,00.000 6-10 4-0 2-4500,00-1.00,000 . . . .4-0 2-4 1-2250.000-500,000... ..... 3.... 5-4 1-2 0-110D.000-250.000- . .1-2 0-1 0

'Selection of urban and actual number of stations per area vw,! be jontly determmnd by EPA and tho Stato eflcncy.*High concentration-exceedmg level of the secondary NAAQS by 20 percent or more+* Medium concentration-exceeding secondary NAAOS.I Low concentration-less than secondary NAAQS.

It is recognized that no PM.o samplers willbe designated as PM.o reference orequivalent methods until, at the earliest,approximately six months after promulgationof PMo NAAQS and the reference andequivalent method requirements. Even thoughnon-designated PMio samplers will have beencommercially available, and a small numberof samplers will have been in use by EPA.other agencies, and industry, there will not be

,enough ambient PM.o data to determineambient PMo levels for all areas of thecountry. Accordingly, EPA has providedguidance 17 on converting ambient IP. data toambient PM,, data. Ambient IPs, data aredata from high volume samplers utilizingquartz filters or dichotomous samplers, bothwith inlets designed to collect particles

nominally 15 pm and below. Also Included inthe guidance are procedures for calculatingfrom ambient TSP data the probability thatan area will be nonattainment for PM.O. Fordetermining the appropriate number ofNAMS per area, the converted IP, data orthe probabilities of PMo nonattalnment areused in Table 4, unless ambient PMo data areavailable. If only one monitor Is required Inan urbanized area, it must be a category (a)type. If an evaluation of the sources of PM1oas described in section 2.8 indicates that themaximum concentration area Ispredominantly influenced by roadwayemissions, then the category (a) stationshould be located adjacent to a major roadand should be a microscale or middle scale.A microscale is preferable but a middle scale

10446

Federal Register f Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules7

is also acceptable if a suitable microscalelocation cannot be found, However, if thepredominant influence in the suspectedmaximum concentration area is expected tobe industrial emissions, and/or combustionproducts (from oth6r than an isolated singlesource), the-category (a)lstation should be amiddle scale or neighborhood scale. A middlescale exposure is preferable to aneighborhood scale in representing themaximum concentration impact from multiplesources, other than vehicular, but aneighborhood scale is acceptable, especiallyin large residential areas that bum oil, wood,and/or coal for space heating.

For those cases where morb than onestation is required for-an urban area; theseshould be at least one station for category (a)and one station for category (b) neighborhood

TABLE 4.-PMo National Air Monitoring Station CriterialApproximate number of stators per area]

Population category Kah Mc:.Um LOWconcentraions-~ c===otta~n- ccnra~n"

>1.000.000 6-10 4-a 2-4500.000 to 1.000,000 4-8 2-4 1-2250.000 to 5D0,000 3-4 1-2 0-1100.000 to 250.000 1-2 0-1 0

"Seection of urban areas and actual number of stations per agrea wZI be loly dctermined by EPA " ft Sth - y.bHigh concentration areas are those for ,"ch: Arbient PM* data onsent [P. data Covrtn-d to FIA. 5112w arr4a"nl

concentrations exceeding either PMo NAAOS by 20 percent or more; or the pcba_ 1y of PM. noaa,'=,44. c:'c!u: d ft=TSP data, is 95 percent or greater.

MedUm. concentration areas are those for which: Ambient PMo data or amb;nt IPs data ccnert,2d to PR.1 ehew an-± reconcentrations exceeding either 80 percent of the PMi. NAAOS; or the prob'!"ty of PM. nonattaa'nnent. ce=.,-d ftrcm TSPdata, is >20 percent and <95"percent

Low concentration areas ar those for which: Ambient PM, data or ambient IP,. data coe~rted to PMK 0,4-w 'r*n;concentrations less than 80 percent of the PM,. NAAQS; or the probab ty of PM,. no tir.nTnt. eae u d from TSP data.is less than 20 percent.

* Procedures-for.esb r -ambient Mo-concentrations from IPt, a a.et ns or forst.,r.,ag the rcb.., y.f nonattamment for PM, gven observed TSP data are provided In reterenco 17.

h. In section4, Table 4 is renumbered Table 5 and is revised to include TSPand PMo as follows:

TABLE 5.-:-SUMMARY OF SPATIAL SCALES FOR SLAMS AND REQUIRED SCALES FOR NAtIS.

5p~~ Scales a;0ppiable for SLAMS Scaes mrked for NASscale TSP SO. CO 0. -NO, Pb PM,. TSP SO, CO o NO, Pb FM.

.icrO. V/ ..... V .V V Vv-jidc ie V V V. V V V I I V

Neighbor-V V V VV VN V V IVVhood.

Urban " V V . V VRegionl.. V V

i. In section 5; the list of references isamended by adding references 1ithrough 17 as follows:.

5. References.

11. Cooper. TA. etal. Summary of thePortlandAerosol Characterization Study.(Presented at the 1979 Annual Air PollutionAssociation Meeting, Cincinnati, OH. APCA#79-2,.41.

12. Bradcway, 1KM. and FA. Record..NationalAssessment of the Urban ParticulateProblem. GCA Technology Division. Bedford,

'. MA. Prepared for U.S. EnvironmentalProtection Agency, Research Triangle Park,

NC. EPA Publication No. EPA-450/3-76-035.July 1978.

13. U.S. Environmental Protection Agency,Air Quality Criteria for Particulate Matterand Sulfur Oxides, Volume 2. EnvironmentalCriteria and Assessment Office.-ResearchTriangle Park. NC. December 1981.

14. Watson. J.G. J.C. Chow, and J.J. Shaw.Analysis of Inhalable and Fine ParticulateMatter Measurements. EnvironmentaLResearch and Technology Inc. Concord. MA.Prepared for U.S. Environmental ProtectionAgency. Research Triangle Part. NC. EPAContract No. 68-02-2542. March 1982 draf

15, Record. F.A. and LA. BacLEvaluationof Contribution of Wind Blown Dust from. theDesert Levels of Particulate Matter in DesertCommunities. GCA Technology Division.

scale obiectivesas discussed in Section 3.Where three or more stations are required.the mix of category (a) and (b) stations is tobe determined on a case-by-case basis. Theactual number of NAMS and their locationsmust be determined by EPA Regional Officesand the Stateagencies, subject to theapproval of the Administrator as required by§ 58.32. The Administrator's approval Isnecessary to insure that individual stationsconform to the NAMS selection criteria andthat the network as a whole Is sufficient Interms of number and location for purposes ofnational analyses. As required under theprovisions of section 2.2 of Appendix C. allPMo NAMS that were previously designatedas TSP NAMS must concurrently collectambientTSP and PMao data for a one-yearperiod beginning when each NAMS PMIosampler is put into operation.

Bedford. MA. Prepared for U.S.Environmental Protection Agency. ResearchTriangle Park. NC. EPA Publication No. EPA-450/2-80-078. August 1980.

18. Koch. P.C. and H.E. Rector. OptimumNetwork Design and Site Exposure Criteria -for Particulate Matter. GEOME"Technologies, Inc.. Rockvile. MD. Preparedfor U.S. Environmental Protection Agency.Research Triangle Park NC. EPA ContracfNo. 68-02-3584. March 1983.

17. Frank, N. and T. Pace. Procedures forEstimating Probability of Nonattainment of aPMto NAAQS Using Total SuspendedParticulate or Inhalable Particulate Data.OAQPS, US. Environmental ProtectionAgency. Research Triangle Park. NC.February 1934.

Appendix E-[Amended]

12. Appendix E is amended as follows:

a. The Table of Contents is amendedby adding a new section 8 andrenumbering the original sections Ithrough11 as sections 9 through 12 asfollows:

8. Particulate Matter (PMI)8.1 Vertical Placement8.2 Spacing from Obstructions8.3 Spacing from Roadways8.4 Other considerations9. Probe Material and Pollutant Sample

Residence Timer10. Waiver Provisions11. Discussion and Summary12. References

b. In section 1, the last sentence of thesecond paragraplr is amended bychanging the term "section 9" to"osection 10.-

c. In section 2.1. the first sentence isremoved.

d. In section 2.3, the fourth and fifthsentences are removed. In the firstsentence of the second paragraph. thelast word "diminished" is replacecbythe word "enhanced" and the remainderof section2.3 is revised to read:'Todetermine the impact of TSP from motorvehicles iMis desirable for NAMS andSLAMS category 'a' monitors to belocated in the enhanced portion of theplume. For neighborhood or largerscalesites, they should be located beyond theconcentrated particulate plumegenerated by traffic and not so closethat the roadway totally dominates themeasured ambient concefitration. FigureI shows the location requirements for,TSP monitors with respect to roadwaysof 3000 vehiclesper day or greater. Themicroscale sitemust be between'5 and15 meters from the nearesttraffic laneand between 2 and 7meters inelevation. Setback distances and.vertical placement of the sampler inletfor middle scale and neighborhood scalesites are also showrimFigure1:'

10447

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

e. The text of section 2.4 is revised toread: "In order to minimize the impact ofwind blown dusts, stations should not

20

Ln

4-.g.)15

4J

4J

10

0.

E

be located on bare ground. Additionalinformation on TSP probe siting may befound in reference 10."

f. Figure 1 in section 2 is revised by anew Figure I as shown.

5 10 15 20 25 30 35Distance from Edge of Nearest Traffic Lane, Metersa

aApplies where ADT > 3000

Figure 1. Acceptable Areas for TSP Micro, Middle and Neighborhood Scale Monitors

g. Section 8 is revised to read as•follows:

8. Particulate Matter (PMo).8.1 Vertical Placement-Although there

are limited studies on the PMo concentrationgradients -around roadways or other groundlevel sources, References 1, 2,4, 18 and 19 ofthis Appendix show a distinct variation in thedistribution of TSP and Pb levels nearroadways. TSP, which-is greatly affected bygravity, has large concentration gradients,both horizontal and vertical, immediatelyadjacent to roads. Lead, being predominatelysub-micron in size, behaves more like a gasand exhibits smaller vertical and horizontalgradients than TSP. PMko, being intermediatein size between these two extremes exhibitsdispersion properties of both gas andsettleable particulates and does showvertical and horizontal gradients.30 Similar tomonitoring for other pollutants, optimalplacement of the sampler inlet for PMomonitoring should be at breathing heightlevel. However, practical factors such asprevention of vandalism, security, and safetyprecautions must also be considered when

siting a PM1o monitor. Given theseconsiderations, the sampler inlet formicroscale PMio monitors must be 2-7 metersabove ground level. The lower limit wasbased on a compromise between ease ofservicing the sampler and the desire to avoidre-entrainment from dusty surfaces. Theupper limit represents a compromise betweenthe desire to have measurements which aremost representative of population exposuresand a consideration of the practical factorsnoted above.. For middle or larger spatial scales,increased diffusion results in verticalconcentration gradients that are not as greatas for the microscale. Thus, the requiredheight of the air intake for middle or largerscales is 2-15 meters.

8.2 Spacing from Obstructions-If thesampler is located on a roof or otherstructure, then there must be a minimum of 2meters separation from walls, parapets,penthouses, etc. No furnace or incinerationflues should be nearby. This separationdistance from flues is dependent on theheight of the flues, type of waste or fuel

burned, and quality of the fuel (ash content).In the case of emissions from a chimneyresulting from natural gas combustion, as aprecautionary measure, the sampler shouldbe placed at least 5 meters from the chimney.

On the other hand, if fuel oil, coal, or solidwaste is burned and the stack is sufficientlyshort so that the plume could reasonably be,expected to impact on the sampler Intake asignificant part of the time. other buildings/locations in the area that are free from thesetypes of sources should be considered forsampling. Trees provide surfaces forparticulate deposition and also restrictairflow. Therefore, the sampler should'beplace at least 20 meters from trees.

The sampler must also be located awayfrom obstacles such as buildings, so that thedistance between obstacles and the sampleris at least twice the height that the obstacleprotrudes above the samples. Samplingstations that are located closer to obstaclesthan this criterion allows should not beclassified as neighborhood, urban, or regionalscale, since the measurements from such istation wduld closely represent middle scale

10448

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

stations. Therefore, stations not meeting thecriterion should be classified as middle scale.

There must be unrestricted airflow in anarc of at leasL270 around the sampler. Sincethe intent of the category- (a) site is tomeasure the maximuxe concentrations from aroad or point source, there must be nosignificant obstruction between a road orpoint source and the monitor, even thoughother spacing from obstruction criteria aremet. The predominant direction for theseason with the greatest pollutantconcentration potential must be included inthe 270* arc.

8.3 Spacng from Roads. Since emissionsassociated with the operation of motorvehicles contribute to urban area particulate

matter ambient levels, spacing from roadwaycriteria are necessary for ensuring nationalconsistency in PMlo sampler siting.

The intent is to locate category (a) NAMSsites in areas of highest concentrationwhether it be from mobile or multiplestationary sources. If the areas Is primarilyaffected by mobile sources, then the monitorsshould be located near roadways with thehighest traffic volume and at separationdistances most likely to produce the highestconcentrations. For the microscale station.the location must be between 5 and 15 metersfrom the major roadway. For the mlddle scalestation, a range of acceptable distances fromthe roadway is shown in Figure 2. This figurealso includes separation distances between a

roadway and neighborhood or larger scalestations by default. Any station. 2 to 5 metershigh. and further back than the middle scalerequirements will generally be neighborhood.urban or regional scale. For example.according to Figure 2. if a PMia sampler isprimarily influenced by roadway emissionsand that sampler is setback 10 meters from a30,000 ADT road. the station should beclassified as a micro scale, if the samplerheight is between 2 and 7 meters. If thesampler height is between 7 and 15 meters,the station should be classified as middlescale. If the sampler is 20 meters from: thesame road. it vrill be classified as middlescale; iff40 meters., neighborhood scale: and if110 meters an urban scale.

Figure 2. Acceptable Areas for PH10 Micro, Middle, Ileighborhood, and Urban Sam-plers

DISTANCE FROM NEAREST TRAFFIC LANE. n

10449

Federal Register / Vol. 49. No. 55 / TuesdRv March 9n la/A I ..... ,-I,,,- - .1'F Q LA = 'V' aU

It is impqrtant to note that the separationdistances shown in Figure 2 are measuredfrom the edge of the nearest traffic lane of theroadway presumed to have the mostinfluence on the site. In general, thispresumption is an oversimplification of theusual urban settings which normally haveseveral streets that impact a given site. Theeffects of surrounding streets, wind speed,wind direction and topography should beconsidered along with Figure 2 before a finaldecision is made on the most appropriatespatial scale assigned to the sampling station.

8.4 Other Considerations. For these areasthat are primarily influenced by stationarysource emissions as opposed to roadwayemissions, guidance in locating these areasmay be found in the guideline documentOptimum Network Design and Site ExposureCriteria for Particulate Matter.29

Stations should not be located in anunpaved area unless there is vegetativeground cover year round, so that the impactof wind blown dusts will be kept to a.minimum.

h. The original section 8 "ProbeMaterial and Pollutant SampleResidence Time" is redesignated assection 9.

i. The orignial section 9 "WaiverProvisions" is redesignated as section10.

j. The original section 10 "Discussionand Summary" is redesignated assection 11; the Table 5 therein is revisedto read as follows:

11. Discussion and Summary

TABLE 5.-SUMMARY OF PROBE SITING CRITERIA

Distance from-Height above supporting structure, Other spcingPollutant Scale ground. meters Cerameters cr tla

SVertical IHorizontal-FI 4

Microscal6................

Middle, neighborhood, urbanand regional scale.

1. Should be >20meters from tvs.

2. D:stance fromsampler toobstacle, such 0abuildings, must beat teast tia theheight theobatacto protrudosabove the-samplorb

3. Must haveunte,-trictod ailow270' around thosamplcr,

4. No furnace orIncineration fluoshould be neabty.'

5. Must havemrnumum spaclngfrom roads.

1, Should be >20motors from trees,

2 Distance fromsampler toobstacle, such asbuildfngo, must beat teast t Wco theheight theobstaclo protrudesabove thesamper., b

3. Must haveunrestricted a!rte270' around thosampler,

4, No furnace ora Incnoration fluos

srhould be noby'5. Must have

minimum spacingfrom roads, ThI3varies with spatilcae (ca Figure

1)1, Should be >20motors from tres.

2 Distance fromInlet probe toobstacle. such abuild;ng3, mut beat least twice thehoIght thoobstacle protrudesabove the Intelprobeob

3. Must haveunrostlcted alilow270' around theInlet probe, or180' if probe Is onthe vkfo of abuilding.

4. No umanco orIncineration flueoshould be raarby'

10450

So= ......... ...... Anl ..... ............... . ..... 3 to1 . . ..

2 to 15 ...........

Federal Register*/ Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules i 10451

TABLE 5.-SUMMARY OF PROBE SMNG CRrTERIA-Continued

D1'co fromPotj.ant Scale 4god sc r .:f €fe

rmelcm-

Co I M.cro 3± %

Midde ncighborhood 3 to 15

NO, AN _ 3 to 15 -

1. JW-t be >10m6-ers ti=m ='OM

KWztectzn ardsht', be at a

Z M, A to 2-C0m'eters from e'eof ncati traf

3. Must h.reo

Inet prebo.

,armvfred aew27W aomd =~Wet prote. cc

ve .6 o a a

2. S;&%g fremmad$s varme %4traffic (tce Tat!*1).

1. Stmbld be >20Wes.s fra tiees

2. D0cc Irwet po, toobs±l. such ashl* Gs M4st beat kast too T

hcis ft -

Cbstade txoenatcabo em rect

3. Must hIvo

270 .rwrd CAJelt ¢ccbe. o

I SW dt probe Es onft Weo el a

4. SN fn omrcdse v~c5sntfr~!t (See Tato2).

Ineers from trees2. D-Estsi0 from

wet Pete tocbs such asWbti-M. MCIu be

at kca-A twO t

cbs'Ude MhdcsU

probo.h3. Must trve

=vVYiesicd a~rrvw270* ammd telrtd pecb. or1801 1 cbe is cn

4. Spac153 U"o

tral~.c (coo Tablo

1. Stm%4d be >20,nccy from trees

2. toaso romM3*T:;!e tocbWv such ashstfnV mast beat least W06c 1:111

ebstact. peatnsdIs

3. Must twosuvstimed aerscw270* arcund te

Street Carimc

N. o fsrice cchairaton &*$VnWr be rmcaty.

Ant 3 to 15

Pb Micro 2 to 7 -

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

TABLE 5.-SUMMARY OF PROBE SITING CRITERIA-ContinuedDistance from

He'ght above supporting structure,Pollutant Scale gVou meters ootmeters - t6.a

Vertical Horizontal-

Middle, neighborhood, urban 2 to 15 -4and regional.

Middlle, neighborhood,and regional scale.

urban 12 to

5. Must be 5 to 15meters from ma]orroadway.

1. Should be >20meters from tress.

2. Distance fromsampler toobstacle, such asbuildings must beat least twice theheight theobstacle protrudeabove thesampler.

3. Must haveunrestricted airflow270' around thesampler.

4. No furnace orincineration fluesshould be nearby.-

5. Spacing fromroads varies withtraffic (see Table4).

1. Should be >20meters from trees.

2. Distance fromsampler toobstacle, such asbuidirngs, must betwice the heightthe obstacleprotrudes abovethe sampler.

3. Must haveunrestricted airflow270' around thesampler except forstreet canyonsites.

4. No furnace orincineration fluesshould be nearby.

5. Spacing fromroads varies withtraffic (see Figure2).

1. Should be >20meters from trees.

2. Distance fromsampler toobstacle, such asbuildings, must beat least twice theheight theobstacle protrudesabove thesampler.

3. Must haveunrestricted airflow270' around thesampler.

4. No furnace orIncineration fluesshould be nearby.

5. Spacing fcomroads varies withtraffic (see Figure2).

When probe Is located on rooftop, this separation distance is in reference to walls, parapets, or penthouses located onthe roof.

bSites not meeting this criterion would be classified as middle scale (see text).Distance Is dependent on heisht of furnace or incineration flues, type of fuel or waste burned, and quality of fuel(sulfur. ash or lead cantent). This as to avoid undue influences from mnor pollutant sources.

k. The original section 11 [References]is redesignated as section 12; and thelist of references is amended by addingreferences 29 and 30 as follows:

12. References.

29. Koch, R. C. and H. E. Rector. OptimumNetwork Design and Site Exposure Criteriafor Particulate Matter. GEOMETTechnologies, Inc., Rockville, MD. Preparedfor U.S. Environmental Protection Agency,Research Triangle Park, N.C. EPA ContractNo. 68-02-3584. (March 1983.)

30. Burton, R. M. and J. C. Suggs.Distribution of Particulate Matter From theRoadway of a Philadelphia Site.Environmental Monitoring SystemsLaboratory, U.S. Environmental ProteoctionAgency, Research Triangle Park, N.C.(September 1983 Draft).

Appendix F--[Amended]

13. Appendix F is amended as follows:

a. The following is added to the end ofthe table of contents:

2.7 Particulater Matter (PMo)2.7.1 Site and Monitoring Information2.7.2 Annual Summary Statistics

b. In section 2.2, the title is revised,subparagraph 2.2.2 is revised, andsubparagraph 2.2.3 is added to read asfollows:

2.2 Total Suspended Particulates l(TSP)

2.2.2 Annual Summary Statistics. Annualarithmetic mean (pg/m 10] as specified InAppendix K of Part 50. Daily TSP valuesexceeding the level of the'24-hour PM1

NAAQS and dates of occurrence. If morethan 10 occurrences, list only the 10 highestdaily values. Sampling schedule used ouch asone every six days, one every three days, etc.Number of additional sampling days beyondsampling schedule used. Number of 24-houraverage concentrations in ranges:

Range Numberof Valuo

0 to 60 (pg/m) ... ........... ........................61 to 120.-..o ... ....... ......... ............... ... ..........

121 to 300 ....... .............. ......181 to 24D0 -. -.------.......... ................... ...... .241 to 300 ... ... ..................... ........ ,

301 to 360 ............. ...... ..... ...............361 to 420 ... ...... . . ....... ......Greater tutn 420 ........ .................. .... ..... ..........

2.2.3 Episode and Other UnscheduledSampling Data. List episode measurements,other unscheduled sampling data, and datesof occurrence. List the regularly scheduledsample measurements and date of occurrencethat preceded the episode or unscheduledmeasurement.

c. Section 2.7 is added to read asfollows:

2.7 Particulate Matter (PMo)2.7.1 Site and Monitoring Information.

City name (when applicable), county name,and street address of site location. SAROADsite code. Number of daily observations.

2.7.2 Annual Summary Statistics. Annualarithmetic mean (ttg/m) as specified InAppendix K of Part 50. All daily PMlo valuesabove the level of the 24-hour PMto NAAQSand dates of occurrence. Sampling scheduleused such as once every six days, once everythree days; etc. Number of additionalsampling days beyond sampling scheduleused. Number of 24-hour averageconcentrations in ranges:

10452

PI fo ................. .. .. .... Micr ..... ............. ... 2 to 7. ._. . . . .

Federal Register I Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

NumberRange of vaues

O to 30 (gfm31 to-6061 to9091 to 120121 t1150151 to.180181 to 210Greater than 210

2.72.3 Episode and Other UnscheduledSampling Data. List episode measurements,other unscheduled sampling data and datesof occurrence. List the regularly scheduledsample measurement and date of occurrencethat preceded the episode or unscheduledmeasurement

Appendix G-[Amended]

14. Appendix G is amended asfollows:

a. Paragraph 2.f. is revised to read asfollows:

2. Definitions.

f. "Critical pollutant" means the pollutantwith the highest subindex during thereporting period.

b. In the first paragraph of section 3,the term "TSP" is removed in the third

sentence and replaced by the words"PM1o", the word "particulate" isremoved and replaced with "PMzo" inthe fourth sentence, and the term "hi-volume" is removed twice in the fourthsentence and replaced by the words"reference or equivalent method."

c. In section 7, the words "totalsuspended particulates (TSP)" aredeleted in the second sentence of thefirst paragraph and replaced by thewords "particulate matter (PM10)," thefirst sentence of the second paragraph isremoved and the word "six" in thesecond sentence of the secondpaiagraph is replaced by the word"five".

d. Section 7.2 is revised to read asfollows:

7.2 Example Computation.Suppose a PMio 24-hour concentration of

283 jig/m is observed. The PM,0 subindex Iscalculatd using equation 1 as follows: InTable 1, the observed concentration ofX1=283 mg/Wm lies between 180' and 350'pg/m3, therefore this computation is carriedout for the second segment (j=2). For thissegment, X, 1=180 and X,.,=350. withcorresponding subindex values for I.2=100and I.3=200. The computation Is as follows:

Ir-I=,2 200-100 100It= (283-X,.2)+I= - (283-180)+100= - X103+100=161

350-180 170

Therefore, the PM,0 subindex is 1=161. If fourother pollutant subindices calculated in asimilar manner from observations on thesame data were: h,=0, I=0, I4=0, and Is=0,then the overall index is reported as themaximum of these values:

PSI-max161,0,0,20,30)=161

A typical report might contain thefollowing statement- "Today's air qualityindex is 161 which is regarded as unhealthful.The responsible pollutant is particulatematter. This report represents conditionsprevailing over most of the downtown urbanarea for the previous 24-hour period ending atnoon today." If the index were forecast forthe next day, the following additionallanguate might also be used: "The currentforecast is for improved air quality tomorrowwith the index not expected to exceed 80."

e. In Table 1, in the sixth columnentitled 1-hr. 03. the number 118 isremoved and replaced with 120, the term"mg/m " is removed from the heading

of the fifth column and replaced with"mg/m'," the fourth column entitledTSPxSO (jtg/ms2) is removed, and thesecond column is revised to read asfollows:

Table 1.-Breakpoints for PSI in MetricUAitsI

. . . 411I44M . . .gtmv155

'350'420.500'603

SASt the conccn"atn kvc.s am Vrd fc? ntra.-'opurp~s orrly. Tho er1t.l Imecs w.1 be d~rrX- at fttL" of prorm jaton o! to st=zrd=,

f. In Table 2, the third column entitledTSP x SO 2(fg/mrXppm) is removed.

-The levels are used for illustrative purposesonly. The actual level will be determined at the timeof promulgation or the standard.

10453

Federal Register / Vol. 49, No. 55 / Tuesday, March 20, 1984 / Proposed Rules

g. Figure 2 is revised to read as follows:

Figure 2. PSI function for suspended particulate matter, PM10

400

300

200

U iuu ZUU 3UU 40U 5U0 6UU

PARTICULATE MATTER PM10

(24-hour Running Average, ug/m3)

*All of the values used are for illustrative purposes only and will bereplaced with the appropriate air quality standards, federal episodelevels, and significant harm level at the time of promulgation.

l h. Figure 6 is removed.[IR Doe. 84-0802 Filed 3-19-84: 845 ami

BILLING CODE 6560-20-M

40 CFR Part 53

[AD-FRL 2491-7]

Ambient Air Monitoring Reference andEquivalent Methods

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

SUMMARY: Elsewhere in this issue of theFederal Register the U.S. EnvironmentalProtection Agency (EPA) is proposing

revisions to the national ambient airquality standards for particulate miatterand is also proposing a new referencemethod for the determination ofatmospheric concentrations of PM1o, aproposed new indicator for particulatematter. Since the proposed referencemethod includes a sampler that wouldbe specified primarily by performance,EPA, in conjunction with that action, isherein proposing performancespecifications, explicit test procedures,and other requirements applicable toreference and equivalent methods forPMo. These PMo specifications and testprocedures are analogdus to existingspecifications and test procedures forreference and equivalent methods for

US- ignircant Harm uu-I Level

I

i- / 20"*

II- 50*

18 PRIMARY NAAQS

5 ANNUAL PRIMARY NAAQS

I I

other criteria pollutants contained in 40CFR Part 53. Therefore, EPA isproposing appropriate revisions to addthe new PMo requirements to Part 53.EPA is also proposing some minorclarifications to existing provisions ofPart 53, pertaining to the otherpollutants for which ambient air qualitystandards exist.DATE: Comments must be received on orbefore May 21, 1984.ADDRESS: Comments, preferably Induplicate, should be sent to PublicDocket No. A-82-43, U.S. EnvironmentalProtection Agency, Central DocketSection (A-130), West Tower Lobby,Gallery , 401 M Street, S.W.,Washington, DC 20460. The docket maybe inspected at this address between thehours of 8:00 a.m. and 4:00 p.m., Mondaythrough Friday. A reasonable fee may becharged for copying services.FOR FURTHER INFORMATION CONTACT.Larry J. Purdue, Chief, MethodsStandardization Branch (MD-77),Quality Assurance Division,Environmental Monitoring SystemsLaboratory, U.S. EnvironmentalProtection Agency, Research TrianglePark, North Carolina 27711 (919-541-2665).SUPPLEMENTARY INFORMATION:

BackgroundElsewhere in this issue of the Federal

Register, EPA is proposing to revise thenational ambient air quality standards(NAAQS) for particulate matter (40 CFRPart 50) and is proposing a newreference method (Appendix J) for thedetermination of ambient concentrationsof particulate matter with anaerodynamic diameter less than or equalto a nominal 10 micrometers (PMlo).

Similar to the current manualreference methods for Total SuspendedParticulate Matter (TSP) and lead, theproposed PMio reference method wouldrequire a sampler for collectingparticulate samples for subsequentanalysis. However, while the samplerrequired in the TSP and lead methods Isexplicitly specified by design anddimensions and must be reproducedprecisely, EPA is proposing to specifythe sampler in the new PMo referencemethod primarily by performance,together with explicit test procedures tobe used to determine acceptability. Thisapproach allows for the use of currentlyavailable and tested sampler designswhile providing greater flexibility toencourage improvements andinnovations in future sampler designs, Itis also consistent with the approachused for specifying reference methodsfor several other criteria pollutants (CO,

10454

100