IS)
 o
 W
 CO
 C
      &ER&
                  United States
                  Environmental Protection
                  Agency
           Office of Air Quality
           Planning and Standards
           Research Triangle Park NC 27711
EPA-450/2-78-036
OAQPSNo. 1.2-1O
August 1978
                  Air
Guideline Series

Supplementary
Guidelines for Lead
Implementation
Plans
 0)
I
i

-------
I
                                           EPA-450/2-78-038
|                                          OAQPS No. 1.2-104
I
I
1                Supplementary Guidelines
I             for Lead Implementation Plans
i
i
i
i
i
i
i
i
I                    U.S. ENVIRONMENTAL PROTECTION AGENCY
                        Office of Air, Noise, and Radiation
I                      Office of Air Quality Planning and Standards
                     Research Triangle Park, North Carolina 27711
                             August 1978

I

-------
                  OAQPS GUIDELINE SERIES

The guideline series of reports is being issued by the Office of Air Quality
Planning and Standards (OAQPS) to provide information to state and local
air pollution control agencies; for example, to provide guidance on the
acquisition and processing of air quality data and on the planning and
analysis requisite for the maintenance of air quality.  Reports published in
this series will be available - as supplies permit - from the Library Services
Office (MD-35), U.S.  Environmental  Protection Agency, Research Triangle
Park, North Carolina  27711;  or, for a nominal fee, from the National
Technical Information Service,  5285 Port Royal Road, Springfield,  Virginia
22161.
                   Publication No. EPA-450/2-78-038

                         (OAQPS No. 1.2-104)

-------
I

I
_                                          PREFACE

                 This guideline appeared in  draft  form  in  November  of 1977  and
•          was referred to in the preamble  to the proposed  regulations  on  lead
•          implementation plans,  which appeared in  the Federal  Register of
*          December 14, 1977, (42 FR 63087).   This  final  edition reflects  a
I          number of changes  from the draft version.   The significant revisions
            are as follows:
|               --Revision of section 3.2 concerning HATREMS.
g               --Revision of section 4.3 on  projecting automotive lead emissions
™                 to correct  errors  in the  units  in the equations  and provide
•                 values for  several  expressions.
                 --Inclusion of a  new section  (4.4)  on  air quality  modeling.
J               --Revision of Chapter 5 on  siting of urban  area ambient air
—                 quality monitors for lead to reflect a  number of comments.
•               --Deletion of the draft Chapter 7,  which  was  a  brief descrip-
•                 tion of the emission sampling technique.
                 --Revision of the inorganic lead  testing  method that appeared
•                 in Appendix A,  and the inclusion  of  a test  method for alkyl
                   lead, which appears as the  new  Appendix B.
I               --Addition to the material  on deposition  of particles and  gases
•                 in the new  appendix D (Appendix C in the  draft)  with  more
                   recent,  more representative material; this  new material  appears
•                 as Appendix E.

I

I

I

-------
     --Inclusion of a paper that describes  the modified rollback  model,
       required as a minimum in several  of  the analyses in  the  State
       implementation plans; this appears as  Appendix G.
     There are also a number of changes  that  were made to reflect the
revision of the national  ambient air quality  standard from a monthly
average to a quarterly average.
                               iv

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                           ACKNOWLEDGEMENTS



     This guideline was prepared under the general  editorship  of  John

Silvasi and Joseph Sableski of the Plans  Guidelines Section, Standards

Implementation Branch, Control Programs Development Division.   CPDD

gratefully appreciates the following contributions:

     Chapter 3—Reporting Requirements
     Chapter 5—Ambient Lead Monitoring



     Chapter 7--Determination of Lead Point Source

                Definition

     Appendix A—Procedures for Determining

                 Inorganic Lead Emissions from

                 Stationary Sources

     Appendix IB—Procedure for Determining Alkyl

                 Lead Emissions from Alkyl Lead

                 Manufacturing Plants
Jacob Summers

   MDAD

Alan Hoffman

   MDAD

James Dicke

   MDAD

Bill Mitchell,

Rodney Midgett

   EMSL

Bill Mitchell/

Rodney Midgett

   EMSL
     Appendix C—Projecting Automotive Lead Emissions    James  Wilson

                 for Roadway Configurations                MDAD

     CPDD gratefully acknowledges the permission  granted by Pergamon

Press and Thomas W.  Horst to reproduce the article,  "A Surface Depletion

Model for Deposition from a Gaussian Plume," which appeared 1n Atmospheric

Environment in 1977.

     In addition, CPDD wishes to thank those who  offered comments  and

suggestions on the draft of this guideline.

-------
                        TABLE OF CONTENTS
Chapter                                                             Page

1.0  INTRODUCTION	   1

2.0  GENERAL IMPLEMENTATION PLAN DEVELOPMENT	   3

     2.1  Subpart A--General Provisions	   3
     2.2  Subpart B--PIan Content and Requirements	   5
     2.3  Subpart C--Extens1ons	  10
     2.4  Subpart D--Maintenance of National Standards	  1?.
     2.5  Forthcoming Requirements	  12
3.0  REPORTING REQUIREMENTS	  15

     3.1  A1r Quality Data Reporting	  15
          3.1.1  SAROAD System	  15
          3.1.2  Reporting Formats	  1(5
          3.1.3  Coding Procedures	  21
          3.1.4  Data Flow	  21
          3.1.5  References for Section 3.1	  22
     3.2  Emissions Data Reporting	  23
          3.2.1  HATREMS System	i	  24
          3.2.2  HATREMS Reporting Formats.	  26
          3.2.3  Coding Procedures	  34
          3.2.4  Data Flow	,	  35
          3.2.5  References for Section 3.2	  38

4.0  ANALYSIS AND CONTROL STRATEGY DEVELOPMENT	  39

     4.1  Background Concentrations	  39
     4.2  Lead Emission Factors	  40
     4.3  Projecting Automotive Lead Emissions	  43
          4.3.1  Lead Emissions from Automobiles	  43
          4.3.2  Lead Emissions from Other Gasoline Powered
                 Vehicles	  45
          4.3.3  Example Calculation of Automobile Lead Emissions...  47
     4.4  Air Quality Modeling	  60
          4.4.1  Modeling in Certain Areas	  GO
          4.4.2  Modeling in Areas Around Significant Point
                 Sources	  61

5.0  AMBIENT LEAD MONITORING	  63

     5.1  Introduction 	  F3
     5.2  Monitoring Scales	  67
     5.3  Site Descriptions	  68
          5.3.1  Roadway Site (Middle Scale)	  68
          5.3.2  Neighborhood Site (Neighborhood Scale)	  69
          5.3.3  Street Canyon Site (Middle Scale)	  70
     5.4  Other Considerations for All Sites	  70
     5.5  Network Design	  71
     5.6  Frequency of Sampling	  72

6.0  NEW SOURCE REVIEW	  73

                                   vi 1

-------
Chapter                                                             Page

7 . 0  DETERMINATION OF LEAD POINT SOURCE DEFINITION ..................  75

APPENDIX A.  Procedures for Determining the Inorganic Lead
             Emissions from Stationary Sources ......................  79

APPENDIX B.  Procedure for Determining the Alkyl  Lead Emissions
             from Al kyl Lead Manuf actur i ng PI ants ................... m
APPENDIX C.  Projecting Automotive Lead Emissions for Roadway Con-
             figurations ............................................ 1Z9

APPENDIX D.  Deposition of Particles and Gases ...................... 155

APPENDIX E.  A Surface Depletion Model for Deposition from a
             Gaussian Plume ......................................... 165

APPENDIX F.  Calculation of Critical Ambient Lead Concentration
             Below Which the NAAQS will be Attained by 1982 Due to
             Mobile Sources in Urbanized Areas ...................... 173

APPENDIX G.  Rollback Modeling—Basic and Modified .................. 179

-------
                              LIST OF FIGURES
                      (Excluding those in Appendices)
Figure                                                            Page
3.1-1  SAROAD Site Identification Form	  17
3.1-2  SAROAD Daily Data Form	  19
3.1-3  SAROAD Composite Data Form	  20
3.2-1  National Emissions Data System (NEDS) Point Source Input
       Form	  30
3.2-2  National Emissions Data System (NEDS) Area Source Input
       Form	  31
3.2-3  Hazardous and Trace Emissions System (HATREMS) Point Source
       Input Form	  32
3.2-4  Hazardous and Trace Emissions System (HATREMS) Area Source
       Input Form	  33
3.2-5  Lead Emisssions Data Flow	  36
4.3-1  Percentage of Burned Lead Exhausted vs. Vehicle Cruise
       Speed	  49

4.3-2      >     C  . E,. . m.   vs. Speed	50
        M=1967    s>1  c>1  nl
                                  ix

-------
1
1
1
•

1

1
•

1
1
•

1

1
^m
•

1

1

1

1


1
1
1
1
1




Table
3.2-1

4.3-1

4.3-2
4.3-3
4.3-4

4.3-5
4.3-6

4.3-7

4.3-8

4.3-9

5-1

5-2

7-1









LIST OF TABLES
(Excluding those in Appendices)

Emission Inventory Data for Use in the Development of
Pb Control Strategi es 	
Probable Pooled Average Lead Content of Gasoline 	


City/Highway Combined Fuel Economy 	
Fraction of Annual Light-Duty Vehicle Travel by Model
Year 	

Fuel Economy Correction Factors by Model Year 	
C . Val ues 	

Values of 3> C . E . m 	
i=1967 Ssl C51 n
Probable Lead Content of Leaded Gasol ine 	
1974
Calculation of i = ]> C,,. . £„ . m. for Example
i=T9~67 lb)1 c>1 n
Calculation 	
Urbanized Areas Greater than 500,000 Population (1970
Census ) 	 ,
Urbanized Areas with Lead Concentrations Exceeding or
Equal to 1.5 ug/m , Maximum Quarterly Mean (1975) 	

Stationary Source Quarterly Modeling Results and Point
Source Definition for NAAQS Quarterly Average of
1 . 5 ug/m 	





X1




Page

27
51

52
53
54

55
56

57

58


59

64
cc


76







-------
 I

 I

 I                                      1.0   INTRODUCTION

                  This  guideline presents  information  on the  development  of implemen-
 I           tation plans  for lead  that are  not  contained  in  EPA's  regulations  for
             preparation,  adoption, and submission  of  implementation  plans, found in
 I           Part 51  of Title 40 of the Code of  Federal Regulations.   In  several  cases,
 _           the guidance  presented herein is referenced in those regulations;  EPA will
 •           use this guidance in determining the acceptability of  a  plan.
 •                A detailed  summary of the  background surrounding  the development
             of the regulations and the guidelines  appears in the preambles to  both
 I           the proposal  (Federal  Register  of December 14, 1977 (42  FR 63087)),
             and the final  version  of the  regulations  pertaining to lead  implementa-
 •           tion plans.

 I

 I

 I

 I

 I

 I

I

I

I

-------
 I

 I
 •                            2.0  GENERAL SIP DEVELOPMENT GUIDELINES
                     The  discussion below follows the outline of those portions of
 |             the 40  CFR 51  regulations, "Requirements for Preparation, Adoption,
 —             and Submittal  of  Implementation Plans," that were not revised to account
 ™             for the lead standard.  These requirements are still applicable to the
 •             lead SIPs as appropriate.  Items of a peripheral nature, such as "Defini-
                tions," are not discussed here.  The State should consult the regula-
 •             tions themselves, rather than this discussion, for the detailed
                requirements.
 •             2.1   SUBPART A—GENERAL PROVISIONS
 •                  --S  51.3   Classification of regions—This section will  not apply  to
                lead SIPs.
 I                  —S  51.4   Public hearings—Before the State submits the plan, a
                compliance schedule, or a plan revision to EPA, it must hold a public
 •             hearing on the plan, schedule or revision.   The State must also give
 •             proper  public notice for the hearing at least 30 days prior  to the
                hearing.   Although this section specifies formal requirements for the
 I              notice  and holding of public hearings, a State can obtain EPA approval
                to  use  alternative procedures that EPA deems adequate.
 I                   --S  51.5  Submission of plans; preliminary review of plans—The
 •              State must submit at least five copies of the plan to the appropriate
                Regional  Office within nine months after EPA promulgation of a primary
 I              or  secondary standard.  The State can obtain from EPA an extension of

 I

 I

I

-------
18 months to submit a plan  to implement  the  secondary standard; this
provision 1s not applicable to lead  plans, however, since the second-
ary standard and primary standard  are  Identical.  Plans  for different
AQCRs within a State can be submitted  as a single document or as
separate documents.
     A State can submit to  EPA a plan  or portions thereof for pre-
liminary review before it is due.
     —S 51.6  Revisions—EPA may  ask  a  State  to revise  its plan for
three reasons:
     (a)  Revisions of a national  standard.
     (b)  The availability  of improved or more expeditious methods of
          attaining the national standards.
     (c)  A finding by EPA  that the  plan is  substantially inadequate
          to attain or maintain the  national standard which it imple-
          ments .
     After notification that a plan  needs revision, the  State has
60 days or any longer period specified by EPA  to submit  the revision.
     --S 51.7  Reports—The State  must report  air quality data to
EPA on a quarterly basis and must  submit the data 1n a specified
format for conversion to machine-readable format.
     The State must also report emissions from point sources whenever
there is a change in emissions, such as  a source coming  into compliance,
a new source beginning operation,  and  a  source ceasing operation.  The
State must also submit the  emissions data in a specified format for
conversion to machine-readable format.   In addition, States must report

-------
on progress 1n plan enforcement and on any substantive  revisions
the State makes 1n the plan other than to rules  and regulations.
EPA intends to revise these requirements  shortly;  the revisions
will cover all the pollutants for which implementation  plans  are
needed, including lead.
2.2  SUBPART B—PLAN CONTENT AND REQUIREMENTS
     --S 51.10  General  requirements—A plan  must  provide  for attain-
ment of national primary air quality standards within three years
after the date of EPA approval  of the plan, although the State can
obtain a two-year extension upon EPA approval of application.  A
plan must provide for attainment of national  secondary  air quality
standards within a reasonable time after  the  date  of EPA approval
of the plan; this provision is  inapplicable to the  lead plans  because
the primary and secondary lead  air quality standards are identical.
Also, the plan for one AQCR must provide  that emissions from  the AQCR
do not interfere with attainment and maintenance of a national standard
in another AQCR.   In addition,  the plan must  provide for public availa-
bility of emission data  from all sources, correlated with  allowable emis-
sions.
     --§ 51.11  Legal  authority—The plan must show that the  State has
the legal authority to—
     — adopt emission limitations;
     --enforce laws, regulations,  and standards, and seek  injunctlve
       relief;
     --abate emissions during an emergency episode;

-------
     —prevent construction  or  modification of facilities that may
       result 1n a violation of a  national air quality standard;
     --obtain information  necessary  to  determine compliance;
     — require source owners to install emission monitoring devices;
       and
     --make emission data  available  to  the public.
     If the plan contains  measures for  transportation and land use
controls (other than new and modified source  review), the plan does
not have to demonstrate that authority  for such measures exists at
the time of plan submission.  In such cases,  however, the plan must
contain a schedule for obtaining the legal authority.
     A State may delegate  authority  and responsibility to a substate
entity for carrying out a  plan  or  portion thereof  if the plan demonstrates
that the substate entity has the legal  authority for carrying out its
responsibility and that the  State  is not relieved  of the responsibility
for carrying out the plan  or portion thereof  if the substate entity
fails 1n its reponsibillty.   (S 51.55 of Subpart D, which pertains to
plans in AQMAs and like areas,  provides an exception to this require-
ment.)
     --S 51.12  Control strategy;   General—This section prescribes
some general requirements  for developing and  evaluating the control
strategy, which is the heart of the implementation plan.  The plan

-------
must demonstrate that the control  strategy  1s  adequate to attain the
standard within the appropriate  period  and  maintain it thereafter.
(Subpart D of the regulations, discussed  below, specifies additional
maintenance requirements  for some  areas.)
     —S 51.13  Control strategy:   Sulfur oxides and particulate matter.
     --S 51.14  Control strategy;   Carbon monoxide, hydrocarbons.
photochemical  oxidants. and nitrogen  dioxide.
     These two sections set forth  requirements specific to the subject
pollutants and do not apply to the lead implementation plans.  The
control strategy requirements for  the lead  implementation plan are
found in a newly-created  Subpart E.
     The process of developing and evaluating  a control strategy for
attainment of the national  standard entails the following:
     —Development of an  emission  and air quality data base.
     —Determination of whether  the air quality standard is being
       violated.
     —Development of alternative  control strategies.
     —Evaluation of the  control strategies by accounting for emission
       reductions and modeling air quality  concentrations.
     --Choice  of an appropriate  strategy.
     Detailed guidance for performing these tasks appears in EPA's
Air Quality Analysis Workshop. Volume 1 - Manual.
 C1r1llo, R.R.,  et al.,  A1r Quality Analysis Workshop. Volume 1 -
 Manual.  Prepared for the  Environmental Protection Agency, Office of
 A1r and Waste Management, Office  of Air Quality Planning and Stand-
 ards, Research  Triangle Park,  NC  27711.  November 1975.

-------
     --S 51.15  Compliance schedules—The  plan  must  contain  legally
enforceable compliance schedules that set  forth the  dates  by which
all sources must be in compliance with any applicable  portion of  the
plan.  Schedules extending over a period of more than  one  year from
the date of adoption must provide for legally enforceable  increments
of progress toward compliance.   All  schedules must provide for compli-
ance by the statutory attainment date or the end of  the  period covered
by an extension to the attainment date.  Revisions to  compliance  schedules
constitute revisions to the implementation plan; enforcement orders that
extend beyond the date for attainment of the national  standard must be
issued in accordance with the requirements of Section  113(d) of the
Clean Air Act.
     --S 51.16  Prevention of air pollution emergency  episodes--As
discussed in section 3.5 of the preamble to the proposed regulation,  this
section will not apply to lead implementation plans.
     --§ 51.17  Air quality surveillance—These regulations  pertain to
pollutants other than lead and therefore do not apply  to the lead imple-
mentation plans.  The requirements for lead air quality  surveillance
appear in a newly-created § 51.17b.
     --§ 51.17a  Air quality monitoring methods—This  section prescribes
procedures for obtaining approval to use nonconforming analyzers  (i.e.,
those analyzers that do not use the reference or equivalent  methods),
methods with nonconforming ranges, and methods  that  have been modified
by users.

-------
 I
 I
 I                  —S  51.18   Review of new sources and modifications — The plan must
 _             provide legally  enforceable procedures to enable the State to determine
 ™             whether the construction or modification of a facility will result in
 •             violations of a  portion of the control strategy or will interfere with
               attainment or maintenance of a national standard.   The State must also
 •             have  the  ability and procedures to prevent construction or modification
               of  facilities that will result in control strategy violations or inter-
 •             fere  with a national standard.  Before approving or disapproving requests
 •             to  construct affected facilities, the State must provide opportunity
               for public notice of and comment on the action.
 •                  EPA  intends to revise these requirements shortly; the revisions
               will  cover all pollutants for which implementation plans are needed,
 I             including lead.
 •                  --S  51.19   Source surveillance—The plan must provide for moni-
               toring the status of compliance with any rules and regulations which
 •             constitute the control strategy.  As a minimum, the plan must provide:
                    --Procedures for requiring owners or operators of sources to
 |                   maintain  records of emissions and report periodically to the
 _                   State ;
                    --Periodic  testing and inspection of sources;
H                  --A system  of detection of violations of rules and regulations
                     through enforcement of a visible emission limitation and for
 |                   investigating complaints (the provision concerning visible
                     emission  limitations 1s inapplicable to the lead plan);
—
I
I

-------
     "Procedures for bbtalning and  maintaining data on emissions
       reductions achieved through transportation control measures; and
     —Procedures to require  sources to  install and use emission
       monitoring devices  (this provision  is  also inapplicable to
       the lead plan).
     --S 51.20  Resources—The plan  must provide a description of the
resources available to the State and local  agencies at the time of
plan submission and of any additional  resources needed to carry out
the plan during the five-year period following its submission.
     --S 51.21  Intergovernmental cooperation—The plan must pro-
vide assurances for the State to submit  to other States air quality
and emission data.  Also,  the plan must  identify local agencies that
will participate in implementing the plan  and their responsibilities.
     --S 51.22  Rules and  regulations—The State must adopt all rules
and regulations necessary  for attainment and  maintenance of the national
standard.  The plan must contain copies  of all such rules and regula-
tions.
2.3  SUBPART C--EXTENSIONS.
     --S 51.30  Request for two-year extension--The Governor of a
State may, at the time he  submits a  plan to implement a primary stand-
ard, request EPA to extend the 3-year  period  for attainment of the
primary standard for a period not exceeding two years.  To obtain the
                               10

-------
extension, the State must demonstrate a number of items,  particularly
that the necessary technology or alternatives will  not be available
soon enough to permit full implementation of the control  strategy
within three years.
     —S 51.31  Request for 18-month extension—Upon request of a
State, EPA may extend, for a period not exceeding 18 months, the
deadline for submission of a plan to implement a secondary air quality
standard.  As discussed above, this provision is inapplicable to the
lead plans, because the primary and secondary ambient lead standards
are identical.
     --S 51.32  Request for one-year postponement—The Clean Air Act
Amendments of 1977 have rendered this section inapplicable;  EPA will
amend Part 51 to reflect this.
     —S 51.33  Hearings and appeals relating to request  for one
year postponement—This section prescribes the procedures relating
to a request for a one year postponement under § 51.32.   The Clean  Air
Act Amendments of 1977 have rendered this section inapplicable; EPA will
amend Part 51 to reflect this.
     —S 51.34  Variances.  This section requires States  to  submit
variances to regulations as revisions to the plan under S 51.6.
                                 11

-------
2.4  SUBPART D—MAINTENANCE OF NATIONAL STANDARDS
     This subpart pertains to the detailed analysis to determine whether
a certain area will  maintain a national air quality standard and to the
plan to maintain the standard.  EPA must select the areas to which the
regulations of this  subpart apply, and so the regulations do not apply
to all areas.
     The subpart consists of sections 51.41 through 51.63 inclusive.
Topics of the sections concerning the analysis are submittal date,
analysis period, guidelines, projection and allocation of emissions,
projection of air quality concentrations, description of data sources,
data bases, description of techniques, accuracy of calculations, and
submittal of calculations.  Topics of the sections concerning the plan
are demonstration of adequacy, strategies, legal  authority, intergovern-
mental cooperation,  surveillance, resources, and submittal.  Two other
sections cover both  the analysis and the plan and concern data availa-
bility and the use of alternative procedures.
2.5  FORTHCOMING REQUIREMENTS
     In addition to  the above mentioned regulations, EPA will shortly
publish additional requirements that account for the Clean Air Act
Amendments of 1977.   These requirements will appeare either as regula-
tions or policy guidance.  The new requirements will cover the following
topics:
     —Provisions for review of new sources in nonattainment areas (this
       will not apply to lead plans).
     --Additional transportation-related provisions.
     --Accounting for stack heights.
     --Assessing adequacy of plan in relation to long-term fuel supplies.
                                   12

-------
 I

 I
                  --Prevention of significant deterioration.
 •                --Permit requirements.
 •                --Indirect source review.
                  --Delegation of authority to local governments.
 I                --Interstate pollution abatement.
                  --Consultation with governmental entities at the local and Federal
 I                  level.
 m                --Planning procedures to allow local governments more authority
                    in developing and implementing plans.
 fl                --Noncompliance penalties.
                  --Permit fees.
 |                --Composition of State air pollution boards.
 M                --Provisions prohibiting loss of pay of employees at facilities
 *                  that use supplemental control systems.
 •                —Provisions for public notification of dangers of air pollution.
                  --Protection of visibility 1n certain areas.
 H                —Emergency episode reporting.
 _                --Energy or economic emergency authority.
 ™                --Suspension of transportation control measures.
 •                --Measures to prevent economic disruption or unemployment.

 I

 I

 I

 I

I

-------
I
I
                                     3.0  REPORTING REQUIREMENTS
                3.1  AIR QUALITY DATA REPORTING
I                   The quarterly air quality reporting requirements  specified  in  40
                CFR 51.7(a) will apply to lead as  well  as the other criteria pollutants.
|              The air quality data for lead will be stored in  the Storage  and  Retrieval
«              of Aerometric Data (SAROAD) System and utilized  by EPA,  States,  and
*              private individuals.
j|              3.1.1  SAROAD System
                     The SAROAD System includes several  data files which include  the site
I              file, the parameter-method file, geographical files, raw data files, and
—              the summary file as well as the software necessary to  update these  files
™              and generate reports from the data.
•                   The site file contains a listing of all  sites reporting data and
                includes a description of the physical  characteristics of the site.  Each
I              site is assigned a unique code and this  code must  be attached to  data to
                identify the sampling site.  Only  data with  a valid site code will  be
«•              accepted on the data bank.
•                   The parameter-method file contains  an entry for each valid  combina-
                tion of sampling and analysis method  for each pollutant.   Only data with
•              a valid entry will be accepted on  the data bank.   The  code '92'  has been
                assigned for lead data collected by Hi-Vol and analyzed  by atomic absorp-
I              tion.  Additional codes may be requested through the Regional  Offices.
m                   The geographic files contain  the specific SAROAD  codes  and the
                corresponding names and are utilized  to  print location names on  standard
•              reports.

I

I

-------
     The raw data files contain 80-90 million individual  data values  for
the time period of 1957-present as collected and reported by EPA,  State,
and local control agencies.  These data are available upon request in
standard raw data listings and are utilized by programs  to generate
summary statistics.
     The summary file contains quarterly and annual  summary statistics
which include frequency distributions, arithmetic and geometric means
and standard deviations, first and second maximums,  violation counts,
etc.  These summary statistics are available in standard reports.
3.1.2  Reporting Formats
     The standardized input formats which will be utilized to report  lead
data include:  Figure 3.1-1 SAROAD Site Identification Form, Figure 3.1-2
SAROAD Daily Data Form, and Figure 3.1-3 SAROAD Composite Data Form.
     The Site Identification Form is utilized to register a new site  or
change information for an existing site.  The coding instructions  are
given in Section 3.4.1 of AEROS User's Manual.
     The Daily Data Form is utilized to report air quality concentrations
for lead as determined from individual 24-hour integrated samples. The
coding Instructions are given in Section 3.4.3 of AEROS  User's Manual.
     The Composite Data Form is utilized to report data  values which
are composites of individual 24-hour integrated samples.   For lead, these
composites could be done on a calendar month to reduce the resources
required to analyze individual samples.  The coding instructions are
given in Section 3.4.4 of AEROS User's Manual.
                                16

-------
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
                                     Figure  3.1-1
THE REPORT IS REQUIRED BY LAW
42 USC 1857; 40 CFR 51
 Form Completed By.
ENVIRONMENTAL PROTECTION AGENCV
   National Aerometric Data Bank
 Research Triangle Park. N. C 27711

 SAROAO Site Identification Form

 	Date	
                                                                      .New
Revised
D
TO BE COMPLETED BY THE REPORTING AGENCY
(A)

.14-3«l
,37-Sli
City

52
0 0
60 61
JTM Zone

60 61
IH\
118-76

State Project
City Name (23 characters)
County Name (15 characters)
Population (right justified)

S3 54 55 56 57 56 59
Longitude Latitude
Deg. Mm. Sec. Deg. Mm. Sec.
W| II N j |
62 63 64 65 66 67 69 69 7O ;> 11 7] 74 75 76
Easting Coord., meters Northing Coord., meters

6"* 63 64 65 66 67 66 69 70 /I 72 73 74 75 76

Supporting Agency (61 characters')
Supporimg Agency, continued


(14-79'
Optional Comments that will help identify
the sampling site (132 characters)

ni
114-791




DO NOT WRITE HERE
State Area
A
1 23456
Agency Project
n
Site
|
; 8 9 m
11 12 13
Time
Region Zone Action


; > 7* -9 ao
State Area
B
Site
1
1234567 69 10
Agency Project SMSA Actio


II f n 14 IS 16 17 60
State Area
C
173456
Agency Project
.... 1 1 1
II I? 13
State Area
D
1 r 3 4 5 6
Agency Project

II 17 13
Site

7 8 9 1C
Action
n
80
Site
— r~i — i

; e 9 10
Action

60
                                                                  State      Area
                                                       Site
(E).
             Abbreviated Si.e Address (25 characters)
OMB No. 158-R0012
Approval expires 2/77.
                                I.34S67B9IO
                             Agency   Project         Action

                              ~~  CD
                                      f 13
                                           17

-------
                                 Figure  3.1-1   (cont'd)
                             SAROAD Site Identification Form (continued)
TO BE COMPLETED BY THE REPORTING AGENCY
                                                           00 NOT WRITE HERE
              Sampling Site Address (41 characters)
Cneck (he ONE
major category thai
best describes the
location of the
sampling site.
1.CU CENTER CITY
2. CU SUBURBAN
3,1   I RURAL
4.1   I REMOTE
Specify
units	
Address, continued

  Next, check the subcategory
  thai best describes the domi-
  nating influence on the sampler
  within approximately a 1-mile
  radius of the sampling site
      1  Industrial
      2. Residential
      3. Commercial
      4  Mobile
      1  Industrial
      2  Residential
  	I 3. Commercial
      4  Mobile
      1. Near urban
      2. Agricultural
      3. Commercial
      4. Industrial
      5. None of the above
               Elevation of sampler above ground
Specify
units 	
                                                               State
                                                       Area
                                                                                          Site
                                                             M    I!   T   I    I    I
                                                             1    2   3   4   5
                                                              Agency
                                                               D
                                                         Project
Station Type
                                                                             County Code
                                                                            51   M   41   60
                                                                          AQCR Number
                                                                            61   67  63
                                                                                AOCR Population
                                                                64  fiS   66   67   6B  t>9   '0   M
                                                                       Elevation/Gr
                                                                        7?   73   J4
                                                             Elevation'MSL
                                       T.   76   II   78
                                                                    Action
                                                                    D
           Elevation of sampler above mean sea level

Circle pertinent time zone    EASTERN     CENTRAL
MOUNTAIN   PACIFIC    YUKON   ALASKA   BERING
HAWAII
                                             18

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                                   Figure  3.1-2
                                ENVIRONMENTAL PROTECTION AGENCY
                                   National Aerometric Data Bank
                                 Research Triangle Park, N. C. 27711

                                     SAROAD Daily Data Form
24-hour or greater sampling interval

 2~
                                THE REPORT IS REQUIRED BY LAW
                                42 USC 1857;40CFR 51
                       Agency
                                                          State
    OMB No. 158-R0012

    Approval expires 2/77.



Area       Site
                      City Name
                                                          2345
                                                                         7  8   9   10
Site Address
Project
Name
PARAMETER
Code
Day
19 20
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
St Hr
21 22

































































23 24 2
Method





5 26 27
Units



:e 29 3" 31
3J 34 35 36





























































































































DP

12
Time Interval
Name
PARAMETER
Code





37 18 39 41) 41
Method Units

•l?


43
47 4


































44 45
3 4H 5"





























































































DP
n
4 13
Agency
D
Project lime rear
m n m
12 13 14 15 16
Name
PARAMETER
Code





51 *' 53 51 0
Method Units

56




5
5' 58 09
61 62 63 64





























































































































DP
D
60
IV
[
onin
IE]
7 18
Name
PARAMETER
Code





65 66 67 68 t
Method Units

71





9
DP
n
71 72 73
75 76 77 78




























































































































74
               43210
                                    43210        43210        432'u
                                           19

-------
oo
 i
r—

CO



 c
 3
 cn
        E!
u
J
ci 0- Ld
: »p.
I £JjH 	 IN
~ o >-
U H
U.
O 1/1
* z "*
CD
• SR*
irt iii 1 [r«.
a 1 	 1-
« al 	 |u>
" S!US
fl
u '
i
Ul
a
!
7
UJ
J



















































































TOTAL
% FILTER
EQUIVALENT


                                      20

-------
I
I
                  3.1.3  Coding  Procedures
                      The  coding procedures for each form will not be discussed 1n detail
I                since the appropriate sections of AEROS User's Manual were referenced
                  above.  In  addition to the AEROS User's Manual, the AEROS Manual of Codes2
I                is  required as a  reference to code the data.
m                    The  following additional rules will help to reduce the errors iden-
                  tified by standard edit checks:
I                    (1)   Utilize the correct standard format to submit data.
                      (2)   Utilize the complete 12-digit site code which was assigned
|                          by the  Regional Office or the National Air Data Branch.
£                    (3)   Utilize only valid combinations of pollutant-method-interval
*                          unit codes.
•                    (4)   Code missing values as blanks and values below the minimum
                            detectable for the sampling-analysis method as '0000'.
I                3.1.4  Data Flow
—                    The  lead air quality data are submitted with other air quality data
•                in  SAROAD standard input formats in the quarterly report.  The complete
•                data flow including specific data processing steps and example edit
                  error messages is discussed in Section 7.0.0 of AEROS User's Manual.
•                    After  the data have been updated, they are available for retrieval
                  from the  data bank through routine reports which are documents in Section
•                2.3.0.0 of  Summary and Retrieval Manual.

I

I

I

I

-------
3.1.5  References for Section 3.1
     1.   AEROS Manual Series Volume II:   AEROS User's  Manual,  EPA-450/2-
76-029,  USEPA, OAQPS.
     2.   AEROS Manual Series Volume V:   AEROS Manual  of Codes, EPA-450/2-
76-005,  April, 1976, USEPA, OAQPS.
     3.   AEROS Manual Series Volume III:  Summary and Retrieval,  EPA-450/2-
76-009,  May, 1976, USEPA, OAQPS.
                                22

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.2  EMISSIONS DATA REPORTING


     Under subpart E to 40 CFR Part 51, States are required to develop

a control strategy for lead and to submit this as part of the State

Implementation Plan for lead.  This control strategy must include a

summary of the baseline emission inventory and a detailed emissions

inventory for point and area sources of lead.   As indicated in sub-

part E, 51.86 (b), this section of the guideline describes the data

formats and data bases that will be utilized to process these data

and identifies the sources that must be reported.


     Since the National Emissions Data System (NEDS) has the capa-

bility to store emissions for only particulate, sulfur dioxide, carbon

monoxide, hydrocarbons, and nitrogen dioxide, the Hazardous and Trace

Emissions System (HATREMS) has been developed to calculate and store

emissions data for lead as well as other possible future criteria pol-

lutants and noncriteria pollutants.


     Although HATREMS is discussed in this section as the data base that

will be utilized to store emissions data for lead, the submission of  data

on HATREMS point and area source forms can not currently be considered a

legal requirement.  After the HATREMS point and area source forms have been

approved by the Office of Management and Budget (OMB), the submission of  data

in HATREMS format will be required and a revision sheet will  be published

for this document to indicate that approval has been obtained.   Although

not currently required, data in HATREMS format may be submitted by the con-

trol agency, and it will be stored in HATREMS.   The emissions data for lead

sources must be submitted in NEDS format, and the NEDS data will  be utilized

by HATREMS to calculate lead emissions data.
                                23

-------
3.2.1  HATREMS System

     To minimize lead emissions data  collection and  reporting  require-
ments, HATREMS utilizes existing NEDS data to calculate and  store lead
data.  Since HATREMS utilizes NEDS data,  the identifiers from  NEDS
which include state, county, AQCR, plant, point, and source  classification
code (SCC) are utilized by HATREMS.   In addition to  utilizing  NEDS point
and area source data, HATREMS includes update and report programs as  well
as the following files:  (1) point source emission factor file,  (2) area
source emission factor file, (3) point source emissions data file, and
(4) area source emissions data file.

     The point source emission factor file contains  an entry for each valid
SCC which emits lead.  Each entry contains the following:

     (1)  Emission factor - emission  rate of lead per unit process.
     (2)  Default multiplier - average lead content  in the process
          material.  This parameter is utilized the  same way sulfur
          content is utilized to calculate S02 emissions.
     (3)  Default multiplier units -  units in which  the lead content  is
          expressed  (% or ppm).
     (4)  Control efficiency multiplier - the % of the NEDS  control
          efficiency which is applied to lead emission.
     (5)  Pollutant  flag - indicates  the NEDS pollutant control  efficiency
          to utilize.  For lead, the particulate control efficiency will  be
          utilized and the control efficiency multiplier will  be 100.

     The area source emission factor file contains an entry for each
valid area source category (ASC) which emits lead; each entry contains
an emission factor,  a default multiplier, and the default multiplier
units.  These parameters are the same as the corresponding parameters in
the  point source emission factor file.
                                  24

-------
I
I
I
I
                                                            Iioo-w noo)
                                         u _ nn >/ rr ,/ nu >,     inn
I
I
I
I
I
I
I
I
I
I
I
I
I
I
     The point source emissions data file contains the emissions data
for individual processes within plants.  This file is routinely recreated
by combining operating rates and control efficiencies from NEDS with
emission factors from the point source emission factor file utilizing
the following formula:
                         M = OR x EF x DM x     100         (1)
                                      2000

where:
     M = emissions in T/Y
     OR = operating rate from NEDS
     EF = emission factor from HATREMS point source emission factor file
     DM = default multiplier from HATREMS point source emission factor
          file
     W = NEDS control efficiency for the pollutant flagged
     Y = control efficiency multiplier from the HATREMS point source
         emission factor file.

     As mentioned previously, the HATREMS point source emissions data
file is routinely updated from NEDS data utilizing formula 1.  Point
source data are also manually updated as described in the following sec-
tions.  Once a data record has been manually updated, data from NEDS will
not override it.

     The area source emissions data file contains the emissions data for
each ASC for each county.  This file is routinely recreated by combining
area source rates from NEDS with emission factors from the area source
emission factor file utilizing the following formula:
                         M = SR x EF x DM
                                   2000                (2)

-------
where:
     M  = emissions in T/Y
     SR = area source rate for individual  counties
     EF = emission factor from HATREMS area source emission factor file
     DM = default multiplier from HATREMS  area source emission factor
          file.

As mentioned previously, the HATREMS area  source emissions data file is
routinely updated from NEDS utilizing formula 2.  Area source data are
also manually updated as described in the  following sections.  Once a
data record has been manually updated, data from NEDS will not override
it.

     Table 3.2-1 lists the data items which are utilized from NEDS as
well as the data items which are stored in HATREMS.  The data items which
are utilized from NEDS can also be updated and stored in HATREMS utilizing
the input forms which are described below.

3.2.2  HATREMS Reporting Formats

     To insure that the volume of data can be automatically processed,
HATREMS utilizes standarized input formats.  Since HATREMS utilizes NEDS
data, the input formats include Figure 3.2-1, NEDS Point Source Form, and
Figure 3.2-2, NEDS Area Source Form, as well as Figure 3.2-3, HATREMS Point
Source Form, and Figure 3.2-4, HATREMS Area Source Form.  The NEDS point
and area source forms have been utilized for several  years and will not be
discussed further.

     The HATREMS point source form is utilized to manually update HATREMS
with data which are not available from NEDS.  The form defines six unique
card formats for data coding.  The first two cards are identical to the
first two for NEDS except the card code.  The other cards (3P, 4P, 5P, and
6P) allow inputting data to HATREMS which are specific to lead.  Multiple
                                  26

-------
I

I
                            Table 3.2-1:  EMISSION INVENTORY DATA FOR USE
                             IN THE DEVELOPMENT OF Pb CONTROL STRATEGIES
I

I

I

I
                I.   GENERAL SOURCE INFORMATION
•                   A.  Establishment name and location (address and Univer-     NEDS
                         sal Transverse Mercator grid coordinates)
|                   B.  Person to contact on air pollution matters               NEDS
                     C.  Source Classification Code (SCC)                         NEDS
                                                                               System in
                                                                               Which Data
                                                                               Will be Stored
                     D.  Operating Schedule                                       NEDS
                         1.  hours/day
                         2.  days/week
                         3.  week/year

                     E.  Year in which data are recorded                          NEDS
                      F.  Future activities (% increase in production or through-  HATREMS
                         put in 10 and 20 years)
•               II.  FUEL COMBUSTION

•                   A.  Number of boilers                                        NEDS

                     B.  Type of fuel burning equipment for each boiler           NEDS

|                   C.  Rated capacity of each boiler, BTU/hr                    NEDS

—                   D.  Types of fuel burned, quantities and characteristics
•                       1.  Type of each fuel used                               NEDS
•                       2.  Maximum quantity per hour                            NEDS
                         3.  Average quantity per hour                            HATREMS
                         14.  Quantity per year                                    NEDS
                         5.  Lead content of throughput                           HATREMS
                         6.  Heat content of fuel                                 NEDS

•                   E.  Percent used for space heating and process heat          NEDS

                     F.  Air Pollution control equipment
•                       1.  Type                                                 NEDS
•                       2.  Pb collection efficiency (actual), %                 HATREMS


I



I



I

-------
                        Table  3.2-1  (cont'd)
     G.   Stack data
         1.   List stacks  by  boilers  served
         2.   Stack height,  ft.
         3.   Stack diameter  (inside, top),  ft.
         4.   Exit gas  temperature, °F
         5.   Exit gas  volume,  cfm

     H.   Emission data
         1.   Estimate  of  Pb  emissions (tons/year)
         2.   Results of any  stack tests  conducted  (tons/year)

III.  MANUFACTURING ACTIVITIES

     A.   Process name  or  description for each product
                                                              System  in
                                                              Which Data
                                                              Will be Stored
                                       NEDS
                                       NEDS
                                       NEDS
                                       NEDS
                                       NEDS
                                       HATREMS
                                       HATREMS
                                       NEDS
     B.   1.   Quantity of raw materials  used  and handled  for each  NEDS
             product  or quantity  of each  product manufactured,
             maximum  quantity per hour  and average  quantity per
             year.
         2.   Pb content
             grinding.
by weight) for ore crushing and
     C.   Air pollution control  equipment  in  use
         1.   Type
         2.   Pb collection  efficiency  (actual),  %

     D.   Stack data
         1.   List stacks  by equipment  served
         2.   Stack height,  ft.
         3.   Stack diameter (inside, top),  ft.
         4.   Exit gas temperature, °F
         5.   Exit gas volume,  cfm

     E.   Emission data
         1.   Estimate of Pb emissions  by  the source
         2.   Results of any stack tests  conducted

IV.  REFUSE DISPOSAL

     A.   Amount and description of refuse generated,  quantity
         per year

     B.   Percent of total that is combustible
HATREMS
                                       NEDS
                                       HATREMS
                                       NEDS
                                       NEDS
                                       NEDS
                                       NEDS
                                       NEDS
                                       HATREMS
                                       HATREMS
                                       NEDS


                                       NEDS
                                28

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                        Table 3.2-1 (cont'd)

                                                               System 1n
                                                               Which Data
                                                               Will be Stored

     C.  Method of disposal                                        NEDS

     D.  Description of on-site disposal method, if applicable
         1.  Type of Incinerator                                   NEDS
         2.  Auxiliary fuel used                                   NEDS
         3.  Air pollution control equipment
             a.  Type                                              NEDS
             b.  Pb collection efficiency (design), %              HATREMS
         4.  Emission data
             a.  Estimate of Pb emissions by the source            HATREMS
             b.  Results of any stack tests conducted              HATREMS

     E.  Waste oil combustion tnew SCC)
         1.  Amount burned (1(T gal)                               NEDS
         2.  Pb content (35 by wt)                                  HATREMS
         3.  Estimated Pb emissions                                HATREMS

V.   AREA SOURCES

     A.  Gasoline combustion    ~
         1.  Amount consumed (10  gal)                             NEDS
         2.  Pb content (gPb/gal)                                  HATREMS
         3.  Estimated Pb emissions                                HATREMS

     B.  Coal combustion        3
         1.  Amount consumed (10  tons)                            NEDS
         2.  Pb content (ppm by weight)                            HATREMS
         3.  Estimated Pb emissions                                HATREMS

     C.  Oil  combustion         ~
         1.  Amount consumed (10  gal)                             NEDS
         2.  Pb content (% by wt.)                                 HATREMS
         3.  Estimated Pb emission                                 HATREMS

     D.  Solid waste incineration
         1.  Amount burned (10  tons)                               NEDS
         2.  Estimated Pb emissions                                HATREMS
                                 29

-------
CM
CD
                                          30

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
tf>
oS
S3
>tt
i»
(— UJ
CD _J zt
•r- «£ O °
u_ z oe
e>

UJ
<

- S
M
Z
_.,_ — o
•>
3

"* «j
t "5
. ~

- 5- "i
>- ° il
ijj
o
CO
en
r»
CD
f^.
fc
S
in
r^
*t
ro
r-»
CN
r*.
r*.
o
r*.
en
to
£
s
s
m
ID
-^
t£>
ro
Uf
rsj
U)
5
o
to
en
m
Ul
r->
m
to
in
ili
^
in
m
CNJ
m
in
o
m
m
v
«»
r^
^
§
m
^
*y
«*
ro
^
(N
•W
Jft
§
cn
ro
CO
ro
_S
a>
n
s
^
ro
ro
eg
ro
ro
S
m
IS
00
rw
r\
to
CN
vn
rvj
rr
tN
rs
CN
CN
CN
(N
cn
00
r».
to
in
"»
£?
CN
T
O
-
^
:












-
















































u

uoipv




vt
™

0
ct ~
"S o
I^T
*-
£ m
° £
Sla


TO
•0 0
£2
_J
rs
u.
«6 -
£ _ 3
o Q 2

«/n
** 0
O '
°"c=>



" S
E —
CD ""



»fl

5 s

1 I
S-S,

S-,
O u_

UJ
-j
^ 6 r1
Z -o <_3
O w T-
S£s
t—
txi
Z
§ - u.
-"=; °
gas
tr
UJ

O cp m
CO ^


£ 6
C "~
*^ "c>

0
00
en
CD
r-
£
LO
L/>
s
s
IN
r~
£
o
(£>
1C
U)
TD
to
-m
UJ
*T
vO
ro
'J3>
CN
tD

e
ID
tv>
ID
CO

~\&
lT>
^
r\i
o
r*
<












_-
'-










—

















--













h
1
u

uoipv
W Hod
5
I «-
K o
	 O
or

S" wi
_i * „
U.' 3.- C3
i =•£,
' W 	 :.
UJ
O
u
•S J5
» S
S'S.
*D ™



J? M
* ™
3C "
•— ro
o S


—i JSf
^> ^ 1/1
^ ** 7;
UJ "^ C3
5 s-a.
0 X ~
s
CJ

t; «
£0
no """"



=
c ^

| 2


z
K = «
* •crx
UJ JE 2
a.
o

-
— tn
i/i <^
O,' O
cc •"
-£ C
1^
O
z:
i — —
DC — p
- 1^
o -H — •
z
t—
ar
o "J5
"E "
£ o
or

0
GO
<7>
r--.
s
fc
i/1
r^
in
r-»
«*•
r--
ro
r>J
^
S
(Tl
(£>

£
S
-i/>
t£>
^f
U?
~?n
U>
Tj1
U)
CD
0
t£>
in
a
r-
m
(O
m
o>
m

o
m
O)
OO
«*
r>.
^r
LO
«•
in
ty
5
cn
*j
rsi
*T
"*
0
•O-
Ol
ro
oo
n
r^.
n
XD
M
S
«I
n
n-
rj
ro
ro
0




QC *t
*s* —
^ 2*
* Js
UJ =
o
UJ
o -r1
^ -S •»
t? S —
2 a:
= «
5
en
•^
or
«> _¥
^5
"2 °
i
_j
•o
£
^ H
£ 'S

n
Z 0
5 •«>
2 s
o 3
0- o S,
55"
^ — 0

t-->

-T 75
S °
o *3>


o °
"5 °
S 2
t/1
_J
*" — .
UJ o ^
S *

*
T5 1
""=>

„

| 0
1 0
o H-



£ s
S ; ^
£ u o
< 0
-J
(^J
£• u
•r o
E H-
o
OJ

r^-
CO
r^
k
S
m
r»
r**
r^
CM
r^
f*»
O
r*.
(?)
(O
oo
U)
r-
u>
1C
in
in
(£>
^
1C
f)
(O
rj
to
to
o
lO
o>
m
ou
\n
if)
ID
m
m
«T
in
u-
tN
un
in
?
m
a
v
to
*s
m
*T
5
^r
rsi
v
0
s
00
r^.
u
LA
fi
S
ro
ro
CN
fO
fO
O
CTJ
fsl
CO
(N
K-
rN
1C
(NJ
LT)
rs(
rsi
(*
rsi
rs
Csl
O
(Nl
C
CO
r^
^
L
«
n
Cs
^
O
-*
<





































































w G
£
uoipv ^
r^
5M i S
"C £ S
" -

* - £
H E
o o r-

^ * t ~ «
"a^Sa

s- S
0 < c ,
r-J >= u
- gm <-
a. ° a

C3

t/>
<






J

>




1
3
r
•>
•)
9
4
}
3
5
1
1
3
•i

^

•

1
•>
J
n
t
0
t
f
•
'
t
t
t
r
t
r
•>
j

>
1
r




•
X

c
r
t


^
=

3
^
J

^
"
n
I
3
•o
u

wipV
































































U-,
Z
£

O
CO
n
0
^
D
n

Ul
u>
V
t£>
tc
(N
«>
U>
O
VD
en
m
m
m
r^.
in
to
m
m
Lf)
«
m
n
m
CN
in
in
o
in
en
^f
oo
«•
r»-
^
s
m
v
*t
*t
CO
^
C"s*
*t
*
C
en
CO
03
r*)
f
V3
m
S
«•
fO
n
n
(M
C
ro
C
en
p>»
CO
r>
r-x
r^
ID
r\j
m
CJ
r
ro
rg
(NJ
rsj
(N
S
C
CO
r-
*
i
*
ii
rs,
^
°
(O
<





































































31

-------
PO
 I
CM
      i
B
M
O)
CO


S
                i
                ss
                                 NOI13V
                                  !=xg
gs

NOtiOV

















«
Z
W '
Z
O











a "
iz
E < **PV.
>g
E0 J
HRUPUT
JUNEjSEPT
ANNUAL 1
DEC MAR,
FFH 1M»Y 1
C
£
•u =
§|
t
0
u
CdOM

• £
3
£
P
^
S
a.
10



t? '
*
«


s
"
-fe-
(D
U
S


O
.

o
c
^5«





>, ui
l. 4



: w
§5


E « §
J
i>

8 =



-
jo uv:
t
£
^
•j
?
^
10
a
£
P
?
f-
g
ffi
(O
o
o
8
s


o
o
(O
0
CD
0



ci
to
a
s
9
?
T-
9
S?
$
«?
«¥
—
0
a
5
S
u
R
v
P
S
-
c
s
s
fj
S]
ft
?•

<;
—
o
«
e
t
n
»
a.
»


































































5S
j ^
NOU.3V













*
z
z
0












OOHi3H

i
u
TON/ TR
+ -
OOHA3H




VI
VI
P«RAMETtF<
+ -
QOHJ.3N





CONTROL
EQUIP.
EFFICIENCY
t'OMin"!"


















\
a.
_£
t! =
n
s
oaiu

§ ~
o
O
^

























K
lo
o
!o
^
*
0.
o































T1
$
*
*
•r
¥
.X.

O)
10
?t
Is
n
R
^
V.
ss
n
t;
s
R;
Sj
R
N
_i
^
N
O






— ^
^



—4




^

















O UJ
cc o
a u
KOIiDV

















« l
Z >
z .
o (
o '





































'
o
tP
(P
3|
8


iil
O
O
O
S
£3
5
*£~
rt_
o
fi
^
S
g
?
^.
$
9
S
*_
^
J
O
«*
^
R
^
R
?
K
K
Irt
S
S
s
rV
^
A











a.
«
















































































































£s
< 1
«.-
1°
0 °
£ g

• S»
3o£
ss§
i"S
SS;
d-5
= |!
» 2 *
1 S 5


                                                    32

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
CM
cn
       O-
       Z
       u
       ce

       O
o
en
to
0 h-
ID 1C
i in
*• i
io>


15T
'"T
5!
pj
—
o
en
CO
o"-
TO l£
1 in
iy^j.
ro
eg
—
O
ov
5?
o1^
N U>
1^)
— 1
fO
PO
—
o
01
CD
r^

i m
i
ro
f\J
-


NOI13V














r

E

0
U











w» "T""
° 2
s >-
•5^-
E i-


ttJ £" k-
i! > £
o t: o
^<"-

£
a

v
\-
o
LO
etu-H
'°K
c
0 V
= "8
"S w
^
-)CW


«DDV


XiuftO"^


94 DiC

<
—
<







-




.






"

























































-




.






~





































































































































































































-




-





-
•




.











































































































































































































































































-



































































































































































































































































































































































































































•






-








i














































































































































~










































































































































































































.
-






~


































                                                33

-------
cards 4P, 5P, and 6P are for multiple processes per point,  multiple  pol-
lutants other than lead per process, and multiple comments  at the point,
process, and pollutant level.

     The HATREMS area source form is utilized to supplement NEDS  area  source
data.  The form describes only one card which may be input  for each  area
source category pollutant combination per county which is  reported.

3-2.3  Coding Procedures

     To insure that the data are correctly coded, the coding procedures for
NEDS and HATREMS which appear in AEROS Users  Manual 1 should be utilized.
Codes which are necessary to complete the forms are located in AEROS Manual
of Codes.2

     Table 3.2-1 lists the data items which must be collected.  Since
HATREMS utilizes NEDS for specific data items, sources of  lead must  be
coded and updated in NEDS as well as HATREMS.  Since NEDS  coding  procedures
have been utilized for several years, no further discussion will  be  given.
The control agency must insure that the NEDS  data are correctly coded
because they will be utilized by HATREMS.

     The coding procedures for the HATREMS point source form are  given in
Section 3.6.2 of AEROS User's Manual.1  These instructions  are general for
all HATREMS coding, and the following additional instructions should also
be followed for lead:

     (1)  Since HATREMS utilizes NEDS data and a NEDS form must be submitted
for each lead source, cards IP and 2P are not required for lead sources  in
NEDS.  Card 3P is not required unless a third control device exists.  In
order for HATREMS to utilize NEDS data, the HATREMS identifier (state,
county, plant, point, SCC) must be the same as the NEDS identifier.
                               34

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
     (2)  All individual data items on cards 4P and 5P must be manually

updated.  When any data on 4P or 5P are updated, the remaining data for

cards 1P-5P of that point are transfered from NEDS to HATREMS.  After these

data are transfered from NEDS, all  future updates for that point must be

manually input utilizing the HATREMS point source form and the correct

update action code.


     The coding procedures for the HATREMS area source form are given in

Section 3.6.1 of AEROS User's Manual.'   These instructions are general for

all HATREMS coding and apply for lead data.


3.2.4  Data Flow


     As discussed in Section 3.2.1, the HATREMS point and area source data

files are created from the NEDS point and area source data.   The States

will be provided this initial emission inventory and will be required to

utilize the coding procedures from Section 3.2.3 to compile the lead inven-

tory for SIP development and for reporting to EPA.   The data flow for the

lead emissions data is presented in Figure 3.2-5.


     The States will be provided with the following reports to assist

them in compiling an emissions inventory for lead:   (1) HATREMS point

source report for lead point sources from the NEDS data base, (2)  HATREMS

emission summary report for lead point and area sources from the NEDS

data bases, and (3) a plant name report for all plants in NEDS.   The

State should initially review the point source and emission summary

reports to ensure that all sources  are included.   If a known lead source

is missing, the plant name report should be reviewed to determine if it

is in NEDS.  The source could be in NEDS but missing from HATREMS if the

source classification code is incorrect or if the source classification

code is not in the HATREMS point source emission factor file.
                             35

-------
e\j
  •
CO
         i—i
•r-    S!
U_    LU


      o
                                                              36

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
     The States must code and submit NEDS and HATREMS point source

forms for all sources of lead that emit five tons or more per year

to update the emission inventory to the common base year which was

utilized in the strategy development.  The States must code and submit

NEDS and HATREMS area source forms for all counties for which the data

were collected for the demonstration of attainment.  The NEDS forms

must be completed to update existing sources and add new sources to

produce a lead emission inventory for a common base year.  The HATREMS

form must be completed to supply data which can not be stored in NEDS.


     The NEDS and HATREMS forms or computer readable format will be sub-

mitted to the appropriate Regional Office.  The NEDS data will be routinely

processed utilizing the procedures defined in AEROS Users'  Manual.  These

data will be updated on the standard monthly update schedule.  The HATREMS

data will be converted to computer readable format and forwarded to the

National Air Data Bank.  These data will be updated to HATREMS on a

standard schedule.


     The HATREMS report programs will be provided to the Regional Office

NEDS contact to provide data requests to States as well as  the other user

community.  New report programs will be developed as requested by the user

and identified as being useful.
                                     37

-------
3.2.5  References for Section  3.2

1.  AEROS Manual  Series,  Volume II:   AEROS  Users' Manual, EPA-450/2-76-029,
    December 1976, USEPA, OAQPS.
2.  AEROS Manual  Series,  Volume V:   AEROS Manual  of  Codes,  EPA-450/2-76-
    005, April  1976,  USEPA,  OAQPS.
3.  AEROS Manual  Series,  Volume III:   Summary  and Retrieval, EPA-450/2-76-
    009, May 1976, USEPA, OAQPS.
4.  Development of HATREMS Data Base and  Emission Inventory Evaluation,
    EPA-450/3-77-011, April, 1977,  USEPA, OAQPS.
                                38

-------
I
I
I
I
I
I
I
I
I
I
I
            4.0  ANALYSIS AND CONTROL STRATEGY DEVELOPMENT

     As indicated above in section 2.2, detailed guidance on the process

of analysis and control strategy development appears in EPA's Air Quality

Analysis Workshop, Volume I—Manual.

4.1  BACKGROUND CONCENTRATIONS
                                                                  2
     Natural concentrations of lead in the air have been estimated  to
                    3
be about 0.0006 pg/m  resulting mainly from airborne dust containing

                     3               4
10 to 15 ppm of lead.   Chow, et al.,  have recommended that 0.008

ug/m  be considered as the baseline concentration for atmospheric
                lead for the continental United States.  Most areas, however, will
experience greater concentrations that cannot be accounted for by

sources in the immediate vicinity or by natural  background.   The

majority of these concentrations not accounted for by nearby sources

or background can probably be attributed to airborne lead transported

from sources outside the study area in question.  Therefore, a plan

to control emissions in the study area will not reduce the concentra-

tions due to the outside sources.  To account for the phenomenon of

transport of lead particulate matter from outside the study area,
                II
                 Cirillo, R.R., et al., Air Quality Analysis Workshop, Volume 1--
                 Manual.  Prepared for the Environmental Protection Agency, Office
I                 of Air and Waste Management, Office of Air Quality Planning and Stand-
                 ards,  Research Triangle Park, N.C. 27711.  November 1975.
                 (EPA-450/3-75-080a).
                2
I                 Patterson, C.C., Contaminated and Natural Lead Environments of Man.
                 Archives of Environmental Health.  11:334-363, 1965.
                3
                 Chow, T.J., and C.C. Patterson.  The Occurrence and Significance of
I                 Lead Isotopes in Pelagic Sediments.  Gsochim.  Cosmochim. Act. (London).
                 2^:263-308. 1962.
                4
                 Chow, T.J., et al., Lead Aerosol Baseline:  Concentration at White
                 (Mountain and Laguna Mountain, California.  Science 178:401-402.
                 October 1972.



I

                                                   39
I

-------
States should assume a "background" equal  to the levels of airborne
lead in a representative nonurban area that is not significantly influ-
enced by stationary or mobile lead sources.
     For the purposes of SIP development,  States may account for the
future reduction of background lead concentrations that may result
from the federal programs for the reduction of lead in gasoline, the
prohibition of the use of leaded gasoline  in catalyst-equipped vehicles,
and reduced gasoline consumption in vehicles in future years.
     States should discuss their choice of background lead concentra-
tions with cognizant persons in the appropriate EPA Regional Offices.
4.2  LEAD EMISSION FACTORS
     In performing the analysis to determine whether an area needs
a control strategy to attain and maintain  the national air quality
standard for lead, a State will have to compile an inventory of lead
emission sources and the quantity of emissions produced by each source.
Emission factors that relate source activity to the quantity of emis-
sions appear in EPA's Control Techniques for Lead Air Emissions,  and
two other reports concerning lead emissions.     One of these reports
also describes procedures by which States  can quantify fugitive lead
emissions from stationary sources; that report also provides additional
          891011
references '''   on the quantification  of fugitive emissions.  The other
report  provides emission factors for resuspended lead dust from paved
roads.
 Control Techniques for Lead Air Emissions.  PEDCo Environmental,
 Inc., Cincinnati, Ohio.  Prepared for the U.S. Environmental Protec-
 tion Agency, National Environmental Research Center under Contract
 No. 68-02-1375, Task Order No. 32.  January 1977.   Report No.  EPA-450/2-77-012.
                               40

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
 Zoller, John M., et al.  A Method for Characterization  and Quantifica-
 tion of Fugitive Lead Emissions from Secondary  Lead  Smelters, Ferro-
 alloy Plants, and Gray Iron Foundries.   Prepared  by  PEDCo Environ-
 mental, Inc., for the U.S.  Environmental  Protection  Agency, Office of
 Air and Waste Management, Office of Air Quality Planning arid Standards,
 Research Triangle Park,  NC 27711.   January  1978.   EPA-450/3-78-003.
 (Currently under revision; copies  not available as of  this printing.)

 Maxwell, Christine M., and Daniel  W.  Nelson.  A Lead Emission Factor
 for Reentrained Dust from a Paved  Roadway.  Prepared by Midwest
 Research Institute for the U.S. Environmental Protection Agency, Office
 of Research and Development, Industrial Environmental  Research Labora-
 tory, Research Triangle  Park, NC  27711.  April 1978.  EPA-450/3-78-021.

technical Guidance for Control of Industrial Process Fugitive Parti-
 culate Emissions, PEDCo  Environmental,  Inc., Cincinnati, Ohio.  Pre-
 pared for the U.S. Environmental Protection Agency,  Office of Air
 Quality Planning and Standards, under Contract  No. 68-02-1375, Task
 Order No. 33.  Publication No. EPA-450/3-77-010.   March 1977.
g
 Technical Manual for the Measurement of Fugitive  Emissions:
 Upwind-Downwind Sampling Method for Industrial  Fugitive Emissions.
 U.S. Environmental Protection Agency, Industrial  and Environmental
 Research Laboratory, Research Triangle  Park, N.C., April 1976.
 Publication No.  EPA-600/2-76-089a.

  Technical Manual for the Measurement of Fugitive Emissions:  Roof
 Monitor Sampling Method for Industrial  Fugitive Emissions.  U.S.
 Environmental Protection Agency, Industrial and Environmental
 Research Laboratory, Research Triangle  Park, N.C., May 1976.
 Publication No.  EPA-600/2-76-089b.

  Technical Manual for Measurement  of Fugitive Emissions:  Quasi-Stack
 Sampling Method for Industrial Fugitive Emissions.   U.S. Environmental
 Protection Agency, Industrial and  Environmental Research Laboratory,
 Research Triangle Park,  N.C., May  1976.   Publication No. EPA-600/2-
 76-089C.
                              41

-------
1
1

1
1

1

1
1
^M
1

1

1
•
*
1

1
1
•
1

1



1

1

4.3 PROJECTING AUTOMOTIVE LEAD EMISSIONS
Lead emissions from mobile sources are calculated based on emis-
sions at different vehicle speeds, the lead content of gasoline, and an
average fleet fuel economy. The lead content of gasoline and fuel
economy are a function of the calendar year of interest. The background
for the computation procedure presented below appears in Appendix C of
this guideline. Appendix C also presents lead emission rates for seven
example roadway configurations.
4.3.1 Lead Emissions from Automobiles

4.3.1.1 Individual Roadways—The emission rate from automotive sources
from an individual roadway (line source) is calculated by the following
equation:
en,s ' asPbnT
TTT (D
n,s x
where: en s = emission rate for calendar year n and speed s (g/road
mile-day);
ag = percentage of lead burned that is exhausted; available
from Figure 4.3-1 (nondimensional; expressed as a decimal);
for roadway portions subject to full-throttle acceleration
(0-60 mph), assume a « 10.0;
Pbn = probable pooled average lead content of gasoline in year
n from Table 4.3-1 (g/gal);
T = average daily traffic (vehicles/day);
fn s • average fleet fuel economy for calendar year n and speed
s; calculation described below (vehicle-road mile/gal).
To calculate the emission rate in units of grams/meter-second, e« _ can
n,s
o
be corrected by dividing by 1.39 x 10.

43


-------
The term, f    is calculated by the following equation; the calculation
           n ,s
is based on a base year of 1974:
   n,s
i = 1967  [Cs,iEc,imi]
                     -74
EnCt
                                                 (2)
     :  C  .  = speed-dependent fuel economy correction factor for model
        s,i
              year i; calculation is described below in equation (3)
              (nondimensional);
       E- ,•  = city/highway combined fuel economy for model year i from
        c > i
              Table 4.3-3 (mi./gal.);
       m.   = fraction of annual travel by model year i vehicles from
              Table 4.3-4; assume 1974 model year vehicles are one year
              old, 1973 model year vehicles are two years old, etc.
              (nondimensional);
       Ej.  = base year (1974) fuel economy from Table 4.3-2 (mi./gal.);
       E    = average fleet fuel economy for projection year n from
              Table  4.3-2 (mi./gal.);
       C.   = traffic flow correction factor; C. = 1.2297 for free-
              flow traffic; C.  = 0.866 for city (stop-and-go) traffic
              (nondimensional).
     C_ j, the nondimensional speed-dependent fuel economy correction
      s ,1
factor for model year i, is calculated by the following equation:
                                 44

-------
I
I
                 Cs,i
I             where: A = correction factors from Table 4.3-5;
_                    S = vehicle speed  (miles/hour). [Note: S  =1.]
•                  To simplify computation, the values of C   . are reproduced in Table
                                         1974                S>1
               14.3-6 and the values of   "">    C  .E  -m. are reproduced in tabular
                                       i=1967   Sjl Cj1 ^
               form in Table 4.3-7 and in graphical form in Figure 4.3-2.
•             4.3.1.2  Area Source Automotive Emissions—Equation (1) may be used to
               calculate automotive emissions as area sources rather than specific line
•             sources, but the term "T" should be replaced, with the term "V", the
•             daily vehicle-miles traveled (VMT) in the area  (vehicle-miles/day).
               This substitution enables the user to employ VMT data, which is more
I             readily available for area mobile sources than average daily traffic.
               Also, the term en   will  now be expressed in g/day.  The determination
                               In ,s
               of a  and f    should be  based on the average vehicle speed for the
                   s      n 5 s
•             specific area.
               4.3.2  Lead Emissions from Other Gasoline Powered Vehicles
I                  Motorcycles and diesel -powered vehicles are assumed to emit quantities
               of lead that are insignificant compared to other gasoline-powered vehicles.
|                  There are no known measurements of lead emissions from either
•             light- or heavy-duty trucks.  Therefore, for purposes of calculating
™             emissions, the percentage of lead burned that is exhausted from these
•             vehicles at various speeds is assumed to be the same as that for automobiles
               (Figure 4.3-1).
I

I

I

-------
     Light-duty gasoline-powered trucks are assumed to have the same
gasoline economy as automobiles; new light-duty trucks are assumed to
require the use of non-leaded gasoline to meet emissions standards for
CO and hydrocarbons through the use of catalysts.   Therefore, the emission
rate for light-duty gasoline-powered trucks is calculated using the same
procedures and parameters as for automobiles.
     Heavy-duty gasoline-powered trucks are assumed to burn leaded
gasoline for all future years.  Also, their fuel economy is different
from that of light-duty trucks.  Therefore, the emission rate for heavy-
duty gasoline-powered trucks is calculated by equation 1, but the following
parameters are modified:
     Pb  = probable lead content of leaded gasoline in year n from
           Table 4.3-8.
     f   = average fleet fuel economy in calendar year n = 5.7 miles/gal
           (taken from Kennedy, G.J., et al.  Exhaust Emissions from
           Heavy-Duty Trucks Testing on a Road Course and by Dynamometer.
           Society of Automotive Engineers, Automobile Engineering
           Meeting, Detroit, Mich., October 13-17, 1975).
                                  46

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4.3.3  Example Calculation of Automobile Lead Emissions

     Problem:  For a city street with a speed of 16  miles  per hour and

average dally traffic of 28,000 vehicles, calculate  the  lead emission  rate

for the year 1983.

     Solution: We use Equation (1):
  e83,16 = a!6Pb83T
             r83,16
                             (1A)
     From Figure 4.3-1, for a cruise speed of 16  mph,  we find that  approxi-

mately 10.5 percent of the lead being burned 1s emitted.   Therefore,  a-ig  =

0.105.

     From Table 4.3-1, we find that Pbg3, the probable pooled average

lead content for 1983, 1s estimated to be 0.25 g/gal.

     We are given the average daily traffic, T, as  28,000 vehicles/day.

     We must now calculate the average fleet fuel economy for 1983  by using

Equation 2:

               L4
   83,16
            1^-1967
E83Ct

 E74
                                                                  (2A)
     To do this, however, we must first  calculate  C,g  ^ ,the  speed-dependent

fuel economy correction factor for model years  1967  through  1974  using

Equation (3):
        -
                            (3A)
                                 47

-------
The A. coefficients appear in Table 4.3-5.   Table 4.3-9 presents the



results of that calculation.   That table also presents the values for



E  .  from Table 4.3-3 and m.  from Table 4.3-4 for each model year,
 ^51                        I


together with the product of these three factors and the sum of the



products.  As Table 4.3-9 indicates,

   1974


   2    CTK iEr ,-m, = 11.34.

  i=1967   ID'1 Cs1 ]





The factors £-,. and EQ~ from Equation (2A)  are found in Table 4.3-2;



E7A = 12.4 vehicle-road mi./gal. and EQ- =  19.1 vehicle-road mi./gal.
 / *T                                   GO


Since the roadway is a city street, the traffic flow correction factor



C.  =  0.866.  Substituting this information  into Equation (2A), we obtain:







  fB, -,,  = 11.34 x 19.1 x 0.866
   OO » I D    --I--U--..L .!--.---

            12.4



          = 15.1 vehicle-road mi./gal.



Substituting the above results into equation (1A), we obtain:
  e83 16 = 0-105 x 0.25 g/gal x 28 x IP3 vehicles/day


                   15.1 vehicle-road miles/gal.
         = 48.7 g/road mile-day.



In units of g/m-sec, this becomes
               48.7	  = 3.50 x 10"7 g/m-sec.
                     p~

            1.39 x 10°
                                     48

-------
                        I 1
                                  73
                     Tv!
     TJJ


            44--
            ±t
                                     4-
                                       —\-\
                            41=
   344
4ff
-nh
                     lip
                                              JZ
                  -TT
                                                tt
                                                                i
          --n-
    ! iLJi
                                       "rr
j_
          4-

                             "H"
                              TS1F
                                                 2'I
   XPT-
                                              -4-H-
                                            -i-
      it:

                                                    :tt:

                                                           it
                                       49

-------

-------
1
1
1

1
1
1

1

1
1

1

1
1
1

1
1
1
1
1
1





Year
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990






TABLE 4.3-1
PROBABLE POOLED AVERAGE
LEAD CONTENT OF GASOLINE
(grams/gal)
Lead Content
2.0
1.7
1.4
1.0
0.8
0.5
0.5
0.5
0.34
0.25
0.19
0.15
0.13
0.11
0.09
0.08
0.05



51

-------
                                                                               I
                                                                               I
                            TABLE 4.3-2
                    AVERAGE FLEET FUEL ECONOMY                                 |
                           (miles/gallon)

            Calendar Year                   Fuel Economy
               1974                            12.4
               1977                            13.3
               1978                            14.0
               1979                            14.8                             |
               1980                            15.7
               1981                            16.8
               1983                            19.1
               1985                            21.7
               1990                            26.2                             •
               1995                            27.4
Ref:   U.S.  Environmental Protection Agency, A Report on Automotive
       Fuel  Economy, Washington, D.C., February, 1974.
       15 USC 2002,  enacted December 22, 1975.                                  I

                                                                               I

                                                                               I

                                                                                I

                                                                                I

                                                                                I

                                                                                I

-------
I

I

I                                          TABLE 4.3-3
•                               CITY/HIGHWAY  COMBINED FUEL ECONOMY
                                           (miles/gallon)

™                         Model  Year                          Fuel Economy
•                            1974                               15.15
                              1973                               14.89
I                            1972                               15.20
                              1971                               15.24
•                            1970                               15.42
•                            1969                               15.47
                              1968                               15.60
|                            1967                               16.15

I

I

I

I

I

I

I

I

I

-------
                            TABLE 4.3-4
                FRACTION  OF ANNUAL LIGHT-DUTY VEHICLE
                       TRAVEL BY MODEL YEAR
            Age                                 Fraction of
           Years                               Annual Travel
Ref:   AP-42, Supplement  5
            1                                      .112
            2                                      .143
            3                                      .130
            4                                      .121
            5                                      .108
            6                                      .094
            7                                      .079
           >8                                      .213
                                54

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                             TABLE  4.3-5

            FUEL ECONOMY CORRECTION FACTORS  BY MODEL YEAR

                   (NORMALIZED TO 32.7 MILES/HOUR)
Model Year
Pre-controlled 2.4697E-2   7.55258E-2

   1968        1.75941E-2  6.90954E-2

   1969        3.43032E-3  7.24956E-2

   1970        6.00124E-3  6.90443E-2

   1971        9.76255E-3  6.84494E-2

   1972        8.57745E-2  7.0882E-2

  1973-74      6.29988E-2  5.96559E-2
-2.42452E-3  4.01469E-5

-2.01359E-3  3.19426E-5

-2.18976E-3  3.54015E-5

-1.98463E-3  3.13931E-5

-1.96781E-3  3.13719E-5

-2.15219E-3  3.57324E-5

-1.59874E-3  2.59441E-5
Ref:  AP-42, Supplement  8

Fuel Economy Correction  Factor  = AQ+A1S+A2S2+A3S3+A4S

                      where  S  = vehicle speed
                            A/L
-2.68893E-7

-2.12343E-7

-2.36485E-7

-2.09286E-7

-2.11167E-7

-2.44316E-7

-1.84877E-7
                                  55

-------
LO   OO
 I    LU
CO   =>
CO
*^
CO
o>

CM
r-.
^T\
t^ i

r""
(^TS
\J1

0
/TX
w i

cn
10
rr\
tj i

00
co
fT\
u 1

("•^
10
^v
U 1
cc
UJ
^





*!• «3- LO LO
CM CM r-« cn
co LO LO r^
r— co co cn
cn r— ^ co
CO LO ^ CO

r^ p~- cn cn
O CM 00 O
CO LO LO CO

LO t^ C3 l~~
O CM CT> r—
CO LO LO CO

LO CM <0 CO
I — •* O CM
CO LO I*- 00

r^ i*^ oo LO
r— CO CT> r-~
CO LO LO CO

f^* LO ^^ *J
^j- f-^ CO ^"
CO LO 1 — CO


ac LO o LO o
Q- r— r— CM
o
Ij -1
LU
Q.
OO

CO
CO
00
LO

cn

cn
cn
CO

o
O
cn

CO
o
cn

co
o
cn

O
CM



LO
CM





LO
LO
cn
CM

o

CO
LO
cn

cn
LO
cn

CM

cn

co
LO
cn

LO
px.
cn


o
CO





^^
CM
O
CO
cn
o

co
CM
O

CO
CM
O


CM
O

O
CM
O

LO
r—
O


LO
CO





CO

0
cn
CO



LO
0


10
o

o
LO
o

o
10
o

CO
^^
0


0






LO
^«
r—
r—
CM
f-.
*—

00
cn
O


cn
o

CO
CO
o

cn
CO
o

cn
LO
O


LO
^





LO
co
I—
cn
CO


*t

i—

CO
r™"
f—


O
r—


O
•—


pv»
o


o
LO





co
co
r—
r™
CO
CO


cn
o


CO
o
r—

co
cn
o

CO
cn
o

CO
LO
o


LO
LO





LO
cn
o
CO
^«
*—

CM

O

co
1^
o

CM
LO
O

CM
V0
O

LO
l*—
O


o
10




                                                       56

-------
I

I
                                           TABLE  4.3-7
I                                        W
                               VALUES OF  4%7   CS).  ECf1

                            Speed (MPH)               Value
I                              5                      5.13
                               10                      8.49
I                             15                    10.95
•                             20                    12.74
                               25                    14.08
•                             30                    15.11
                               35                    15.92
I                             40                    16.57
-                             45                    17.04
"                             50                    17.27
•                             55                    17.16
                               60                    16.54
I

I

I

I

I

I

I

I

-------
                        TABLE  4.3-8

         PROBABLE LEAD  CONTENT OF  LEADED  GASOLINE*
                        (grams/gal)
Year
1974
1975
1976
1977
1978
1979
1980
1981 and beyond
Lead Content
2.0
1.9
1.9
1.6
1.6
1.2
1.6
2.0
*Source:   Appendix C of this  Guideline,  Table  5-
                            58

-------
1
1
1
1
1
••


1

1

1
1

1
1
1
1
1
1
1
1





TABLE 4.3-9
CALCULATION OF 1 ^67 Cl6,1Ec,1m1 FOR EXAMPLE CALCULATION



Model Year
pre-1968
1968
1969
1970
1971
1972
1973
1974









Fraction of
Fuel Economy City/Highway Annual LDV
Speed Correction Combined Fuel Travel by
Factors Economy Model Year
C16,1 Ec,1 mi
0.759 16.15 0.213
0.725 15.60 0.079
0.732 15.47 0.094
0.718 15.42 0.108
0.716 15.24 0.121
0.799 15.20 0.130
0.702 14.89 0.143
0.702 15.15 0.112
2






59

1974 Fuel
Economy
at 16 mph
Cl6Ec,im1
2.61
0.893
1.06
1.20
1.32
1.58
1.49
1.19
= 11.34









-------
4.4  AIR QUALITY MODELING
4.4.1  Modeling In Certain Areas
     Sections 51.83 and 51.85 of 40 CFR Part 51  require a demonstration
of attainment using a modified rollback model  as a minimum, but pro-
vide that a dispersion model  may be used if desired.   Section 51.83
requires the analysis of an entire urbanized area if that area has
                                                  o
measured lead concentrations  in excess of 4.0 ug/m,  quarterly
mean, measured since January  1, 1974.   Section 51.85  requires an
analysis of the area in the vicinity of any air quality monitor that
has recorded lead concentrations in excess of the lead national stand-
ard concentration.
     Based solely on national data available to EPA,  EPA estimates
that there are few areas that have concentrations greater than the
standard concentration and fewer still with concentrations greater than
4.0 jjg/m , quarterly mean.  There are other data available to State
and local air pollution control agencies, however, that may indicate
that other areas have concentrations in excess of the concentrations
specified in the criteria for performing the analysis.  Many areas have
already been analyzed for other pollutants using dispersion models and
therefore, some of the work needed to run a dispersion model for lead,
such as allocation of area source emissions, may have already been
done.  For such areas, EPA encourages appropriate States to employ a
dispersion model where possible, since such a model will generally
yield more accurate estimates of air quality concentrations.
                                60

-------
I

                   Where States elect to use modified rollback, however, it must
                                                                      112
              be one of the versions described by de Nevers and Morris   or an
              equivalent.  The reprint of the de Nevers and Morris paper appears as
•            Appendix G of this Guideline.
•            4.4.2  Modeling in Areas Around Significant Point Sources
                   40 CFR 51.84 requires State implementation plans to contain a
I            calculation of the maximum lead air quality concentrations and the
              location of those concentrations resulting from the following point
|            sources for the demonstration  of attainment:
«                 --Primary lead smelters.
                   --Secondary lead smelters.
I                 --Primary copper smelters.
                   --Lead gasoline additive  plants.
|                 --Lead-acid storage battery manufacturing plants that produce
_                   1200 or more batteries  per day.
™                 --Any other stationary source that emits 25 or more tons per
•                   year of lead or lead compounds.
                   The regulation requires that a dispersion model be used for these
I
I
analyses.  The dispersion models that are acceptable are described in
the section on point source models for sulfur dioxide and particulate
                                                1 3
matter in EPA's Guideline on Air Quality Models.
              12
                de Nevers,  Noel,  and Roger Morris,  Rollback  Modeling—Basic and
               Modified.   Paper no.  73-139, presented  at  the 66th  Annual  Meeting
 Inuu i i ieu.   reader  nu.  /j-io^, prtibenueu O.L trie ooui Miiriua i  net
 of the Air Pollution  Control Association, Chicago, Illinois,
 June 24-28, 1973.
13
  •Guideline on Air Quality Models.   Monitoring and Data Analysis Division,
 Office of Air Quality Planning and Standards, U.S. Environmental  Protec-
 tion Agency, Research Triangle Park,  N.C.  EPA-450/2-78-027 (OAQPS No.
 1.2-080).   April, 1978.
I
                                                 61

-------
     These models do not account for deposition of large particles
that may be generated by such sources,  however.  States  that  wish to
account for such deposition may use the methods described in  Appendices
D and E of this Guideline.   Appendix D  is an exerpt from Meteorology
and Atomic Energy,   and provides a description of a source depletion
model.  Appendix E is a reprint of an article in Atmospheric  Environment,
and provides a description  of a surface depletion model, which is com-
putationally more complex than a source depletion model, but  more
accurate.  The article provides comparisons of the two models.  Per-
sons interested in the surface depletion model are also directed to
a further discussion on limitations of that model, which appeared in
a later edition of Atmospheric Environment.
  Slade, David H. (ed.), Meteorology and Atomic Energy 1968.   Prepared
  by Air Resources Laboratories, Research Laboratories, Environmental
  Science Services Administration, U.S. Department of Commerce, for
  the Division of Reactor Development and Technology, U.S.  Atomic
  Energy Commission, Oak Ridge, Tennessee.  July 1968.  Pp. 202-208.
  Horst, Thomas W., "A Surface Depletion Model for Deposition from a
  Gaussian Plume."  Atmospheric Environment, Vol. 11, pp.  41-46.  1977.
  Doran, J.C., and T.W. Horst, "Long-Range Travel of Airborne Material
  Subjected to Dry Deposition and a Surface Deposition Model  for Deposi-
  tion from a Gaussian Plume."  Atmospheric Environment, Vol. 11,
  p. 1246.  1977.
                                 62

-------
 I
 •                                5.0  AMBIENT LEAD MONITORING
             5.1  INTRODUCTION
 |               According to the requirements for State Implementation Plans (SIPs),
 _          two air quality lead monitors are required in each urbanized area (as
 •          defined by the U.S. Bureau of the Census)--
 •               --that has a 1970 population greater than 500,000; or
                  --where lead air quality levels currently exceed or have exceeded
 I          1.5 yg/m3, quarterly arithmetic mean, measured since January 1,  1974.
             Lists of areas that meet these criteria appear in Tables 5.1 and 5.2.
 I          Table 5.2 is based only on limited data available to EPA, however.   There
 •          are other data available to State and local air pollution control agencies
             that may indicate that other areas have concentrations in excess of 1.5
 •          yg/m3, quarterly arithmetic mean.  In addition, EPA can require  additional
             monitors in those areas and any other areas.
 •               The monitoring networks operated by State and local agencies will
 •          consist of sites in three general categories:  State and Local Air Monitoring
             Stations (SLAMS); National Air Monitoring Stations (NAMS), which  are a  subset
 I          of SLAMS; and Special Purpose Monitoring (SPM).
                  The strategy for State and local agencies to perform monitoring
 |          under these three categories was developed by the Standing Air Monitoring
 _           Work Group (SAMWG) and is more fully described in "Air Monitoring Strategy
 •           for State Implementation Plans," EPA-450/2-77-010, OAWM, OAQPS,  RTP, N.C.
 •           June, 1977.
                  Although the SAMWG did not specifically address monitoring  of ambient
 J           lead, the concepts developed by the SAMWG were incorporated into the
             requirements for State Implementation Plans (SIPs) for lead.  The termi-
•           nology of the SAMWG concerning the site categories (viz., SLAMS, NAMS, and

I
I

-------
                               TABLE 5.1

           URBANIZED AREAS* GREATER THAN  500,000 POPULATION
                            (1970 Census**)
AQCR#              AREA                 AQCR

043   New York, N.Y.-Northeastern       106
      New Jersey                        015
024   Los Angeles-Long Beach, Calif.     193
067   Chicago, 111.-Northwestern        244
      Indiana                           080
045   Philadelphia, Pa.-N.J.             120
123   Detroit, Mich.
030   San Francisco-Oakland,  Calif.      176
119   Boston, Mass.                     217
047   Washington, D.C.-Md.-Va.           078
174   Cleveland, Ohio                   173
070   St. Louis, Mo.-111.               215
197   Pittsburgh, Pa.                   223
131   Minneapolis-St. Paul, Minn.       018
216   Houston, Texas                    028
115   Baltimore, Md.                    050
215   Dallas, Texas
239   Milwaukee, Wise.                  160
229   Seattle-Everett, Wash.             024
050   Miami, Fla.
029   San Diego, Calif.                 184
056   Atlanta, Ga.                      004
079   Cincinnati, Ohio-Ky.              174
094   Kansas City, Mo.                  049
162   Buffalo, N.Y.                     042
           AREA

New Orleans, La.
Phoenix, Ariz.
Portland, Ore.-Wash.
San Juan, P.R.
Indianapolis, Ind.
Provi dence-Pawtucket-
Warwick, R.I.-Mass.
Columbus, Ohio
San Antonio, Texas
Louisville, Ky.-Ind.
Dayton, Ohio
Fort Worth, Texas
Norfolk-Portsmouth, Va.
Memphis, Tenn.-Miss.
Sacramento, Calif.
Ft. Lauderdale-Hollywood,
Fla.
Rochester, N.Y.
San Bernardino-Riverside,
Calif.
Oklahoma City, Okla.
Birmingham, Ala.
Akron, Ohio
Jacksonville, Fla.
Spri ngfield-Chi copee-Holyoke,
Mass.-Conn.
 *As defined in U.S. Bureau of the Census, "1970 Census Users'  Guide;"
  U.S. Government Printing Office, Washington, D.C., 1970 (p.  82).

**U.S. Bureau of Census, "U.S. Census of Population: 1970; Number of
  Inhabitants; Final Report PC (1)-A1; United States Summary."   U.S.
  Government Printing Office, Washington, D.C. 1971.
                                   64

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
                              TABLE 5.2

             URBANIZED AREAS  WITH  LEAD-AIR CONCENTRATIONS
        EXCEEDING OR EQUAL  TO 1.5  jjg/nT, MAXIMUM QUARTERLY MEAN
                                (1975)


       :R#                     AREA
     004          Birmingham,  Ala.

     015          Phoenix,  Ariz.
     031          Fresno,  Calif.
     024          Los Angeles—Long  Beach, Calif.
     028          Sacramento,  Calif.
     024          San Bernardino—Riverside, Calif.
     029          San Diego, Calif.
     030          San Francisco—Oakland, Calif.
     030          San Jose, Calif.
     036          Denver,  Colo.
     043          New York, N.Y.--Northeastern N.J.
     042          Waterbury, Conn.
     042          Springfield, Chicopee-Holyoke, Mass.--Conn.
     045          Wilmington,  Del.—N.J.
     045          Philadelphia,  Pa.—N.J.
     047          Washington,  D.C.—Md.--Va.
     067          Chicago,  111.--Northwestern Ind.
     131          Minneapolis--St. Paul, Minn.
     070          St. Louis, Mo.— 111.
     013          Las Vegas, Nev.
     148          Reno,  Nev.
     184          Oklahoma  City, Okla.
     151          Scranton, Pa.
     244          San Juan, P.R.
     200          Columbia, S.C.
     202          Greenville,  S.C.

     207          Knoxville, Tenn.
     018          Memphis,  Tenn.—Miss.
     215          Dallas, Tex.
     153          El  Paso,  Tex.
     216*         Houston,  Tex.
Source:   Data from EPA's  Environmental Monitoring Support Laboratory,
         Statistical  and  Technical Analysis Branch.

-------
SPM), does not appear In the lead SIP regulations,  however,  since that
terminology has not yet been incorporated into the  monitoring  requirements
for SIPs for other pollutants.   The terminology has been  retained for the
following discussion.
     The basic approach of the  strategy adopted by  SAMWG  is  that the
majority of State and local needs for monitoring data  will be  met with
the SLAMS network.  These needs include, for example,  developing,
tracking, and revising the control strategy of the  SIPs;  ensuring
compliance with the national ambient air quality standards (NAAQS);
measuring local air quality trends; determining potential  ambient
problems throughout urban areas; and understanding  specific  source
impacts.
     The primary national data  needs are met through the  use of NAMS,
which are a subset of the SLAMS stations.  The data from  the NAMS
stations are routinely reported to EPA Headquarters.  These  NAMS sites,
usually two per major urbanized area, are selected  to  provide  for
national trends assessments and overall SIP progress and  must  also  be
standardized in terms of their  operation, instrumentation, and placement
as  required by Section 319 of the Act.
     The two primary objectives of stations in the  NAMS network are (1)
to  tnonitor areas which are believed to experience the  highest  ambient
concentrations for averaging times consistent with  the NAAQS,  and  (2)  to
measure pollutant exposure to the public over the averaging  time of the
standard.  Thus, the lead NAMS  fall into two site categories:  (1)
stations located in the area of peak or maximum concentrations (middle
scale), and (2) stations which  combine poor air quality with high
population density (neighborhood  scale).  Both stations are  to avoid
                                  66

-------
 I
 •          single point source influences to the extent possible since NAMS are
            for national trends and conditions.  The SLAMS network could be used
 I          for single source problems since this network is designed primarily for
 •          local strategy development and trends.
                 Some SLAMS stations will also be located in areas experiencing
 •          highest ambient concentrations.  The distinction is that NAMS will not
            be sited within exposure settings that are unique to certain urban
 I          areas.  For example, NAMS will not be sited in confined street canyons
 •          since such exposure settings occur only in a relatively few urban areas.
                 From time to time, there are also data needs which would be
 •          impractical to meet with a routine network operation.  These special
            needs are met with special purpose monitoring.  Examples of these
 I          data needs include studying a specific source impact, evaluating why
 _          a particular site is not attaining a NAAQS, determining detailed
            spatial air quality gradients in a particular area, and validation of
•
            air quality models.

            5. 2 MONITORING SCALES
                 The monitor siting procedures which follow are required for SLAMS
_
•          and NAMS monitoring.  These procedures focus on the relationships between
•          a monitoring objective and geographical location.   The link between
            objective and location is made with the spatial scale of representative-
|          ness.  The spatial scale of representativeness is defined as the area
            around a site which has reasonably homogeneous air quality.  The goal in
•          siting monitors 1s to match the scale most appropriate for the monitor-
•          1ng objective of a station to the spatial scale of the ambient air
            monitored at a particular location.
I

I

-------
     As a result, the spatial  scale of representativeness  1s  described  in
terms of the physical dimensions of the air near the station.  The
scales of most interest for lead monitoring include  the  following:
     --Middle Sca1e--defines the concentration  typical of  areas  up  to
       several city blocks in size with dimensions ranging from  about
       100 meters to 0.5 kilometers.
     —Neighborhood Scale—defines concentrations within some extended
       area of the city that has relatively uniform  land use  with
       dimensions in the 0.5 to 4.0 kilometer range.
     These scales can be described by types of situations  which  frequently
occur within an urban area.  These situations are discussed below  in
terms of a physical description of a monitoring site measuring in  one
of the spatial scales of representativeness.
5.3  SITE DESCRIPTIONS
5.3.1  Roadway Site (Middle Scale)
     The roadway site must be located adjacent  to major  roadways.
These roadways may be arterials, freeways, expressways,  or interstate
highways with total average daily traffic (ADT) exceeding  50,000.   If
there is no roadway in the urbanized area with  ADT exceeding  50,000,
then the roadway with the largest traffic volume should  be selected.
The intent of this site is to obtain representative  worst  case measure-
ments of lead where people may be reasonably expected to be exposed.
In addition, it should be considered representative  of concentrations  in
the area measured as well as similar areas within the urban area.
                                 68

-------
 I
 M                   It  should  be  recognized that  in some situations, a roadway with a
                lower traffic volume may  be better suited for the monitoring site because
 I              of the fact that people may reside, work, or play closer to the roadway.
                In these situations, the  fact that population exposures occur closer to
 |              the roadway should override the traffic volume criterion for selecting
 ^              the monitoring  site.  The site should be placed at least five meters but
                no greater than 15 meters from the closest traffic lane.  It would be
 •              preferable to locate the  site in an area where the roadway passes through
                areas where people reside or work.  The monitor should be placed near
 |              residences and  no  greater than 5 meters above ground level.
                     Roadways that are at or below grade level should be selected in
 •              preference to elevated roadways where pos: ble,  since the differing
 •              heights  of elevated roadways make  representative long-term monitoring
                of high  concentrations very difficult.  This does not totally preclude
 I              monitoring near ^levated  roadways  if the agency believes that population
                exposures are significant.  Again, population exposure would override
 •              this siting consideration.  The monitors must not be placed near areas
 •              such as  toll gates or metered ramps since EPA has found  that the highest
                lead concentrations occur where traffic is moving at fairly high and
 I              constant rates  of  speed.   In monitoring lead from roadways that are
                below grade, the monitors  should be placed near the road but not actually
 I              in the cut section itself, since this would not be representative of
 •              population exposure.
                5.3.2 Neighborhood Site  (Neighborhood Seal ej_
 •                    This site must be  located in an area of high traffic and popula-
                tion density but not necessarily near a major roadway.  A minimum
I              _____
                FY-77 Catalyst Research  Project Report, Office of Research and
•              Development, Environmental Protection Agency (under preparation).
I

-------
separation distance of 15 m between the monitor and the  nearest roadway
(over 2,000-3,000 vehicles per day) is  recommended for purposes of
data comparability.  Areas at or near play areas or schools  are good
candidates for monitoring since children are the most susceptible  members
of the population at risk.  It would also be desirable for monitors  in
this site category to be placed so that they measure the combined  impact
of mobile and stationary sources.  The  monitors must be  placed  as  close
to the ground level as practicable, but no greater than  5 meters above
ground level, so that measurements will be taken as close to the breathing
zone as practicable.
5 3.3  Stre_et Canyon Site (Middle Scale)
       This type of SLAMS site should be located in an area  of  high
traffic and population density.  The site should be chosen to provide a
measure of the influence of the immediate source on the  pollution
exposure of the population.  These sites would typically be  located  in
downtown areas in roadway corridors or  canyons which have high  traffic
density (but not necessarily the largest traffic volume  in the  urbanized
area) and relatively uniform and tall buildings lining both  sides
of the street.  Monitors must be placed no higher than 5 meters
above ground level.
5.4  OTHER CONSIDERATION FOR ALL SITES
     If the sampler is located on a roof or other structure, then
there should be a minimum of 2 meters separation from walls, parapets,
penthouses, etc.  The sampler should be placed at least  20 meters from
                                   70

-------
 I
            trees since trees absorb particles as well  as restrict air flow.
 •          The sampler (except street canyon sites)  should also be located
 •          away from obstacles such as buildings, so that the distance between
            obstacles and the sampler are at least twice the height that the
 •          obstacle protrudes above the sampler.  There should be unrestricted
            air flow in at least three out of the four major wind directions except
 •          for street canyon sites.  One of the three should be the predominant
 •          direction for the season with the greatest pollution potential.
            5.5  NETWORK DESIGN
 I               Since the major objective of ambient lead monitoring by State
            and local agencies is to support the SIPs, the vast majority of  moni-
 •          toring sites should be located where the  lead concentrations are
 •          expected to be the highest.  Priority should be given to those locations
            which contain significant segments of the susceptible population at  risk.
 I          The siting criteria discussed above are required only for NAMS and SLAMS
            type monitoring.
 I               For NAMS, at least one of each of two types of sites are required
 _          to meet the objectives stated previously.  These are the roadway site
            and the neighborhood site.   It is believed that these two types  of
 I          sites will provide data which will enable representative air quality
            characterizations on a national scale.
 |               For additional SLAMS monitoring, SLAMS roadway sites do not need
 _          to meet any minimum traffic volume criterion (i.e., the 50,000 vehicle
 •          per day traffic criterion is not required).   This is to allow State  and
 •          local agencies to measure where significant air quality problems exist
            or to provide data for the development of spatial air quality gradients.
I          The 50,000 VPD requirement was intended only for MAMS monitors so that,

I
I

-------
as a minimum,  peak concentration measures  would be obtained.   The  street
canyon site is needed since many urban areas  (primarily  in the  eastern
part of the U.S.) have a significant portion  of the inner city  with
street canyons.  Therefore, for SLAMS purposes, this type of site may
be required to allow for situations where segments of the population
might be exposed to unhealthful levels of lead.
     For SPM, the type of site used could be  any of the  ones used for
SLAKS and NAMS.  The site could be other than those discussed below
because of the special nature or unique objective of the monitoring
project.
5.6  FREQUENCY OF SAMPLING
     The minimum acceptable sampling frequency for a quarterly  average
is once every six days.  This is identical  to the schedules currently
required for total suspended particulate and  ensures that each  day  of
the week should appear at least twice in the  90-day sampling period,
thereby reducing possible bias which may result from day-of-the-week
variations.  For a typically placed monitoring site with an average
concentration near the standard, the true concentration  could be
expected to lie within +_ 15 percent of the sample mean 95 percent
of the time for a one-in-six-day sample.  This is based  on an observed
coefficient of variation of 30 percent.  The  precision could be reduced
to +. 10 percent of the sample mean 95 percent of the time for a one-1n-
three-day sample.  This increase in precision, however,  does riot warrant
the significant national costs of monitoring.  Some areas, however, may
have sufficient reasons to intensify their sampling frequency based upon
local needs for data.
                                   72

-------
I

I

                                    6.0  NEW SOURCE  REVIEW
•                Procedures  for the review of new stationary  lead  sources will be
             similar to those for sources of other pollutants.   As  explained  1n the
•           preamble to the  proposed regulations, EPA  1s  developing  regulations  for
•           new stationary sources  of all  pollutants.   Procedures  for performing
             review of new stationary sources appear 1n EPA's  Guidelines  for  A1r  Quality
I           Maintenance Planning and Analysis, Volume  10;   Procedures for Evaluating
             A1r Quality Impact of New Stationary  Sources.


I

I

I

I

I
•            "Guidelines  for  A1r Quality Maintenance Planning and Analysis, Volume
              10:   Procedures  for Evaluating A1r Quality  Impact of New Stationary
I              Sources."  U.S.  Environmental Protection Agency, Office of A1r Quality
              Planning and Standards,  Research Triangle Park, NC 27711.  OAQPS No.
              1.2-029R,  October,  1977.
I

I

I

I

-------
I
I                               7.0  DETERMINATION OF LEAD POINT

_                                      SOURCE DEFINITION
I

I
I

I
                    Estimates  of maximum quarterly lead air quality concentrations

               from several  stationary source categories were made using the Single

               Source (CRSTER)  Model.  The emission rates used in this analysis were
I

|             taken  from  EPA's Background Support Document for Economic Impact

I

I
                                                                  p
               Assessment of the Lead Ambient Air Quality Standard.   A summary of
               the modeling results is provided in Table 7-1.

                   An estimate of a lead point source definition was extracted from

               these data by ising the following expression:

I

I                 q = 77^T
•                        p  p


I             where:  Q   = the approximation to the point source definition  (t/y);


I
                                                            o
                       S   = the assumed NAAQS for lead (fig/m );
                                                                     2
                       x   = the maximum quarterly concentration (fig/m ); and
•                    Q   = the lead emission rate (t/y).

              This equation yields the emission rate above which the standard is violated.

I                 The results of this calculation for all the sources appear in Table
                                                                                 3
              7-1.  The lowest point source definition for a standard of 1.5 jjg/m

              quarterly arithmetic mean, is 2 tons/year,  based on the estimated  impact

              from hypothetical grey iron foundries and ferroalloy plants.
•             User's Manual for Single Source (CRSTER)  Model.   U.S.  Environmental
•             Protection Agency, Office of Air Quality  Planning and  Standards,
               Research Triangle Park, NC 27711.   EPA-450/2-77-013.   July 1977.

I             Background Support Document for Economic  Impact  Assessment of the
™             Lead Ambient Air Quality Standard.   U.S.  Environmental  Protection
               Agency, Office of Air and Waste Management,  Office of  Air Quality
               Planning and Standards.  January 12, 1978.
                                                 75

-------
           TABLE  7-1:   STATIONARY SOURCE QUARTERLY MODE.IN6
        RESULTS AND POINT SOURCE DEFINITION EOR NAAQS QUARTERLY
                         AVERAGE OF 1.5 ug/irT
                                   Maximum
Industry
Type
Primary Lead
Primary Copper
Tetraethyl Lead
Mfg.
Secondary Lead
Grey Iron
Ferroalloy
Lead
Emission
Rate (t/yr)
110
94
243
63
2.1
0.28
Quarterly
Concentration
(ug/nn
2.5
2.6
12.5
15
1.5
0.24
XP/QP,
(ug/nr)
t/yr
0.023
0.028
0.051
0.24
0.71
0.86
Point 5
Definil
(t/yi
65
54
29
6
2
2
Battery Manufacture
 (500 BPD) w/o
 PbO production     1.6

Battery Manufacture
 (6500 BPD) w/PbO
 production           23
0.6
4.8
0.38
0.21
*Emission rates above which the NAAQS will be violated.
                                   76

-------
I

•              The predominant caus* of the elevated concentrations from these two
•         facilities was assumed to be fugitive emissions.  The emission factors
           used to estimate the fugitive emissions from these facilities were derived
•         indirectly from ambient particulate data and assumptions concerning lead
           content of the particulate natter.  Thus, they are of questionable validity.
I         The point source definition will only be used in two functions:  emission
•         inventory development and possibly new source review.  A slightly larger
           point source definition would rpt unduly affect the accuracy of the control
I         strategy analysis or the number of sources that would have to be reviewed
           under the new source review  requirements (which will not be included in
|         the lead implementation plan regulations but will appear in a subsequent
«         rulemaking).   Because of this  and ths uncertainties in the emission factors,
           and because State resources  for  gathering lead emission data are severely
•         limited, it appears  that a  less  stringent point source definition of 5
           tons/year is warranted.
I

I

I

I

I

I

I

I

I

-------
                APPENDIX A

PROCEDURES FOR DETERMINING THE INORGANIC LEAD
I
I
I
I
I
I
I
I                                EMISSIONS  FROM  STATIONARY SOURCES
I
I
I
I
I
I
I
I
I
I
I

-------
I

                        PROCEDURES FOR DETERMINING THE INORGANIC LEAD
I                           EMISSIONS FROM STATIONARY SOURCES

I

I

I

I

I

I

I

I

I
•                        ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
                                OFFICE OF RESEARCH AND DEVELOPMENT
I                             U.S. ENVIRONMENTAL PROTECTION AGENCY
.                         RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
                                          MARCH 1978
I

I

I

I

-------
                PROCEDURE FOR DETERMINING THE INORGANIC LEAD
                     EMISSIONS FROM STATIONARY SOURCES

1.   Principle and Applicability
     1.1       Principle.  Participate and gaseous lead emissions are
withdrawn isokinetically from the source.  The collected samples are
digested in acid solution and analyzed by atomic absorption spectrometry
using an air acetylene flame.
     1.2       Applicability.  This method is applicable for the determination
of inorganic lead emissions from stationary sources.
 2.   Range, Sensitivity, Precision, and Interferences
     2.1       Range.  The upper limit can be considerably extended by dilution.
For a minimum analysis accuracy of +_ 10%, a minimum lead mass of 100 yg should
be collected.
     '
-------
I

I
.                2.4        Interferences.   Sample matrix effects may  interfere with
*           the  analysis  for  lead  by  flame  atomic absorption.   If the analyst suspects
•           that the  sample matrix is  causing erroneous results, the  presence of  these
             matrix  effects  can  be  confirmed and  frequently corrected  for  by carrying out
I           the  analysis  using  the Method of Standard Additions.
_                          High concentrations of copper may  interfere with the analysis
•           of lead at  217.0  nm.   This  interference can be avoided  by analyzing the samples
•           for  lead  using  the  283.3  nm lead line.
             3.    Apparatus
I                3.1        Sampling Train.   A schematic of the  sampling train used in this
             method  is shown in  Figure  A-l.   Complete construction details are given in APTD-
•           0581; commercial  models of this train are also available.  For changes from
•           APTD-0581 and for allowable modifications of the  train  shown  in Figure A-l,
             see  the following subsections.
I                          The operating and maintenance procedures  for  the sampling train
             are  described in  APTD-0576.  Since correct usage  is important in obtaining
 •           valid results,  all  users  should read APTD-0576 and  adopt  the  operating and
 •           maintenance procedures outlined in it, unless otherwise specified herein.
             The  sampling  train  consists of  the following components:
             tapered  leading edge.  The angle of taper shall be < 30°, and the taper shall
•               3.1.1       Probe Nozzle.  Stainless steel  (316) or glass with sharp,

I
I
             be  on  the  outside  to  preserve a constant  internal diameter.  The probe  nozzle
             shall  be of  the  button-hook or elbow design, unless otherwise specified by the
             Administrator.   If made of stainless  steel, the nozzle shall be constructed
 I          from seamless  tubing; other materials  of  construction may be used,  subject to
             the approval of  the Administrator.

                                                 83

-------
                                               c

                                               'co
                                               O)

                                               C


                                               "5.

                                               E
                                               co
                                               t/>

                                               T3
                                               03
                                                u

                                               'c
                                                co

                                                ff
                                                o
                                                01
84

-------
I

I
•               A range  of nozzle  sizes  suitable  for  isokinetic sampling  should be
            available,  e.g.,  0.32 to  1.27 cm  (1/8  to 1/2  in.),or larger  if higher
I          volume sampling trains  are  used,  inside diameter  (I.D.)  nozzles  in  increments
M          of 0.16 cm  (1/16  in.).  Each  nozzle  shall  be  identified  and  calibrated  (see
            Section 5.2).
•               3.1.2        Probe  Liner.   Borosilicate or quartz glass  tubing with a
            heating system capable  of maintaining  a gas temperature  at the exit end during
|          sampling of 120°  +_ 14°C (248° + 25°F); note that  lower exit  temperatures are
M          acceptable, provided that they exceed  the  stack gas dew  point.   Since the
            actual  temperature at the outlet  of  the probe is  not usually monitored during
I          sampling, probes  constructed  according to  APTD-0581 and  utilizing the calibration
            curves of APTD-0576 (or calibrated according  to the procedure  outlined in APTD-
|          0576)  will  be  considered  acceptable.
•               Either borosilicate  or quartz glass probe liners may be used for stack
            temperatures  up to about  480°C (900°F); quartz liners shall  be used for tempera-
•          tures  between  480° and  900°C  (900° and 1650°F).   Both types  of liners may be
            used  at temperatures higher than  specified for short periods of  time, subject
|          to the approval  of the  Administrator.  The softening temperature for borosilicate
.          is 820°C (1508°F), and  for  quartz it is 1500°C (2732°F).
™               Whenever  practical,  every effort  should  be made to  use  borosilicate or
fl          quartz glass  probe liners.  Alternatively, metal  liners  (e.g., 316  stainless
            steel, Incoloy  825*, or  other corrosion resistant metals) made  of  stainless
|          tubing be used, subject to  the approval of the Administrator.

*          *Mention of trade names or  specific  products  does not constitute endorsement by
             the  U.S. Environmental Protection Agency.
I
                                                 85

-------
     3.1.3       Pitot Tube.   Type S, as described in Section 2.1  of Method 2,
40 CFR 60 Appendix A, or other device approved by the Administrator.  The Pitot
tube shall be attached to the probe (as shown in Figure A-l) to allow constant
monitoring of the stack gas velocity.  The impact (high pressure)  opening plane
of the Pitot tube shall be even with or above the nozzle entry plane (see
Method 2, Figure 2-6b) during sampling.  The Type S Pitot tube assembly shall
have a known coefficient, determined as outlined in Section  4 of Method 2.
     3.1.4       Differential Pressure Gauge.  Inclined manometer or equivalent
device (two), as described in Section 2.2 of Method 2.   One  manometer shall be
used for velocity head (Ap) readings, and the other, for orifice differential
pressure readings.
     3.1.5       Filter Heating System.  Any heating system  capable of main-
taining a temperature around the filter holder during sampling of 120° j^ 14°C
(248° +_ 25°F), or such other temperature as specified by an  applicable subpart of
the standards or approved by the Administrator for a particular application.
Alternatively, the tester may opt to operate the equipment at a temperature
lower than that specified.  A temperature gauge capable of measuring temperature
to within 3°C (5.4°F) shall be installed so that the temperature around the filter
holder can be regulated and monitored during sampling.   Heating systems other than
the one shown in APTD-0581 may be used.
     3.1.6       Filter Holder.  Borosilicate glass, with a  glass frit filter
support and a silicone rubber gasket.  Other materials of construction (e.g.,
stainless steel, Teflon, Viton) may be used, subject to the  approval of the
Administrator.  The holder design shall provide a positive seal against leakage
from the outside or around the filter.  The filter holder shall be attached
immediately at the outlet of the probe.

                                      86

-------
I
I
_              3.1.7        Impingers.   Four  impingers connected  in  series with  leak-free
           ground  glass  fittings or any  similar  leak-free noncontaminating fittings.  The
I         first,  third, and  fourth impingers  shall be of Greenburg-Smith design, modified
           by  replacing  the tip with a 1.3 cm  (1/2  in.)  I.D. glass tube extending to about
I         1.3 cm  (1/2 in.) from the bottom of the  flask.  The second  impinger shall be of
_         the Greenburg-Smith design with the standard  tip.  The first and second  impingers
™         shall contain known quantities of 0.1 N  HNO-  (Section  4.1.3), the third  shall
•         be  empty, and the  fourth shall contain a known weight  of  silica gel or equivalent
           desiccant.  A thermometer, capable  of measuring temperature to within 1°C (2°F),
P         shall be placed at the outlet of the  fourth impinger for  monitoring purposes.
_              3.1.8       Metering System.   Vacuum gauge, leak-free  pump, thermometers
•         capable of measuring temperature to within 3°C (5.4°F), dry gas meter capable of
•         measuring volume to within 2%, and  related equipment,  as  shown in Figure A-l .
           Other metering systems capable of maintaining sampling rates within 10%
|         of  isokinetic and  of determining sample  volumes to within 2% may be used,
_         subject to the approval of the Administrator.  When the metering system  is
•         used in conjunction with a Pitot tube, the system shall enable checks of
•         isokinetic rates.
                Sampling trains utilizing metering  systems designed  for flow rates  higher
£         than that described in APTD-0581 or APTD-0576 may be used provided that  the
           specifications of  this method are met.
•              3.1.9       Barometer.   Mercury, aneroid, or other barometer capable of
•         measuring atmospheric pressure to within 2.5  mm Hg (0.1 in. Hg).  In many cases,
           the barometric reading may be obtained from a nearby National Weather Service
I         station, in which  case the station  value (which is the absolute barometric
           pressure) shall be requested, and an  adjustment for elevation differences between


i

-------
the weather station and sampling point shall  be applied  at a  rate of minus
2.5 mm Hg (0.1  in.  Hg)  per 30 m (100 ft)  elevation  increase or vice  versa  for
elevation decrease.
     3.1.10       Gas Density Determination Equipment.   Temperature  sensor  and
pressure gauge, as  described in Sections  2,3 and 2.4 of  Method 2, and gas
analyzer, if necessary, as described in Method 3 (40 CFR 60 Appendix A).   The
temperature sensor  shall,  preferably, be  permanently attached to the Pitot  tube
or sampling probe in a fixed configuration, such that the tip of the sensor
extends beyond the  leading edge of the probe sheath and  does  not touch any  metal.
Alternatively,  the  sensor  may be attached just prior to  use in the field.   Note,
however, that if the temperature sensor is attached in the field, the sensor
must be placed in an interference-free arrangement  with  respect to the Type S
Pitot tube openings (see Method 2, Figure 2-7).  As a second  alternative,  provided
that a difference of not more than 1% in  the average velocity measurement
is introduced,  the  temperature gauge need not be attached to  the probe or  Pitot
tube.  (This alternative is subject to the approval of the Administrator.)
     3.2         Sample Recovery.  The following items are needed:
     3.2.1       Probe-Liner and Probe-Nozzle Brushes.  Nylon bristle brushes
with stainless steel wire  handles.  The probe brush shall have extensions  (at
least as long as the probe) of stainless  steel, Nylon, Teflon, or similarly
inert material.  The brushes shall be properly sized and shaped to brush out the
probe liner and nozzle.
     3.2.2       Glass Wash Bottles—Two.
     3.2.3       Glass Sample Storage Containers.  Chemically resistant, boro-
silicate glass bottles, for 0.1 N HNO.,  impinger and probe solutions and washes,
                                     O
1000 ml.  Screw cap liners shall be either rubber-backed Teflon or constructed

                                    88

-------
I

I
•           so as to be leak-free and resistant to chemical  attack by 0.1  N HNO.,.
             (Narrow mouth glass bottles have b.een found to be less prone to leakage.)
•                3.2.4       Petri Dishes.   For filter samples,  glass or polyethylene,
•           unless otherwise specified by the Administrator.
                  3.?.5       Graduated Cylinder and/or Balance.   To measure condensed water
I           to within  2 ml  or 1 g. The graduated cylinder shall  have a minimum capacity of
             500 ml, and subdivisions no greater than 5 ml.   Most laboratory balances are
I           capable of weighing to the nearest 0.5 g or less.
m                3.2.6       Plastic Storage Containers.   Air-tight containers to  store
             silica gel.
•                3.2.7       Funnel  and Rubber Policeman.  To aid in transfer of silica gel
             to container; not necessary if silica gel  is  weighed in the field.
|                3.2.8       Funnel.  Glass, to aid in sample recovery.
•                3.3         Analysis.
                  3.3.1       Atomic Absorption Spectrophotometer.  With lead hollow cathode
•           lamp and burner for air/acetylene flame.
                  3.3.2       Hot Plate.
I                3.3.3       Erlenmeyer Flasks.  125 ml 24/40 f.
•                3.3.4       Membrane Filters.  Millipore SCWPO  4700 or equivalent.
                  3.3.5       Filtering Apparatus.  Millipore vacuum filtration unit, or
I           equivalent, for use with the above membrane filter.
                  3.3.6       Volumetric Flasks.  100 ml,  250 ml.
I           4.   Reagents
—                4.1         Sampling.
™                4.1.1       Filters.  High purity glass  fiber filters, without organic
•           binder, exhibiting at least 99.95% efficiency (<_ 0.05% penetration) on 0.3

i

-------
micron dioctyl phthalate smoke particles.   The filter efficiency test
shall be conducted in accordance with ASTM Standard Method D 2986-71.  Test
data from the supplier's quality control  program are sufficient for this purpose.
Filters shall be Gelman Spectro Grade, or equivalent, with lot assay for Pb.
Reeve Angel 934 AH and MSA 1106 BH filters have been found to be equivalent.
     4.1.2       Silica Gel.   Indicating type, 6 to 16 mesh.  If previously
used, dry at 175°C (350°F) for 2-hr.   New silica gel may be used as received.
Alternatively, other types of desiccants (equivalent or better) may be used,
subject to the approval of the Administrator.
     4.1.3       Nitric Acid, 0.1 N.   Prepared from reagent grade HN03
and deionized, distilled water (Reagent 4.4.1, below).  It may be desirable to
run blanks prior to field use to eliminate a high blank on test samples.  Prepare
by diluting 6.5 ml of concentrated HNO- (69%) to 1 liter with deionized,
distilled water.
     4.1.4       Crushed Ice.
     4.1.5       Stopcock Grease.  HNO,, insoluble, heat stable, silicone grease.
This is not necessary if screw-on connectors with Teflon sleeves, or similar, are
used.  Alternatively, other types of stopcock grease may be used, subject to the
approval of the Administrator.
     4.2         Pretest Preparation.
     4.2.1       Nitric Acid, 6 N.  Prepared from reagent grade HNO., and deionized,
distilled water.  Prepare by diluting 390 ml of concentrated HNO_ (69%) to  1
liter with deionized, distilled water.
     4.3         Sample Recovery.
     4.3.1       Nitric Acid, 0.1 N.  Same as 4.1.3 above.
     4.4         Analysis.
                                    90

-------
 I
 I
               4.4.1       Water.   Deionized,  distilled to conform to ASTM Specification
          D 1193-74, Type 3.
 I             4.4.2       Nitric Acid.  Concentrated ACS reagent grade,  or equivalent.
               4.4.3       Nitric Acid,  50% (V/V).   Dilute 500 ml  of concentrated
 I        HN03 to 1  liter with deionized,  distilled water.
 •             4.4.4       Stock Lead Standard Solution (1000 yg Pb ml"1).   Dissolve
          0.1598  g  of reagent gradp Pb  (N03)2 in about 60 ml  of deionized distilled
 I        water, add 2 ml concentrated HNOo,  and dilute to 100 ml  with dionized,  dis-
          tilled water.
 I             4.4.5       Lead Standards.
 •             4.4.5.1     Solution Sample Standards.   Pipet 0.0,  1.0, 2.0, 3.0,  4.0,
          and 5.0 ml aliquots of the stock lead standard solution  (Reagent 4.4.4) into
 •        250 ml volumetric flasks.   Add 5 ml  concentrated HNO, to each  flask  and dilute
          to volume with deionized,  distilled water.   These working  standards  contain
 |        0.0, 4.0, 8.0, 12.0,  16.0, and 20.0 yg Pb ml"1,  respectively.   Additional
 •        standards at other concentrations  should  be  prepared  in  a  similar  manner as
          needed.
 I             4.4.6       Air.   Of  a quality suitable for atomic  absorption analysis.
               4.4.7       Acetylene.   Of a  quality suitable for atomic  absorption analysis,
 I             4.4.8       Hydrogen  peroxide.  ACS  reagent grade or  equivalent,  35» by
 *                         volume.
 •        5.    Procedure
               5.1          Sampling.  The complexity of this method  is such  that,  in
 I        order to  obtain reliable results,  testers should be trained and experienced
          with the  test procedures.
|             5.1.1       Pretest Preparation.   All the components  shall  be maintained
_        and calibrated according to the procedure described in APTD-0576,  unless

                                                 91
I

-------
otherwise specified herein.   In addition, prior to testing, all  sample-exposed
surfaces shall be rinsed, first with 6 N HNCL then with deionized, distilled
water.
     Weigh several  200 to 300  g portions of silica gel in air-tight containers
to the nearest 0.5  g.  Record the total weight of the silica gel  plus con-
tainer, on each container.  As an alternative, the silica gel need not be
preweighed, but may be weighed directly in its impinger just prior to train
assembly.
     Check filters visually against light for irregularities and flaws or pin-
hole leaks.  Label  the shipping containers (glass or plastic petri dishes) and
keep the filters in these containers at all  times except during  sampling and
analysis.
     5.1.2          Preliminary Determinations.  Select the sampling site anfd
the minimum number of sampling points according to Reference Method 1 or as
specified by the Administrator.  Determine the stack pressure, temperature,
and the range of velocity heads using Reference Method 2; it is  recommended
that a leak-check of the pitot lines (see Method 2, Section 3.1) be performed.
Determine the moisture content using Reference Method 4 or its alternatives
for the purpose of making isokinetic sampling rate settings.  Determine the
stack gas dry molecular weight as described in Reference Method 3.
     Select a nozzle size based on the range of velocity heads,  such that it
is not necessary to change the nozzle size in order to maintain isokinetic
sampling rates.  During the run, do not change the nozzle size.   Insure that
the proper differential pressure gauge is chosen for the range of velocity
heads encountered (see Section 2.2 of Method 2).
                                       92

-------
  I
  I
 _              Select  a  suitable  probe  liner  and probe length such that all traverse
            points  can be  sampled.   For large stacks, consider sampling from opposite
 I         sides of the stack  to reduce  the length of probes.
                 Select  a  total  sampling  time such that  (1)  the sampling time per point is
 |         not  less than  2 min.  (or  greater time interval as specified by the Adminis-
 _         trator), and (2)  a  minimum lead mass of 100  yg is collected in the sample.
 ™         The  sampling time and volume  will therefore  vary from source-to-source.
 I              It is recommended  that the number of minutes sampled at each point be an
            integer or an  integer plus one-half minute,  in order to avoid timekeeping errors,
 J              In some circumstances, e.g., batch cycles, it may be necessary to sample
 _         for  shorter  times at the traverse points and to obtain smaller gas sample
 •         volumes.   In these  cases,  the Administrator's approval must first be obtained.
 •              5.1.3     Preparation of Collection Train.  During preparation and
            assembly of  the sampling train, keep all openings where contamination can occur
 •         covered until  just  prior to assembly or until sampling is about to begin.
                 Place 100 ml of 0.1 HNO., in each of the first two impingers, leave the
 I
 •         third impinger empty, and  transfer approximately 200 to 300 ug of preweighed
 •         silica  gel from its  container to the fourth  impinger.  More silica gel may
            be used, but care should be taken to insure  that it is not entrained and
 •         carried out  from  the impinger during sampling.  Place the container in a
            clean place  for later use  in  the sample recovery.  Alternatively, the weight
 •          of the  silica  gel plus  impinger may be determined to the nearest 0.5 g
 •          and  recorded.
                 Using tweezers or  clean  disposable surgical gloves, place a filter in the
 •          filter  holder.  Be  sure that  the filter is properly centered and the gasket
            properly placed,  so as  to  prevent the sample gas stream from circumventing the
I
                                                 93

-------
filter.  Check the filter for tears after assembly is completed.
     When glass liners are used, install  the selected nozzle using a Viton
A 0-ring when stack temperatures are less than 260°C (500°F) and  an asbestos
string gasket when temperatures are higher.   See APTD-0576 for details.   Other
connecting systems using either 316 stainless steel  or Teflon ferrules  may be
used.  When metal  liners are used, install  the nozzle as above or by a  leak-
free direct mechanical connection.  Mark  the probe with heat resistant  tape or
by some other method to denote the proper distance into the stack or duct for
each sampling point.
     Set up the train as in Figure A-l, using (if necessary) a very light coat
of silicons grease on all ground glass joints, greasing only the  outer  portion
(see APTD-0576) to avoid possibility of contamination by the silicone grease.
     Place crushed ice around the impingers.
     5.1.4       Leak-Check Procedures.
     5.1.4.1     Pretest Leak-Check.  A pretest leak-check is recommended, but
not required.  If the tester opts to conduct the pretest leak-check, the following
procedure shall be used.
     After the sampling train has been assembled, turn on and set the filter
and probe heating systems at the desired  operating temperature.  Allow  time for
the temperature to stabilize.  If a Viton A 0-r1ng or other leak-free connection
1s used 1n assembling the probe nozzle to the probe Uner, leak-check the train
at the sampling site by plugging the nozzle and pulling a 380 mm Hg (15 1n. Hg)
vacuum.
     Note:  A lower vacuum may be used, provided that 1t is not exceeded during
the test.
     If an asbestos string 1s used, do not connect the probe to the train during
                                        94

-------
 I
 I
            the leak-check.   Instead, leak-check the train by first plugging the inlet
            to the filter and pulling a 380 mm Hg (15 in.  Hg)  vacuum (see note immediately
 I          above).  Then connect the probe to the train and leak-check at about 25 mm Hg
            (1 in. Hg) vacuum; alternatively, the probe may be leak-checked with the rest
 |          of the sampling train, in one step, at 380 mm Hg (15 in. Hg)  vacuum.  Leakage
                                                                     3             31
 •          in excess of 4% of the average sampling rate or 0.00057 m /min (0.02 ft  min   ),
            whichever is less, are unacceptable.
 I               The following leak-check instructions for the sampling train  described in
            APTD-0576 and APTD-0581  may be helpful.  Start the pump with  bypass valve fully
 |          open and coarse adjust valve completely closed.   Partially open the coarse adjust
 «          valve and slowly close the bypass valve until  the  desired vacuum is reached.
            Do, not reverse direction of bypass valve; this will  cause 0.1  N HNO^ to back
 I          up into the filter.   If the desired vacuum is  exceeded, either leak-check at
            this higher vacuum or end the leak check as shown  below and start  over.
 |               When the leak-check is completed, first slowly remove the plug from the
 _          inlet to the probe and immediately turn off the vacuum pump.   This prevents the
 ™          0.1N HNO^ in the impingers from being forced backward and silica gel from being
 •          entrained backward.
                 5.1.4.2   Leak-Checks During Sample Run.   If, during the sampling run, a
 J          component (e.g., filter assembly or impinger)  change becomes  necessary, a leak-
 _          check shall  be conducted immediately  before the change is made.  The leak-check
 '          shall be done according to the procedure outlined  in Section  5.1.4.1 above,
 •          except that it shall  be done at a vacuum equal to  or greater  than  the maximum
            value recorded up to  that point in the test.   If the leakage  rate  is found to
                                        13             3-1
            be no greater than 0.00057 m /min (0.02 ft  min" ) or 4% of the average sampling
            rate (whichever is less), the results are acceptable, and no  correction will  need


i

-------
to be applied to the total  volume of dry gas metered;  if,  however,  a  higher
leakage rate is obtained, the tester shall  either record the leakage  rate and
plan to correct the sample volume as shown  in Section  6.3  of Reference Method
5, or shall void the sampling run.
     Immediately after component changes, leak-checks  are  optional; if such
leak-checks are done, the procedure outlined in Section 5.1.4.1  above shall
be used.
     5.1.4.3   Posttest Leak-Check.  A leak-check is mandatory at the conclusion
of each sampling run.  The leak-check shall  be done in accordance with the
procedures outlined in Section 5.1.4.1, except that it shall  be conducted at a
vacuum equal to or greater than the maximum value reached  during the  sampling
run.  If the leakage rate is found to be no greater than 0.00057 m /min (0.02
ft ) or 4% of the average sampling rate (whichever is  less), the results are
acceptable, and no correction need be applied to the total  volume of  dry gas
metered.   However, if a higher leakage rate is obtained, the tester shall
either record the leakage rate and correct  the sample  volume as shown in
Section 6.3 of Method 5, or shall void the  sampling run.
     5.1.5     Sampling Train Operation.  During the sampling run,  maintain an
isokinetic sampling rate (within 10% of true isokinetic unless otherwise
specified by the Administrator).
     For each run, record the data required on a data  sheet such as the one
shown in EPA Method 5, Figure 5-2.   Be sure to record  the  initial dry gas meter
reading.   Record the dry gas meter readings at the beginning and end  of each
sampling time increment, when changes in flow rates are made, before  and after
each leak-check, and when sampling is halted.  Take other readings required
by Figure 5-2 of Method 5 at least once at  each sample point during each time
                                      96

-------
  I
  I
  I         increment and additional readings when significant changes (20% variation
           in velocity head readings) necessitate additional adjustments in flow rate.
  I
 I
           Level and zero the manometer.  Because the manometer level and zero may drift
           due to vibrations and temperature changes, make periodic checks during the
           traverse.
 •             Clean the portholes prior to the test run to minimize the chance of
           sampling deposited material.  To begin sampling, remove the nozzle cap, verify
 |        that the filter and probe are at proper temperature, and that the Pi tot tube
 «        and probe are properly positioned.  Position the nozzle at the first traverse
           point with the tip pointing directly into the gas stream.  Immediately start
 I        the pump and adjust the flow to isokinetic conditions.  Nomographs are available,
           which aid in the rapid adjustment of the isokinetic sampling rate without excessive
 |        computations.  These nomographs are designed for use when the Type S Pi tot tube
 _        coefficient is 0.85 j^O.02, and the stack gas equivalent density (dry molecular
           weight) is equal to 29 +_ 4.  APTD-0576 details the procedure for using the nomo-
 •        graphs.  If C  and Md are outside the above stated ranges, do not use the nomo-
           graphs unless appropriate steps (Shigehara, 1974) are taken to compensate for
 |        the deviations.
 _              When the stack is under significant negative pressure (>_ a water column
 •         the height of the impinger stem), take care to close the coarse adjust valve
 •         before inserting the probe into the stack to prevent 0.1 N HN03 from backing
           into the filter.  If necessary, the pump may be turned on with the coarse
 I         adjust valve closed.
                When the probe is in position, block off the openings around the probe and
•         porthole to prevent dilution of the gas stream.
•              Traverse the stack cross-section, as required by Reference Method 1 or as
I

-------
                                                                                   I

                                                                                   I
specified by the Administrator, without bumping the probe nozzle into the          «
stack walls when sampling near the walls or when removing or inserting the
probe through the portholes.                                                        •
     During the test run, add ice and, if necessary, salt to the ice bath, to
maintain a temperature of less than 20°C (68°F) at the impinger/silica gel         |
outlet.  Also, periodically check the level and zero of the manometer.             _
     A single train shall be used for the entire sample run, except in cases       ™
where simultaneous sampling is required in two or more separate ducts or at        •
two or more different locations within the same duct, or, in cases where
equipment failure necessitates a change of trains.  In all  other situations,
the use of two or more trains will be subject to the approval of the Administrator. _
     Note that when two or more trains are used, separate analyses of the sample   •
fractions from each train shall be performed, unless otherwise specified by the    •
Administrator.  Consult with the Administrator for details concerning the calcu-
lation of results when two or more trains are used.                                I
     At the end of the sample run, turn off the coarse adjust valve, remove the
probe and nozzle from the stack, turn off the pump, record the final dry gas       •
meter reading, and conduct a post-test leak-check, as outlined in Section 5.1.4.3.  •
Also, leak-check the Pitot lines as described in Method 2, Section 3.1; the lines
must pass this leak-check in order to validate the velocity head data.              I
     5.1.6       Calculation of Percent Isokinetic.  Calculate percent isokinetic
(see Section 6.11 of Method 5) to determine whether the run was valid or another    I
test run should be made.  If there was difficulty in maintaining isokinetic rates   •
due to source conditions, consult with the Administrator for possible variance
on the isokinetic rates.                                                            •
     5.2         Sample Recovery.  Proper cleanup procedure begins as soon as the
                                                                                    i

                                                                                    i

-------
 I

 I
 •        probe is removed from the stack at the end of the sampling period.   Allow the
           probe to cool .
 |             When the  probe can be safely handled, wipe off all  external  participate
 •        matter near the tip of the probe nozzle and place a cap  over it.   Do not cap
           off the probe  tip tightly while the sampling train is  cooling down  as this
 8        would create a vacuum in the filter holder, thus drawing liquid from the
           impingers into the filter.
 |             Before moving the sample train to the cleanup site, remove the probe from
 _        the sample train, wipe off the silicone grease, and cap  the open  outlet of
 ™        the probe.  Be careful not to lose any condensate that might be present.  Wipe
 •        off the silicone grease from the glassware inlet where the probe  was fastened
           and cap the inlet.  Remove the umbilical  cord from the last impinger and cap
 |        the impinger.   Either ground-glass stoppers, plastic caps, or serum caps may
 _        be used to close these openings.
 •             Transfer  the probe and filter-impinger assembly to  the cleanup area.  This
 •        area should be clean and protected from the wind so that the chances of contam-
           inating or losing the sample will  be minimized.
 I             Save a portion of the 0.1N HNCL used for sampling and cleanup  as a blank.
           Place 200 ml of this 0.1N HNO., taken directly from the bottle being used into
 I
 •        a glass sample container labeled "0.1N HN03 blank."
 •             Inspect the train prior to and during disassembly and note any abnormal
           conditions.  Treat the samples as follows:
 I              Container No. 1 .   Carefully remove the filter from  the filter  holder and
           place it in its identified petri  dish container.  If it  is necessary to fold
 •         the filter, do so such that the sample-exposed side is inside the fold.  Care-
 •         fully transfer to the petri dish any visible sample matter and/or filter fibers
i

-------
that adhere to the filter holder gasket by using a dry Nylon bristle brush
and/or a sharp-edged blade.  Seal the container.
     Container No. 2.    Taking care to see that dust on the outside of the
probe or other exterior surfaces does not get into the sample, quantitatively
recover sample matter or any condensate from the probe nozzle, probe fitting,
probe liner, and front half of the filter holder by washing these components
with 0.1 N HN03 and placing the wash into a glass or polyethylene container.
Measure and record (to the nearest ml) the total amount of 0.1N HNCL used for
each rinse.  Perform the 0.1N HNCL rinses as follows:
     Carefully remove the probe nozzle and clean the inside surface by rinsing
with 0.1N HNCL from a wash bottle while brushing with a stainless steel, Nylon-
bristle brush.  Brush until the 0.1N HNCL rinse shows no visible particles, then
make a final rinse of the inside surface with 0.1N HNCL.
     Brush and rinse with 0.1N HNCL the inside parts of the Swagelok fitting
in a similar way until no visible particles remain.
     Rinse the probe liner with 0.1N HNCL by tilting the probe and squirting
0.1N HNCL into its upper end, while rotating the probe so that all inside
surfaces will be rinsed with 0.1N HN03.  Let the 0.1N HN03 drain from the
lower end into the sample container.  A glass funnel may be used to aid in
transferring liquid washes to the container.  Follow the 0.1N HNCL rinse with
a probe brush.  Hold the probe in an inclined position, squirt 0.1N HN03 into
the upper end of the probe as the probe brush is being pushed with a twisting
action through the probe; hold a sample container underneath the lower end of
the probe and catch any 0.1N HN03 and sample matter that is brushed from the
probe.  Run the brush through the probe three times or more until no visible
sample matter is carried out with the 0.1N HN03 and none remains on the probe

                                    100

-------
 I

 I
 •        liner  on  visual  inspection.  Hith stainless steel or other metal probes, run
           the  brush  through in  the  above  prescribed manner at least six times, since metal
 •        probes  have  small  crevices  in which  sample matter can be entrapped.  Rinse the
           brush with 0.1N HNCL  and  quantitatively collect these washings in the sample
                              I3
           container.   After the brushing  make  a final 0.1N HNO^ rinse of the probe as
 •        described  above.
                It is recommended that two people be used to clean the probe to minimize
 I        loss  of sample.   Between  sampling runs, keep brushes clean and protected from
           contamination.
 I             After insuring that  all joints  are wiped clean of silicone grease, clean
 m        the  inside of the front half of the  filter holder by rubbing the surfaces with
           a  Nylon bristle brush and rinsing with 0.1N HNCL.  Rinse each surface three
 I        times or more, if needed, to remove  visible sample matter.  Make a final rinse
           of the  brush and  filter holder.  After all 0.1 N HNO, washings and sample matter
                                                              I3
           are  collected in  the  sample container, tighten the lid on the sample container
 •        so that 0.1N HM03 will not  leak out  when it is shipped to the laboratory.  Mark
           the  height of the fluid level to determine whether leakage occurred during
 I        transport.   Label  the container to clearly identify its contents.
 _             Container No.  3.  Check the color of the indicating silica gel to determine
 •        if it has  been completely spent and  make a notation of its condition.  Transfer
 •        the  silica gel from the fourth  impinger to the original container and seal.  A
           funnel  may make it easier to pour the silica gel without spilling.  A rubber
 I         policeman  may be  used as  an aid in removing the silica gel from the impinger.
           It is not  necessary to remove the small amount of dust particles that may adhere
 •         to the  walls and  are  difficult  to remove.  Since the gain in weight is to be
 •         used  for moisture calculations, do not use any water or other liquids to transfer

i

-------
the silica gel.   If a balance is available in  the field,  follow the  procedure
for Container No.  3 under "Analysis."
     Container No.  4.  Due to the large quantity of liquid  involved, the  impinger
solutions are placed together in a separate container.   However,  they may be com-
bined with the contents of Container No.  2 at  the time  of analysis  in order to
reduce the number of analyses required.  Clean each of  the  first  three impingers
and connecting glassware in the following manner:
     1.  Wipe the impinger ball joints free of silicone grease and  cap the joints,
     2.  Rotate and agitate each impinger, so  that the  impinger contents  might
serve as a rinse solution.
     3.  Transfer the contents of the impingers to a 500 ml  graduated cylinder.
The outlet ball  joint cap should be removed and the contents drained through this
opening.  The impinger parts (inner and outer  tubes) must not be  separated while
transferring their contents to the cylinder.
         Measure the liquid volume to within +_ 1 ml. Alternatively, determine
the weight of the liquid to within +_ 0.5 g  by using a  balance.  The volume
or weight of liquid present, along with a notation of any color or  film observed
in the impinger catch, is recorded in the log.  This information is  needed,
along with the silica gel data, to calculate the stack  gas  moisture content
(see Method 5, Figure 5-3).
     4.  Transfer the contents of the first three impingers to Container No.  4.
     5.  Pour approximately 30 ml of 0.1N HNOg into each of the first three
Impingers and agitate the Impingers.  Drain the 0.1N HN03 through the outlet
arm of each Impinger Into the No. 4 sample container.  Repeat this  operation
a second time; Inspect the impingers for any abnormal conditions.
     6.  Wipe the ball joints of the glassware connecting the impingers free
                                      102

-------
I

I
•           of silicone grease and rinse each piece of glassware twice with 0.1N HNO,;
             this rinse is collected in Container No.  4.   (Do not rinse or brush the glass-
•           fritted filter support. )
                  Mark the height of the fluid level to determine whether leakage occurred
•           during transport.  Label  the container to clearly identify its contents.
m                Note:  In steps 5 and 6 above, the total  amount of 0.1N HN03 used for
             rinsing must be measured  and recorded.
I                5.3        Analysis
                  5.3.1      Container No.  3.   This step may be conducted in the field.
|           Weigh the spent silica gel (or silica gel plus impinger) to the nearest 0.5g
«           using a balance.
                  5.3.2       Lead Sample Preparation  and Analysis
I                5.3.2.1     Container No. 1 .  Cut the filter into strips and transfer the
             strips and all loose particulate  matter into 125-ml  Ehrlenmeyer Flask.  Rinse
|           the petri dish with 10 ml of 50%  nitric acid to insure a quantitative transfer
_           and add to the flask.  (Note:   if the total  volume required in Section 5.3.2.3
             will exceed 80 ml, it will be  necessary to use a 250-ml  Ehrlenmeyer flask in
•           place of the 125-ml Ehrlenmeyer flask.)
                  5.3.2.2.   Containers No. 2  and No.  4.   Combine the contents of Containers
|           No. 2 and No. 4 and take  to dryness on a  hot plate.   (Note:  Prior to analysis,
_           the liquid level in Containers No.  2 and/or No. 4 should be checked; confirmation
™           as to whether or not leakage occurred during transport should be made on  the
•           analysis sheet.  If a noticeable  amount of leakage has occurred, either void
             the sample or take steps, subject to the  approval  of the Administrator, to
J           correct the final results.)
                  5.3.2.3   Sample Extraction  for Lead.  Based on the approximate stack gas

                                                  103
-

-------
particulate concentration and the total  volume of stack  gas  sampled,  estimate
the total  weight of sample collected.   Now transfer the  residue  from  Containers
No. 2 and  No.  4 to the 125-ml  Ehrlenmeyer flask that contains  the  filter using
a rubber policeman and 10 ml  of 50% (V/V) HN03 for every 100 yg  of sample
collected  in the train or a minimum of 30 ml  of 50% HNO~,  whichever is  larger.
     Place the Ehrlenmeyer flask on a  hot plate and heat with  periodic  stirring
for 30 min. at a temperature just below boiling.   If the sample  volume  falls
below 15 ml, add more nitric acid.  Add 10 ml  of 3% H?0? and continue heating
for 10 min.  Add 50 ml of hot (80°C) distilled deionized water and heat for
20 min.  Remove flask from heat and allow to  cool.  Filter the sample through
a Millipore membrane filter or equivalent and transfer the filtrate to  a 250-ml
volumetric flask.  Dilute to volume using distilled, deionized water.
     5.3.2.4   Filter Blank.   Determine a filter blank using two filters from
each lot of filters used in the sampling train.  Cut each filter into strips
and place  each filter in a separate 125-ml Ehrlenmeyer flask.  Add 15 ml
of 50% (V/V) HN03 and treat as described in Section 5.3.2.3 (Extraction for
Lead)using 10 ml of 3% H^ and 50 ml  of hot, distilled, deionized water.
Filter and dilute to a total  volume of 100 ml  using distilled, deionized water.
     5.3.2.5   0.1 N Nitric Acid Blank. Take  the entire  200 ml of 0.1 N HN03  to
dryness on a steam bath, add 15 ml of 50% (V/V) HN03, and treat  as described
in Section 5.3.2.3 (Extraction of Lead) using 10 ml of 3% H202 and 50 ml of
hot, distilled, deionized water.  Dilute to a total volume of 100 ml  using dis-
tilled, deionized water.
     5.3.2.6   Lead Determination.  Calibrate the spectrophotometer as  described
in Section 6.1 and determine the absorbance for each source sample, the filter
blank and 0.1N HNO~ blank.  Analyze each sample three times in this manner
                                    104

-------
 I

 I

 •        Make  appropriate  dilutions,  as required, to bring all sample lead con-
 m        centrations  into  the  linear  absorbance  range of the  spectrophotometer.
                If  the  lead  concentration of a  sample is at the low end of the calibration
 •        curve and  high  accuracy  is required, the sample can  be taken to dryness on a
           hot plate  and the residue dissolved  in  the appropriate volume of water to
 I        bring it into the optimum range  of the  calibration curve.
 g             5.3.2.7   Mandatory Check for Matrix Effects on the Lead Results.  The
           analysis for lead by  atomic  absorption  is sensitive  to the chemical composition
 I        and to the physical properties (viscosity, pH) of the sample (matrix effects).
           Since the  lead  procedure described here will be applied to many different sources,
 B        it can be  anticipated that many  different sample matrices will be encountered.
 •        Thus, it is  mandatory that at least  one sample from  each source be checked using
           the Method of Additions  to ascertain that the chemical composition and physical
 ft        properties of the sample did not cause  erroneous analytical results.
                Three acceptable "Method of Additions" procedures are described in the
 |        General  Procedure Section of the Perkin Elmer Corporation Manual.  If the
 _        results  of the  Method of Additions procedure on the  source sample do not agree
 ™        within 5%  of the  value obtained  by the  conventional  atomic absorption analysis,
 •        then  all  samples  from the source must be reanalyzed  using the Method of Additions
           procedure.
 •         6.    Calibration
 m              Maintain a laboratory log of all calibrations.
                6.1        Sampling  Train Calibration.  Calibrate the sampling train com-
 •         ponents  according to  the indicated sections of Method 5 (40 CFR 60 Appendix A):
           probe nozzle (Section 5.1);  Pitot tube  assemble (Section 5.2); metering system
I
105

-------
(Section 5.3); probe heater (Section 5.4);  temperature gauges  (Section  5.5);
barometer (Section 5.7).   Note that the leak-check of the metering  system
(Section 5.6 of Method 5) applies to this method.
     6.2       Spectrophotometer.  Measure  the absorbance of the standard
solutions using the instrument settings recommended by the Spectrophotometer
manufacturer.  Repeat until good agreement  is obtained between replicates.
Plot the absorbance (y-axis) versus concentration  in yg Pb ml"  (x-axis).
Draw or compute a straight line through the linear portion of the curve.   Do
not force the calibration curve through zero, but  if the curve does not pass
through the origin or at least lie closer to the origin than +_ 0.003 absorbance
units, check for incorrectly prepared standards and for curvature in the
calibration curve.
     To determine stability of the calibration curve, run a blank and a standard
after every five samples and recalibrate, as necessary.
7.   Calibrations
     7.1       Nomenclature.
                            2
     A      =  Stack area, m
     (Pb)   =  Total yg of lead in the source samples after correcting for all
               dilutions.
     P.      =  Barometric pressure at the sampling site, mm Hg.
     P      =  Absolute stack gas pressure, mm Hg.
     R      =  Rate of lead emission, g/day.
     T      =  Absolute average dry gas meter temperature, K.
     T      =  Absolute stack temperature,  K.
     v      =  Average stack gas velocity,  m/sec.
                                    106

-------
1
1


1

1



1

1

1



1

1

I


1




vm
III
Vtotal
Y
AH
7.

leakage
Appendi
7.












Total volume of gas sample as measured by the dry gas meter,
corrected
Total gas
for
leakage, m .


•3
sample volume (stack conditions), m .
= Dry gas meter
=
2

Average
calibration factor.
pressure differential across the orifice meter, mm H^O.
Calculate

, if necessary,
x A).
3
the stack gas,


as

Calculate
from data
(5.3) of Method 5, 40 CFR
7.
4
Calculate,
(2-9) of Method 2, 40 CFR


pressure, and
7.
5
using the foil

V

7.
average
blank.

total

6


V
in
the total volume of dry gas metered (corrected


for

outlined in Section 6.3 of Method 5, 40 CFR 60,

the


volume of water vapor and the moisture content

of
obtained in this testing, use Equations (5.2) and
60,
V
60,

Appendix A.


the average stack gas velocity, using Equation
Appendix A; use velocity head (AP), tempe-ature,

moisture data from this field
Calculate
the

test.


total gas sample volume at stack conditions,
owing equation:

• v



T^~

Total Lead
»
s_

m
in
D 4. AH
Hbar TO"
i^
ps
m •
Source Sample.
absorbance for the contribution of
Use the calibration curve and this
the lead concentration
1


1


Calculate
correcting for
of the
7.


sample
7


in
the total
all the di
into the

(A-l)

For each source sample correct



the
the filter b<"ank and the 0.1 N HN03
corrected ibsorbance to determine

the sample aspirated intc the spectrophotometer.
lead content in the orijlnal source sample (Pb)Q;
lutions that were
linear range of the
Total Lead




mack to bring the lead concentration
cpectrophotometer.

Emission. Calculate the total amount of lead emitted


W






-------
from each stack per day by Equation (A-2).   This equation is applicable for
continuous operations.   For cyclic operations, use only the time per day each
stack is in operation  The total  lead emissions from a source will  be the
summation  of results from all  stacks.
           R  =
(Pb)0 v,As
W
total

86400 seconds/day
106 yg/g
                                                                (A-2)
8.   Isokinetic Variation.   Determine the isokinetic variation in the sampling
""ate using Equation (5-7) of Method 5, 40 CFR 60, Appendix A and the raw data
 fhm this testing.
     8.1       Acceptable Isokinetic Results.  The following range sets the
 limit on acceptable isokinetic sampling results:
      If 90% <_ I <_ 100%, the results are acceptable.  If the results are low
 in  compar'.son with the emission standard and I is beyond the acceptable
 range,  or ii i is less than 90%, the Administrator may opt to accept the
 results.   Otherwise, reject the results and repeat the test.
 9.   Alternate T
-------
I

I

•         method  provided  that  the  filter  is  included  in  the analysis  for  lead.
•              9.3       In-Stack Filter.   Use  of  an in-stack  filter is an acceptable
           alternate  method provided that:  (1) the  in-stack  filter  is followed by a  glass-
•         lined probe and  at  least  two  impingers that  each  contain  100 ml  of 0.1N HNO,;
           and (2)  the probe and impinger contents  are  recovered and analyzed for lead.

           10.   Bibliography
|              Analytical  Methods for Atomic  Absorption Spectrophotometry,  Perkin  Elmer
           Corporation,  Norwalk, Connecticut,  September 1976.
•              Annual  Book of ASTM  Standards.   Part 31; Water, Atmospheric Analysis.
•         American Society for  Testing  and  Materials,  Philadelphia, Pa.,  1974.  pp. 40-42,
                Code  of Federal  Regulations.   Title 40, Part 60, Appendix A "Reference
£         Methods."   (As amended in the Federal Register  of August  18, 1977, pp. 41754-
           41789.)
•              Klein,  R. ,  and C. Hach,  "Standard Additions  - Uses  and  Limitations in
•         Spectrophotometric  Analysis," Amer. Lab. 9:  21-27 (1977).
                Martin,  Robert M. "Construction  Details of Isokinetic Source-Sampling
•         Equipment."   APTD-0581.   U.S. Environmental  Protection Agency, Research Triangle
           Park, N.C.,  April 1971.
B              Mitchell, W.O.,  and  M.R. Midgett, "Test Methods to  Determine the Inorganic
•         and the  Alkyl  lead  Emissions  from Stationary Sources," in preparation.
                Rom,  Jerome J.,  "Maintenance,  Calibration, and  Operation of Isokinetic
•         Source  Sampling  Equipment," APTD-0576,   U.S. Environmental Protection Agency,
           Research Triangle Park, N.C., March 1972.
•              Smith,  W.S., R.T. Shigehara, and W.F. Todd,  "A  Method of Interpreting
•j         Sampling Data, " Paper presented at  the 63d Annual  Meeting  of the Air Pollution

                                                109
•

-------
Control  Association, St.  Louis,  Mo., June 14-19,  1970.
     Smith, W.S., et al., "Stack Gas Sampling Improved  and Simplified With
New Equipment,"  APCA Paper No.  67-119,  1967.
     Shigehara, R.T., "Adjustments in the EPA Nomograph for Different Pitot
Tube Coefficients and Dry Molecular Weights,"  Stack Sampling News  2:  4-11
(1974).
                                     110

-------
I
I
I
I
I
I
I
I
•
_


I

I

I

I

I

I

I

I
                                             APPENDIX B


                                PROCEDURE FOR DETERMINING THE ALKYL LEAD

                           EMISSIONS FROM ALKYL LEAD MANUFACTURING PLANTS

-------
I
I
I
                              PROCEDURE FOR DETERMINING THE ALKYL LEAD
                          EMISSIONS FROM ALKYL LEAD MANUFACTURING PLANTS
I
I
I
I
I
I
I
I
                               U. S. ENVIRONMENTAL PROTECTION AGENCY
|                               OFFICE OF RESEARCH AND DEVELOPMENT
                          ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
I                         RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711
I                                          MARCH 1978
i
i

-------
             PROCEDURE FOR DETERMINING THE ALKYL LEAD EMISSIONS
                    FROM ALKYL LEAD MANUFACTURING PLANTS

1.    Principle and Applicability
     1.1        Principle.  Alkyl  lead emissions are extracted from the
source, collected in IC1 solution, and analyzed by atomic absorption spectro-
metry using an air-acetylene flame.
     1.2       Applicability.   This method is applicable  for determining
the lead emissions from alkyl  lead manufacturing plants.

2.    Range, Sensitivity, Precision, and Interferences
     The values given below are typical of the method's capabilities.
Absolute values will vary for individual situations depending on the type of
instrument used, the lead line, and operating conditions.
     2.1       Range.  The upper range of the method is unknown, but, in
theory, the test method as described here can quantitatively collect approx-
imately 10 g of tetraethyl lead.  The upper range could be increased by
increasing the volume of IC1 used and by adding additional impingers.  The
lower limit of the method as described here is approximately 10 yg of lead
alkyl as lead, if the Alternate Analytical Procedure (Section 8.1) is used
and the final dilution  is 25 ml.
     2.2       Analytical sensitivity.  Typical sensitivities for a 1% change
 in absorption (0.0044 absorbance units) are 0.2 and 0.5 yg Pb ml~  for the
 217.0 and  283.3 nm  lines, respectively.
                                     114

-------
I
I
I              2.3       Precision.  The within-laboratory precision, measured as
           the coefficient of variation, was determined at an alkyl lead manufacturing
I         plant vent stack using four trains sampling simultaneously.  The average
_         run concentration for the five sampling runs was 265 to 369 yg Pb m" .
*         The percent coefficient of variation for each run, which is determined
•         by expressing the standard deviation of the run as a percentage of the
           run mean concentration, ranged from 1.8 to 8.2%.  The average precision for
P         all five runs was 3.8% of the mean concentration.
_              2.4       Interferences.  Sample matrix effects may interfere with
•         the analysis for lead by flame atomic absorption.  If the analyst suspects
•         that the sample matrix is causing erroneous results, the presence of these
           matrix effects can be verified and frequently corrected for by carrying
•         out the analysis using the Method of Standard Additions on samples previously
           treated using the Alternate Analytical Procedure (Section 8.1)

•         3.   Apparatus
                3.1       Sampling Train.  The sampling train is shown in Figure B-l ,
•         and component parts are discussed below.
                3.1.1     Probe.  Glass, Nylon or Teflon are acceptable.
»              3.1.2     Impingers.  Smith-Greenburg.  The two Smith Greenburg
•         impingers must be connected in series using a leak-free glass connector.
           Silicone grease may be used if necessary to prevent leakage.
•              3.1.3     Acid Trap.  Mine Safety Appliances Air Line Filter,
           Catalogue Number 81857 with acid absorbing cartridge and suitable connections,
•         or equivalent.
mm              3.1.4     Temperature Gauge.   Dial  thermometer, thermocouple, or
i

-------
                                        c
                                        'to
                                          .
                                        E
                                        CD

                                         01
                                         L_

                                         O)
116

-------
 I

 I
 •          equivalent to measure the temperature of the gas in the stack, in the impinger
             bath and in the dry gas meter to 3°C(5.4°F).
 m               3.1.5     Valve.  Needle valve to regulate sample gas flow.
 •               3.1.6     Pump.  Leak-free diaphram pump, or equivalent, to pull gas
             through the train.  It is suggested that a small tank be installed between
 •          the pump and the rate meter to eliminate the pulsation effect of the diaphram
             pump on the rate meter.
 m               3.1.7     Volume Meter.  Dry gas meter, sufficiently accurate to measure
 •          the sample volume within 2%, calibrated at the selected flow rate and
             conditions actually encountered during sampling, and equipped with a temp-
 •          ature gauge (dial thermometer, or equivalent) capable of measuring temperature
             to within 3°C (5.4°F).
 •               3.1.8     Barometer.  Mercury, aneroid, or other barometers capable
 «          of measuring atmospheric pressure to within 2.5 mm Hg (0.1 in. Hg).  In many
             cases, the barometric reading may be obtained from a nearby National Weather
 V          Service station, in which case the station value (which is the absolute
             barometric pressure) shall be requested, and an adjustment for elevation
 |          differences between the weather station and sampling point shall be applied at
 _          a rate of minus 2.5 mm Hg (0.1 in. Hg) per 30 m (100 ft) elevation increase or
 *          vice versa for elevation decrease.
 •               3.1.9     Pi tot Tube.  Type S, as described in Section 2.1 of Method
             2, 40 CFR 60, Appendix A, or other device approved by the Administrator.  The
 I          Pi tot tube shall be attached to the probe to allow constant monitoring of the
 ._           stack gas velocity during testing for lead.  The Type S pitot tube assembly
 •           shall have a known coefficient, determined as outlined in Section 4 of Method
 •           2 (40 CFR 60, Appendix A).
i

-------
     3.1.10     Differential  Pressure Gauge.   Inclined manometer or
equivalent device as described in Section 2.2 of Method 2, 40 CFR 60,
Appendix A, to be used for velocity head (Ap) readings.
3.2  Sample Recovery
     3.2.1     Wash bottle.   Polyethylene or glass, 500 ml.
     3.2.2      Storage bottles.   Polyethylene or glass, 500 ml, narrow
mouth with screw cap closures.
3.3  Analysis
     3.3.1     Volumetric Flask.   Class A with penny head standard taper
stoppers.  100, 250, 500, and 1000 ml.
     3.3.2     Volumetric Pipets.  Class A.   1,2, 3, 4, and 5 ml.
     3.3.3     Graduated Cylinder.  100 ml.
     3.3.4     Atomic Absorption Spectrophotometer.  Equipped with lead
hollow cathode or electrodeless discharge lamp.
     3.3.5    Acetylene.  The grade recommended by the instrument manu-
facturer should be used.  Change cylinder when pressure drops below 50-
100 psig.
     3.3.6    Air.  Filtered to remove particulate, oil, and water.
     3.3.7    Cleaning.  All  glassware should be scrupulously cleaned.
The following procedure is suggested.  Wash with laboratory detergent,
rinse, soak for 4 hr in 20% (v/v) HNO.,, rinse 3 times with distilled-
deionized water, and dry in a dust free manner.
4.   Reagents
     4.1       Sampling
     4.1.1     Distilled Water. Meeting ASTM specifications for Type 1
                                     118

-------
I

I
•         Reagent Water-ASTM Test Method D-l193-72.  This water must be used in all
           dilutions and solution preparations.
g             4.1.2     Hydrochloric Acid.  Concentrated ACS reagent grade or equivalent.
_             4.1.3     Potassium  Iodide.  Reagent grade.
"             4.1.4     Potassium  Iodide Solution, 25% by weight.  Dissolve 250 9  of
•         potassium iodide in distilled water and dilute to 1 liter with distilled
           water.
|             4.1.5     Potassium  lodate.  Reagent grade.
               4.1.6     Iodine Monochloride Stock Solution, l.OM.  To 800 ml of  25%
•         potassium iodide solution  (Reagent 4.1.4) add 800 ml of concentrated
•         hydrochloric acid (Reagent 4.1.2).  Cool to room temperature.  With vigorous
           stirring, slowly add 135.5  g of potassium iodate and continue stirring until
I         all  free iodine (dark blue color) has dissolved to give a clear, orange-red
           solution.  Cool to room temperature and dilute to 1800 ml with distilled water.
•         The  solution should be kept in the dark to avoid degradation.
•             4.1.7     Iodine Monochloride Absorbing Solution for Lead Alkyl Compounds,
           0.2  M.  Dilute 200 ml of  the IC1 stock solution (Reagent 4.1.6) to 1000 ml
•         using distilled water.  Store in glass bottles in the dark.  Discard any unused
           solution after 60 days.
W             4.2       Sample Recovery.
•             4.2.1     All sample exposed surfaces should be washed with 0.2 M  IC1
           (Reagent 4.1.7).
•             4.3       Analysis.
               4.3.1     Nitric Acid.  Concentrated ACS reagent grade or equivalent.
•             4.3.2     Sample Dilution Reagent, 0.2 M Id (Reagent 4.1.7).
m             4.3.3     Lead Nitrate.  Reagent grade.

i

-------
     4.3.4     Stock Lead Solution, 1000 yg Pb ml"1.  Dissolve 0.1598  q of
lead nitrate (Reagent 4.3.3) in 60 ml of distilled water, add 2 ml  of con-
centrated HN03 (Reagent 4.3.1) and dilute to 100 ml using distilled water.
Store in polyethylene bottle.  Commercially available certified lead standard
solutions may be used.
     4.3.5     Working Standard Lead Solutions for Spectrophotometer Calibration,
0, 4, 8, 12, 16, and 20 yg Pb ml"1.   Pipet 0.0, 1.0, 2.0, 3.0, 4.0, and 5.0 ml
aliquotsof the stock lead standard solution (Reagent 4.3.4) into 250 ml  volu-
metric flasks.  Add 5 ml concentrated HN03 to each flask and dilute to volume
with distilled, deionized water.  These working standards contain 0.0, 4.0, 8.0,
12.0, 16.0, and 20.0 yg Pb ml  , respectively.  Additional standards at other
concentrations should be prepared in a similar manner as needed.
5.   Procedure
     5.1       Sampling.
     5.1.1     Preparation of Collection Train.  Add 125 ml of 0.2 M ICL
to each Smith-Greenburg impinger.  Assemble the train as shown in Figure B-l.
Place crushed ice and water around the impingers.
     5.1.2     Leak-check Procedure.  A leak check prior to the sampling run
is optional; however, a leak check after sampling is completed is mandatory.
The leak-check procedure is as follows:
                                                3     1
     Temporarily attach a suitable (e.g. 0-40 cm  min   rotameter to the outlet
of the dry gas meter and place a vacuum gauge at or near the probe inlet.
Plug the probe inlet, pull a vacuum of at least 250 mm Hg  (10 in. Hg), and
note the flow rate as indicated by the rotameter.  A leakage rate not in excess
of 2% of the average sampling rate is acceptable.  Note:   carefully release
the probe inlet plug before turning off the pump.

                                      120

-------
I
I
•               It is suggested (not mandatory) that the pump be leak checked separately.
            If done prior to the sampling run, the pump leak check shall precede the leak
•          check of the sampling train described immediately above, if done after the
            sampling run, the pump leak check shall  follow the train leak check.  To leak
I          check the pump, proceed as follows:  Disconnect the drying tube from the
M          probe-impinger assembly.  Place a vacuum gauge at the inlet to either the
            drying tube or the pump, pull a vacuum of 250 mm Hg (10 in. Hg), plug or
•          pinch off the outlet of the flow meter,  then turn off the pump.  The
            vacuum should remain stable for at least 30 sec.
|               5.1.3     Preliminary Determinations.  Select the sampling site and the
«          number of points for a velocity traverse of the stack according to Method 1,
            40 CFR 60, Appendix A or as specified by the Administrator.  Determine the
•          stack pressure, temperature, and range of velocity head using Method 2, 40 CFR
            60, Appendix A; it is recommended that a leak check of the Pitot lines (See
|          Method 2, Section 3.1) be done.  Use these measurements to select a point
            in the stack at which to position the inlet to the probe and to determine
            the sampling rate that will be used in the testing.
                 If the volumetric flow in the stack is not expected to change by more
            than +_ 10% during a test run, then single point, constant rate sampling
            will be acceptable.  If the volumetric flow is expected to change by more
            than +_ 10% during the run, then single point, proportional sampling shall be
            required unless directed otherwise by the Administrator.
                 5.1.4     Sample Collection.  Record the initial dry gas meter volume
            reading and temperature, the stack velocity and temperature, and the barometric
            pressure.  Use the criteria in Section 5.1.3 (Preliminary Determinations) to
            decide if proportional sampling or constant-rate sampling will be employed.

-------
     To begin sampling,  position the tip of the sampling probe at  the  sampling
point, connect the probe to the bubbler, and start the pump.   Adjust the
sample flow rate (to the desired proportional  rate or the constant rate as
necessary) to obtain a flow of 1.5 to 3.0 1  min   through the  train as
indicated by the rotameter.   (If proportional  sampling is used, maintain  the
flow through the train within +_  10% of the desired proportional flow  rate.
If constant rate sampling is used, maintain a  constant flow rate within
+_ 10% during the sample.)
     At 15-min intervals or shorter if necessary because of stack  conditions,
record the following information:   (1)  dry gas meter and stack temperature
(2)  volume reading of the dry gas meter and (3)  indicated rotameter  flow
rate.  As necessary add more ice to the impinger box to maintain a temperature
in the impinger box between 0° and 10°C.  At the conclusion of each run,  turn
off the pump, remove probe from the stack, and record the final readings.
Conduct a leak check as in Section 5.1.2.  (This leak check is mandatory).
If a leak is found, void the test run and repeat the run.
     Samples should be taken over such a period or periods as  necessary to
accurately determine the maximum emissions that occur in a 24-hr period.
In the case of the cyclic operations, sufficient tests shall  be made  to allow
accurate determination of the emissions that occur over the duration  of the
cycle.
     5.2       Sample Recovery.  Disconnect the probe from the impingers  and
transfer the impinger contents to a 500-ml polyethylene (or glass) bottle.
Wash each impinger twice with 25 ml of 0.2 M Id solution and add  the washes
to the polyethylene or glass bottle.  Wash the probe with a minimum of 50 ml
of 0.2 M  Id and add this wash to the sample recovery bottle that  contains
                                      122

-------
 I
 I
 I         the impinger  wash.   Seal  and  identify  the container and mark the fluid level on
 ^         the bottle or record the  total weight.
                 For a blank  place  50 ml  of  the  0.2 M Id  solution in a separate sample
 fl         recovery bottle.
                 5.3       Sample Analysis.   Note  the level of liquid in the container or
 |         the weight of the container and  confirm if any sample was lost during shipment.
 £         Note this on  the  data sheet.   If a noticeable  amount of leakage has occurred,
 *         either  void the sample  or use methods, subject to the approval of the
 •         Administrator,  to correct the final  results.
                 Transfer the contents of the storage container to a 500-ml volumetric
—
*
I
 |          flask  and  dilute  to  volume with  0.2 M  Id.   If the expected concentration
 —          of lead  in the  sample exceeds  15 yg Pb ml"  , dilute an aliquot of the sample
 *          with 0.2 M IC1  to bring the lead concentration into the linear absorbance range
 •          of the spectrophotometer as indicated  by the manufacturer.
                 Calibrate the spectrophotometer as described in Section 6.5 and analyze
            the  source  samples by  aspirating them into the flame.  Record the equilibrium
            absorbance.  Analyze each sample three times in this manner and average the
            absorbances.
•               Similarly, aspirate a sample of the IC1 blank and record the absorbance.
            6.   Calibration
I
m              Maintain a laboratory log of all calibrations.
™              6.1       Metering System
•              6.1.1     Initial Calibration.  Before its initial use in the field,
           first leak check the metering system (drying tube, needle valve, pump, rotameter,
•         and dry gas meter) as follows:  place a vacuum gauge at the inlet to the drying
                                                123

-------
tube and pull a vacuum of 250 mm (10 in.) Hg;  plug or pinch off the outlet of
the flow meter, then turn off the pump.   The vacuum shall  remain stable
for at least 30 sec.  Carefully release  the vacuum gauge before releasing
the flow meter plug.
     Next, calibrate the metering system (at the sampling flow rate specified
by the method) as follows:   connect an appropriately sized wet test meter
(e.g., 1 liter per revolution) to the inlet of the drying tube.  Make three
independent calibration runs, using at least five revolutions of the dry
gas meter per run.  Calculate the calibration  factor, y (wet test meter
calibration volume divided  by the dry gas meter volume, both volumes adjusted
to the same reference temperature and pressure), for each run and average
the results.  If any y value deviates by more  than 2% from the average,
the metering system is unacceptable for  use.  Otherwise, use the average as
the calibration factor for  subsequent test runs.
     6.1.2     Post Test Calibration Check.  After each field test series,
conduct a calibration check as in Section 6.1.1 above, except for the
following variations:  (1)   the leak check is  not to be conducted, (2) three,
or more revolutions of the  dry gas meter may be used, and (3)  only two
independent runs need be made.  If the calibration factor does not deviate
by more than 5% from the initial calibration factor (determined in Section 6.1.1),
then the dry gas meter volumes obtained  during the test series are acceptable.
 If the calibration factor deviates by more than 5% recalibrate the metering
system as in Section 6.1.1  and for the calculations use the calibration factor
(initial or recalibration)  that yields the lower gas volume for each test run.
     6.2       Thermometers.  Calibrate  against mercury-in-glass thermometers.
     6.3       Rotameter.  The rotameter need not be calibrated but should be
                                      124

-------
 I
 I
I
           cleaned and maintained according to the manufacturer's instruction.
                6.4       Barometer.  Calibrate against a mercury barometer.
 •             6.5       Spectrophotometer.  Measure the absorbance of the standard
           solutions using the instrument settings recommended by the spectrophotometer
 |        manufacturer.  Repeat until good agreement is obtained between replicates.
 •        Plot the absorbance (y-axis) versus concentration in yg Pb ml    (x-axis).
           Draw or compute a straight line through the linear portion of the curve.
 •        Do not force the calibration curve through zero, but if the curve does not
           pass through the origin or at least lie closer to the origin than +_ 0.003
 |        absorbance units, then check for incorrectly prepared standards and for
 M        curvature in the calibration curve.
                To determine the stability of the calibration curve, run a blank and a
 •        standard after every five samples and recalibrate as necessary.
           7.   Calculations
                       2
A         Stack area, m
 |             7.1       Nomenclature.
 _
                         (pb)0    = Total yg lead in the source samples after
 •                                 correcting for all dilutions.
                          p
                           bar    = Barometric pressure at the sampling site, mm Hg.
                          IP
                           s      = Absolute stack gas pressure, mm Hg.
R       = Rate of lead emission, g/day.
 m      = Absolute average dry gas meter temperature, K.
 s      = Absolute stack temperature, K.
vs      = Average stack gas velocity, m/sec.
 m      = Total volume of gas sample as measured by the dry gas
          meter.
I                                    IIIV* VW < *
                           Y      = Dry gas meter calibration factor.
i

-------
     7.2       Dry Gas Volume.   Correct the sample volume measured  by the dry
gas meter to stack conditions:
                      "Total  •"•<$  %'   *     (*-"

     7.3       Stack Gas Velocity.   Calculate  the stack gas  velocity at stack
conditions using Equation (2-9)  of Method 2, 40 CFR 60  Appendix A;
use velocity head, temperature,  and pressure data from the alkyl  lead testing.
     7.4       Total Lead in  Source Sample. For each source sample correct
the average absorbance for the  contribution of the Id blank.
Use the calibration curve and this corrected absorbance to determine the
lead concentration in the sample as aspirated  into the spectrophotometer.
     Calibrate the total lead content (yg) in  the original source sample
(Pb) ; correcting for all dilutions made to bring the lead concentration into
the linear range of the spectrophotometer.
     7.5       Total Lead Emission.  Calculate the total amount of lead
emitted from each stack per day using Equation (B-2).  This  equation is applicable
for continuous operations.  For cyclic operations, use only the time per day
each stack is in operation.  The total  lead emissions from a source will be
the summation of results from all stacks.
                                 	    r                   (B-2)
                                     J  L106 yg/g   .
8.   Alternate Analytical Method
     8.1       Alternate Analytical Method. The source samples may be
analyzed for lead after converting all  lead to an inorganic form.  This can
be done by gently evaporating an aliquot of the original source sample to
dryness on a hot plate at low heat, heating on low heat until  all color has
                                     126
(PbL V. A
0 5
VTotal
s i
J
86400 sec/day
. 106 yg/g

-------
I

I
•          disappeared,  dissolving the residue in 5%  (v/v) HNCL and diluting to known
            volume with 5%  (v/v) HNCL.  The  samples can then be analyzed for lead using
I
•          the  analysis procedure described in the test method.

I
            9.    Bibliography
I               Analytical Methods for Atomic Absorption Spectrophotometry, Perkin
            Elmer Corp. Norwalk,Connecticut,  September  1976.
•               Code  of Federal Regulations, Title 40, Part 60, Appendix A "Reference
m          Methods"  (As amended in the Federal Register of August 18, 1977. pp. 41754-
            41789).
•               Klein, R., and C. Hach, "Standard Additions, Uses and Limitations in
            Spectrophotometric Analysis", Amer. Lab 9:  21-27 (1977).
I
I
I
I
I
I
I
I
I
     Mitchell,  W.  J.,  and  M.  R.  Midgett,  "Study  to  Develop  Test  Methods  for
Inorganic and  Alkyl  lead Emissions  from  Stationary  Sources,"  in  preparation.
                                     127

-------
  I
  I
  I
  I
  I
  I
  I
 •                                       APPENDIX C
 I                          PROJECTING AUTOMOTIVE LEAD EMISSIONS
 g                               FOR ROADWAY CONFIGURATIONS
 I
 I
 I
 I
 I
 I
 I
I
m                                              129

-------
I

I                                  APPENDIX  C
              Projecting Automotive Lead Emissions for Roadway Configurations

              The characterization of automotive lead emissions has been described in
•      journal articles and in minimal EPA testing.  In this work, it is generally
•      agreed that  (1) 'Only a percentage of  the lead burned in the fuel is exhausted
        and  (2) that the percentage of lead exhausted varies with vehicle operating
I      conditions.  Experimental evidence shows that 70-80 percent of the lead burned
        in gasoline  is emitted into the atmosphere, while 20-30 percent is retained in
•      the  engine,  oil, and exhaust system  (Hirschler, et al., 1964, Ter Haar, et a1._,
•      1972).
              The Hirschler and Gilbert study obtained data from three passenger cars
I      during a city-type driving  cycle.  These data showed that, on the average, the
^      exhaust gas  contained from  28.2% to 44.8% of the lead  burned by the cars.  At
•      the  completion of exhausted lead tests  in all three cars, the amount of lead
•      left in various zones of  the engine and exhaust system was determined.  From
        21.1% to 27.8% of the lead  burned by  the cars was recovered from deposits in
•      the  exhaust  system and engine, and in the lubricating  oil and oil filters.
        Therefore, something less than 75% of the lead burned  could have been  present
•      in  the exhaust gas.
•            During  operation under a  test cycle of moderate speeds similar to city-
        type driving,  three cars  exhausted from 28% to 45% of  the lead burned.  When
•      exhaust  systems were fairly free of  deposits, the emissions of lead were in
        the  20-25% range, but after the cars  had been operated for several thousands
I      of  miles at  moderate speeds,  emissions  increased  into  the 35-55% range.  Indi-
M      vidual tests showed wide  variations  which indicate that a random factor, prob-
        ably deposit flaking in  the exhaust  system, is involved in the process of
•      exhausting lead.
                                                131

-------
      The Ter Haar study calculated a lead balance based on the assumption that
25% of the lead remains in the automobile.  In the emission tests,  an average of
21.4% of the lead burned was emitted from 26 cars on the Federal hot cycles,
while the cold cycles emitted 43.9% of the lead burned.   With the Federal  cycle
weighted 65% for the hot cycles and 35% for the cold cycles^ the calculated
percentage of burned lead emitted during the fall Federal  Cycle is  29.2%.
      The 29% for the full Federal cycle and the 25% which remains  in the car
accounted for 54% of the lead burned.  Ter Haar et al.  assume that  the remain-
ing 46% is emitted during heavy accelerations and decelerations.
      This analysis was tested in a total lead balance study of a car driven
on a mileage accumulation route for 12,000 miles.  The exhaust from the car
was measured with a total filter.  Ter Haar et al. accounted for 84.3% of the
lead burned.  They believe that the 15% unaccounted for was lost during han-
dling of the exhaust system and the total filter.  Although the total lead
exhausted overall was about 55% of the lead burned,  it should be noted that it
was only 36% for the first 6000 miles and increased to 73% for the last 6000
miles.  Therefore, if the exhaust system  is not  changed, it is likely that this
car will continue to exhaust  75% or more  of the  lead burned.

                           Emissions  During City-Type Driving
      Several  studies have been performed measuring  lead emissions under differ-
ent driving  conditions.   The  first comprehensive study of  automotive  lead emis-
sions was performed  by  Hirschler  and  Gilbert.  The  Hirschler and Gilbert  study
measured automotive  lead  emissions for constant  speed dynamometer  tests and a
-city-type driving cycle.  The city-type  driving  cycle was  a  repeated  cycle  of
 idling,  accelerating,  decelerating and  cruising  over 110 miles  at  an  average
                                           132

-------
 I
 m     speed of 22 miles per hour.  The cruise speeds were' 25 and 40 miles per hour.
        The city driving cycle data showed that, on the average, the exhaust qas con-
 •     tained from 28.2 to 44.8 percent of the lead burned by the cars.  These
        numbers were averaged to derive an emission factor of 0.365 (36.5% of burned
 |     lead is exhausted) to estimate the emissions for a city-suburban roadway
 _     configuration.
              In comparison, Bradow (1976) measured lead emissions from three 1970-
 •     1974 model year cars on the Federal Test Procedure cycle.  For these tests,
        the percentage of burned lead exhausted ranged from 10.8 percent to 21.4 per-
 |     cent.
 £           While the Hirschler and Gilbert data for city-suburban driving are from
        1953, 1954, and 1957 model year automobiles, it seems reasonable to assume
 •     that lead emissions from these cars are representative of more recent model
        year vehicles.  Figure 1 shows that the Hirschler and Gilbert data for cruise-
 |     speed driving are within the bounds of the data presented in  other studies
 _     on the percentage of burned lead exhausted.  The fact that the Hirschler and
 *     Gilbert numbers in Figure 1 are reasonable is important since these are the
•
        most complete data that represent city-suburban driving.
 |                            Emissions During Cruise-Speed Driving
 m           Habibl  (1973) tested a 1966 model year vehicle using the Federal mileage
 •       accumulation schedule (Federal Register, 1968) under steady-state constant-
 •       speed  conditions.   Exhaust measurements were made at four nominal test
         mileages during the mileage accumulation schedule.  The steady-state operation
 •       emissions were measured at 20,45, and 70 miles/hour.  The tests ranged from
         200  to 400 miles  in duration and were run on a fuel containing 3 grams of lead

i
i

-------
per gallon.   The lead emissions from these tests are shown in Figure 1.   The
ranges and average emission rates for each of the three cruise speeds are shown
on the graph.
      Ter Haar et a!. studied automotive exhaust particulates by testing 26
cars.  Ten were owned by the Ethyl  Corporation in commuter service and 16 were
employee-owned.  The  Ethyl  Corporation cars were 1963-1968 models that had been
driven from 20,000 to 62,000 miles.   The employee cars were all 1966 models with
17,000 to 92,000 miles of service.   The test conditions included cruise at 25,
45, and 60 ir,iles/hour.  The 26-car average results for the three speeds are
shown in Figure 1.
      Sampson and Springer (1973) and Ganley and Springer (1974) measured auto-
                               3
motive lead using a 1970 350-in  Chevrolet V-8 production engine mounted on a
dynamometer.  A simulated exhaust system was connected to the engine.  The
majority of the tests were performed at a load equivalent to a full-size 1970
Chevrolet crjising at 55 miles/hour.  For each of the two studies, an average
lead emissioi rate is shown in Figure 1.
      Dr. ROT Bradow of EPA, ESRL tested four in-use vehicles for lead emissions
in 1976.  These four vehicles  ranged from 1963-1974 model years and  had 37,000-
54,000 accumulated miles.  Twenty-nine test runs were performed using the
Sulfate Emissions Test, the Fuel Economy Test, and  the Federal Test  Procedure
(Federal  Register No. 221, 1972).  The lead emissions data from the  Sulfate
Emissions Test and the Fuel Economy Test are  included  in  Figure 1 because  these
tests are representative of cruise-speed driving.
      The automotive  cruise speed emissions data  from journal  articles and  EPA
testing were  used to  develop a curve for determining lead emission  factors.
This curve  is  a  conservative estimate of future year lead emissions  and  is
shown on  Figure  1.
                                      134

-------
 I
 I
                                    Emissions  During Acceleration Periods
 ^                  Data  collected  by  Hirschler  and Gilbert and Ter Haar et al. on lead emis-
 *            sions  from  automobiles show  that tt\e highest emission rates occur when cars
 B            accelerate  from  a  stop or  a  low  cruise speed (20 mph) to a much  higher speed
              (60  mph).   Therefore,  an area where vehicles are accelerating on expressway
 |            access ramps  onto  heavily  traveled expressways  has the potential for high con-
 —            centrations of ambient lead.  The  accurate modeling of such a situation  is
 •            limited due to the scarcity  of data on acceleration emissions.   However, data
 •            from the journal articles  by Hirschler and Gilbert and Ter Haar  can be used to
              estimate an emission  factor  for  accelerations to 60 miles per hour.
 •                  Hirschler  and Gilbert  conducted full-throttle acceleration tests on a
              single exhaust 1954 car  and  a dual exhaust 1953 car.  A series of tests were
 •            made on the 1954 model year  car  after 25,000 miles of deposit accumulation in
 •            the  exhaust system.  In  the  first  test,  a series of three full-throttle accel-
              erations from 20 to 60 mph were  made, and the lead exhaust was collected and
 •            classified.  This  was followed by  a test in which exhausted lead was collected
              from a series of nine similar full-throttle accelerations.  The same type of
 •            tests were  run on  the 1953 model year car.  However, the data from the 1953 car
•            are  less complete  because  the sample from the second test of nine accelerations
              was  lost.   The results of  the Hirschler  and Gilbert tests are shown below:
I
I
I
I
I
I
                                                      %  burned  lead  emitted
                                                  Single Exhaust   Dual  Exhaust
                                                    1954 car         1953  car
                                                   (12 tests)       (3  tests)
                  Full-throttle acceleration        870 to 1230         1990
                        20 to 60 mph
                                                 135

-------
      Ter Haar et al. tested 26 1963-1968 model year cars for acceleration emis-
sions.  Ten of these cars were owned by the Ethyl Corporation in commuter service.
These cars ranged from model year 1963-1968 and in accumulated mileage from 20,000-
62,000 miles.  The other 16 cars were .employee-owned 1966 model year cars with
17,000-92,000 accumulated miles.  The results of the Ter Haar study indicate
that an average of 1119% of the lead burned is exhausted during full throttle
 acceleration  (0-70  and  30-70 mph).
      Deriving a specific emission factor for acceleration conditions from the
aforementioned studies  is a difficult task.  Ter Haar's data seem to be more
useful because 26 different cars were tested and more recent model year cars
were used.  While Hirschler and Gilbert made multiple runs, only two cars were
tested and these were 1953-1954 model year vehicles.  Therefore, the best emis-
sion factor estimate for a  full-throttle acceleration from 0-60 mph is that
approximately 10 times  the  lead burned is exhausted by the car.

                             Automotive Lead Projections
      Based on data on  automotive lead emissions under city, cruise speed, and
acceleration conditions, calculations of lead emission rates were made for seven
roadway configurations.  These  seven roadway configurations were chosen to approxi-
mate worst-case  traffic volumes for  the roadway  types considered.  However,  it  is
possible  that particular roadways may have higher traffic volumes, and hence
higher emission  rates than  those modeled here.
      Lead emissions from automobiles are  projected for  future years for  seven
roadway configurations  based on emissions  at different vehicle speeds, the lead
.content of gasoline, and an average  fleet  fuel economy.   Emission rates are  cal-
culated based on equation  (1):
        [Emission  Factor]  [Lead Content of Gas  (g/gal)]  x APT  _  Emi=,.ion  Rjtc    .  .
                     [Fuel  Economy  (miles/gal)]                  (g/road mile-day)

                                      136

-------
 I
 •          Using equation (1), emission rate estimates were made for the seven roadway con-
             figurations.  These results are presented in Table 6.  [Note: To calculate the
 I          emission rate in units of gram's/meter/second, equation (1) should be corrected
             by dividing by 139,017,600.]

 •                                             Fuel Economy
                   Fuel economy projections are based on the average fuel economy standards
 I          (15 USC 2002).  In this law, fuel economy standards are established for 1978,
             1979, 1980 and 1985.  Past year average fuel economies are known through 1973
 |          (EPA 1974).  Fuel economy values for other years are determined using a straight-
 M          line interpolation.  Average fuel economies, based on the fuel economy standards,
             are shown in Table 4 for the projection years.
 V                The fuel economy calculation at each roadway configuration is performed
             to account for the effect of vehicle speed on fuel economy.  The fuel economy
 |          of a car is maximized when it is cruising at 30-40 miles/hour.  In addition, the
 «          fuel economy during highway driving is better than under stop-and-go city driving.
             Therefore, an accurate estimate of fuel economy on a roadway configuration must
 I          take traffic flow characteristics into account.
                   Table 1 shows fuel economies for different model year vehicles.  These
 |          fuel economies are based on a combined city/highway driving cycle.  Combined
 ^          fuel economy is a weighted average of the city and highway estimates based on
 •           Federal Highway Administration studies of average U. S. driving patterns.  This
 M           value (which assumes 55% city and 45% highway driving) is what the average
             driver can expect in overall summer driving on level roads after the car has
 |          .been broken in.  City fuel economy reflects trips for local errands, driving to
 _           work, and general stop-and-go driving  in urban and suburban areas.  Highway fuel
•           economy reflects long-distance driving on non-urban roads and on interstate
•           highways at a speed averaging approximately 50 miles/hour with no stops.
                                                 137
m

-------
      While the values in Table 1  are accurate estimates of nationwide fuel
economy, they are not representative of the fuel economy on specific roadways.
The coefficients in Table 2 can be .used to determine a correction factor to
account for vehicle speed at the study location.  The coefficients in this table
are normalized to 32.7 miles/hour, the average speed of the combined fuel economy
cycle.  These numbers were derived from the fuel economy correction factors pre-
sented 1n an EPA Interim report (EPA 1977).  These correction factors are
based on the Federal Test Procedure with an average speed of 19.6 miles/hour.
The normalized correction factors were calculated by determining the FTP
correction factor at 32.7 miles/hour for each model year, then dividing the A
coefficients for each model year evaluated at 19.6 miles/hour by the correction
factor for each model year for 32.7 miles/hour.  This normalization makes
the Table 2 coefficients compatible with combined cycle fuel economies.
      Fuel economy calculations for the projection years were performed following
these steps:
      1.  Knowing the average vehicle speed for the roadway configuration, the
coefficients from Table 2 and the vehicle speed are used in equation  (2) below
to calculate a speed dependent fuel economy correction  factor for each model year.
                                                     p    o    n
          Fuel Economy Correction Factor = A0+A1S+A2S +A3S +Ai|S        (2)
               where Ai = correction factors  from Table 2
                      S = vehicle speed (miles/hour)
      2.  Multiply the fuel economy speed correction factors for each model year
calculated in  (1) by the City/Highway Combined  Fuel Economy  (Table  1) and  the
fraction of annual travel by model year (Table  3).  The products for  each  model
year  are summed to determine the  base year  (1974) composite  fuel economy.  This
calendar year  composite fuel economy represents the combined cycle  fuel  economy
at average speed S.
                                     138

-------
 I
                  3.  The base year composite fuel economy calculated in step (2) is divided by
 •          the base year (1974) average fleet fuel  economy from Table 4 to determine the
 •          ratio of the fuel economy at the study location to the nationwide average fleet
            fuel economy.  This ratio is then used, to estimate the fuel  economy at the study
 I          location for future years by multiplying it by the average fleet fuel economy
            in Table 4 for each projection year.
 •                4.  The fuel economies calculated in (3) above are representative of
 •          combined city/highway driving.  These numbers should be corrected for a study
            location with traffic that is primarily free-flow or city-type driving rather
 •          than combined city/highway.  To correct for free-flow traffic, multiply each
            fuel economy by 1.2297.  The correction factor for city-type driving is 0.866.
 I          (Austin et al.. 1975).

                                          Lead Content of Gas
 •                The average lead content of gas for future years is calculated based on
            the promulgated lead phase-down schedule  '(Federal  Register. 1976).   The
 I          "pooled average lead content" of gas is based on the percentage of cars using
 •          unleaded gas and the lead content of leaded and unleaded gas.  For the calcula-
            tions in Table 5, the number of Pre-75 and Post-74 vehicles was determined using
 •          the fraction of annual light-duty vehicle travel by model year schedule
            in AP-42, Supplement 5.  The "pooled average lead content" is as given
 I          in the  lead phase-down regulations and the lead content of unleaded gas is assumed
 •          to be 0.05'grams/gallon.  From these numbers, the maximum possible lead content
            of leaded gas can be calculated for each calendar year using equation (3):
 •               [% Pre-75 vehicles] [Lead content of leaded gas (g/gal)]
              + [X  Post-74 vehicles] [Lead content of unleaded gas (g/gal)] •     (3)
 •               [100%] [Pooled average lead content  (g/gal)]

i

-------
Assuming that the lead content of leaded gasoline will  not  exceed  2.0 grams/gallon,
the probable pooled average lead content of gas was calculated.  The lead  content
of gas calculations are summarized in Table 5.

                     Automotive Lead Particle Size Distributions
      The primary work .on particle size distributions has been performed by
Hirschler and Gilbert, 1964 and Habibi, 1973.  Hirschler and Gilbert show that the
weight percentage of exhausted lead greater than five microns increases as vehicle
speed increases.  Their testing was performed on three cars equipped with V-8
engines and manual transmissions.  The cars each had 12,500 accumulated miles.
      The accuracy of Hirschler and Gilbert's data has been questioned in view of
the collection technique—i.e., total collection and subsequent dispersion and
fractionaticn of the particles.  Such operations can lead to agglomeration of
small particles leading to inaccurate size distribution data.  Because of possi-
ble inaccuracies in Hirschler and Gilbert's particle size data for different
vehicle speeds, it is  inappropriate to use these values in modeling roadway con-
figurations.
      Habibi's data on particle  size distributions seems to be the best available.
Habibi measured lead emissions from a 1966 model vehicle equipped with a 327-CID
engine operating on gasoline containing three grams of lead per gallon.  The
vehicle was  driven on  a programmed chassis dynamometer using the Federal mileage
accumulation  schedule  (Federal Register, 1968).  The lead particle size
distribution  data  is presented as a function of  accumulated mileage  in Figure 2.
This  seems  to be the most  accurate available data  on lead particle size distri-
bution emitted  from an automobile under city-type  driving.  While Figure 2 shows
how  particle size  varies as a  function  of  accumulated mileage,  it should also be
noted that  the  particle  size distribution  varies with the mode  of vehicle opera-
tion. The  percentage  of large  particles  increases with  the  severity of operation.
                                    140

-------
I
I
Summary
I                  Table 6 summarizes the emission rates for seven roadway configurations.
              As shown in the table, automotive lead emission rates at a specific location
                                                 I'
              are dependent on vehicle speed, the lead content of gasoline, fuel economy,
•j            and traffic volume.  The relationship of these variables is summarized in
              equation (1).

•                   [Emission Factor] [Lead Content of Gas (g/gal)] x ADT
                     	= Emission Rate  (1)
•                                   Fuel Economy (miles/gal)              (g/road mile-day)

flj                  While total lead balances show that 70-80 percent of the lead burned in
              gasoline is emitted into the atmosphere, this number is not an accurate esti-
|            mate of automotive emissions from a particular roadway.  Estimates have been
—            made for lead emission rates during city-type driving, cruise-speed driving,
™            and acceleration periods so that specific roadways can be analyzed.  Knowing
•            the average speed, volume and flow characteristics of the traffic on a roadway,
              an estimate of the automotive lead emission rate can be made using the techniques
I            described in this paper.

I

I

I

I

I

I
•                                            141

-------
                                REFERENCES


1.    Act of December 22,  1975,  89 Stat.  902;  15  USC  2002.

2.    Austin, T.C.,  R.B.  Michael, and G.R. Service, Passenger  Car  Fuel  Economy
       Trends Through 1976,  Automobile Engineering Meeting, Detroit, Michigan,
       October 13-17, 1975.

3.    Bradow, R.L.   Memorandum on "Lead Emission  Factors",  U.  S. Environmental
       Protection  Agency, Research Triangle Park, North  Carolina, June 22,  1976.

4.    Federal Register, "Control  of Air Pollution from New  Motor Vehicle Engines",
       Vol. 33 ,  No. 2,  Part II, Department of Health, Education  and Welfare,
       January, 1968.

5.    Federal Register, "Control  of Air Pollution from New  Motor Vehicles and New
       Motcr Vehicle Engines",  Vol. 37,  No. 221, Part II,  Environmental
       Protection  Agency, November 15, 1972.

6.    Federal Register, "Control  of Lead  Additives  in Gasoline", Vol. 41, No. 189,
       Environmental Protection Agency,  September 28, 1976.

7.    Ganley, J.T.  and George S.  Springer.   Physical  and  Chemical  Characteristics
       of Particulates in Spark Ignition Engine  Exhaust, Environmental Science
       and Technology, Volume 8, Number 4,  April,  1974.

8.    Habibi, K.  Characterization of Particulate Lead in Vehicle  Exhaust—
       Experimental Techniques, Environmental Science and  Technology,  Volume 4,
       Number 3,  March,  1970.

9.    Habibi, K.  Characterization of Particulate Matter  in Vehicle Exhaust,
       Environmental Science and Technology, Volume  7, Number 3,  March, 1973.

10.  Hirschler, D.A., L.G. Gilbert, F.W. Lamb, and  L.M.  Niebylski.   Particulate
       Lead Compounds in Automobile Exhaust Gas, Industr.  Eng. Chem.  49, 1957.

11.  Hirschler, D.A. and L.F. Gilbert.  Nature of Lead in  Automobile Exhaust Gas,
       Archives of Environmental Health, Volume  8,  February,  1964.

12.  Huntzicker,  J.J., S.K.  Friedlander, and C.I.  Davidson.   Material  Balance
       for Automobile-Emitted Lead in Los Angeles Basin, Environmental Science
       and Technology, Volume 9, Number 5,  May,  1975.

13.  Ter Haar, G.L., D.L. Lenane, J.N. Hu,  and M.  Brandt.   Composition, Size and
       Control of Automotive Exhaust Particulates,  Journal of the Air Pollution
       Control Association,  Volume 22, Number 1, January,  1972.

14.  U. S. Environmental Protection Agency, A Report on  Automotive Fuel Economy,
       Washington, D.C., February, 1974.

15.  U. S. Environmental Protection Agency, Compilation  of Air Pollutant Emission
       Factors (AP-42), Second  Edition, Research Triangle Park,  North Carolina,
       February, 1976.
                                    142

-------
I
I
            16.  U. S. Environmental Protection Agency, Factors Affecting Automotive Fuel
•                 Economy, Office of A1r and Waste Management, Washington, D.C., May, 1976.
            17.  U. S. Environmental Protection Agency, Mobile Source Emission Factors--
_                 Interim Document, Office of Transportation and Land Use Policy, Washington,
•                 D.C., June 1977.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

-------
                Table   1

   City/Highway Combined  Fuel  Economy
             (miles/gallon)

Model Year                    Fuel  Economy
   1974                           15.15
   1973                           14.89
   1972                           15.20
   1971                           15.24
   1970                           15.42
   1969                           15.47
   1968                           15.60
   1967                           16.15
                    144

-------

































CM

d)

JD
fO
I—



































J-
> -C
.0 \
 E
0
CO
1
UJ
CM
LO
^f
CM
«vf-
•
CM
1
ro r^
1 1 •
u» •
CM
c co
o
•r- O
•4-> 4->
o
CU "O •-1
s- cu <£
i- rsi
o •»-
O r—
ro
CM
1
LU
00
LO
CM
LO
LO
•
r>.
>» E

o o
c z
o •• —
(J
LU O
<

cu
ll

CM
i
LU
rv.
CT>
(O
*t
CM





s-
rO
cu
>-
r— •
CU
TJ
0
z:


-a
cu
r—
'o
i.
•M
C
O
u
1
cu
i-
o.
" r
LU
co
*d-
co
CM

CM
1
to

LU
VQ
CM
^1-
CTi

•
co


co

LU
CTl
LO
co

o
•
CM
1



CM

UJ
<•
LO
CTi
O
CTl

VO


CM

UJ
r"—
«d-
CTl
LO
r-..
•—







00
V£>
CTl






l"
LU
LO
00
<*
vo
CO
CM
1
LO
1
LU
LO
i—
O
*d-
LO
•
co


co
i
UJ
vo
r-~
CTi
00

•
CM
1



CM
1
LU
vo
LO
CTl
.
vo
CTI
•
r—
|



CM

LU
•*
C1
V
^a-
00

vo


CO

UJ
LO
LO
CM
VO
r^
CT>







t>^
CTI






~ |
LU
vo

co
^1-
«*
CM

IO
1
LU
«d-
CM
co
r^.
LO
•
CO


co

LU
CTl

CM
LO
r—
•
CM
1




CM
1
LU
CM
CO
00
O

1^.


CM
1
LU
LO
•st-
r^
r*.
IO
00*







CM
r*^
CTI






" i
LU
r^.
i>«.
CO
«d-
00
r—
1
to
1
LU
H—
^f
^1-
CTl
LO
•
CM


CO
1
UJ
*i-
r^.
00
CTI
LO
•
r™
1



CM

LU
CTl
LO
LO
VO
CTl

LO


CM
1
LU
03
00
CTi
CTl
CM
VO





<±
r-.
1
CO
r-
CTl
























^
t^
f^
CTl
r—

•a:
Q_
UJ
^-^*

•M
c
u
3
U
O
o

E
•r~
s~
cu
4->
c
H-l
1
1

U

O
ny
U_
C
O
•t—
-p
o
cu
t
^—
o
0
!>
o
c
o
u
UJ

r—
CD
3
u_



































-o
cu
cu
O.
in

CU
r-"
O
• r—
^:
cu

u

co

o>
i.
O)
_c
3:














145

-------
                            Table   3


              Fraction  of  Annual LKjht-Duty Vehicle

                     Travel  by Model  Year
              Age                       Fraction of
             Years                     Annual  Travel
Ref:  AP-42, Supplement 5
               1                           .112

               2                           .143

               3                           .130

               4                           .121

               5                           .108

               6                           .094

               7                           .079

              *8                           .213
                                      146

-------
I
I
I
Table  4
•                              Average  Fleet  Fuel  Economy
                                      (miles/gallon)

                      Calendar  Year                 Fuel  Economy
I                         1974                        12.4
                           1977                        13.3
I                         1978                        14.0
•                         1979                        14.8
                           1980                        15.7
•                         1981                         16.8
                           1983                        19.1
|                         1985                        21.7
•                         1990                        26.2
™                         1995                        27.4
I

•           Ref:    U.  S.  Environmental  Protection  Agency,
                    A Report on Automotive  Fuel  Economy, Washington, D.C., February, 1974.
I                  15 USC 2002

I

I

I

I

 I

-------
LO
 cu

!a
       cu
o
1/1
ra
CJ3

M-
o
        C
        cu
       •t-l
        C
        o
       o

       -o
        (C
        CD
 C— -
i —  ro
ro T3 C CT
.0 0) O -^
O r-* C_5 I/)
s- o E
CL O "O (13
Q. fd i-
O) cn
*
•X
> CU^
C cn
~O O —
CU C-5 (/)
i— E
O -0 ro
O ro S-
Q_ CU cn
	 1 > 	
i —
ro
uj cn
_pl -»— ^^
-^1, ^ ^^v.
i — i re i/)
-J a E
O — TO
CO £~
s sS
a
UJ
Q ^t
uj i cu
— 1 4-1 f—
^ I/O
^> C T-
D. -C
0)

CU ro
•— cn
.a -a ^.
(O H3 t/)
.a cu E
O — 1 fO
Q- cn
UJ
:z cu
_! J2 r-
O T- 10
c/i (/) cn
cC i/) ^-.
CD O "^3 (/
O- ro E
UJ • _J $-
Q x cn
^C f3 **— **
uj E:
un LO
r^. cu
1 r—
cu o
S- •!-
a. .c
cu
s_

a




*
•K
C3








LL.






LU














O







*
O3







ej-










orv-^ocouoLOLn
CMr-r— r— OOOO
1






f- ^" O CO LO LO LO
r— i— r— O O O O





LO LO LO LO LO LO LO LO
oooooooo





O CM LO LO IJD — cncncnr—
OOOr-~UD'*COCMC\l
1
CO 0 •—
r-^r-^r-^r-^p^r^-coco
cncncncncncncncn




^t" LO cn LO co r— cn co LO
COCMr— r—i— F— OOO
ooooooooo







LOLOLOLOLOLOLOLOLO
OOOOOOOOO





LOLOLOLOLOLOLOLOLO
OOOOOOOOO





o r^* cn co ^" ou cn co o
LocncM«3-ir>r^i--.coo
•~





OOOOOOOOO
CM C>J CM CM CM Cvj CM C\J CM






LO CM
O «*
co ^*






o co r™ CM cn co ^~ r^ o
LOOr-^LOCOCMCMr—
r— r—

CMco«*Lo^r^cocno
cocooocococooococn
oScncncncncncncnoi
                                                                                                                                   Q

                                                                                                                                    I
                                                                                                                                   03

                                                                                                                                   C

                                                                                                                                   3
                                                                                                                                   r-™
                                                                                                                                   O
                                                                                                                                   o
                                                                                                                                          O


                                                                                                                                          X.
                                                                                                                                          II

                                                                                                                                          03
                                                                                                                                   O
                                                                                                                                   o
                                                                                                                                                in
                                                                                                                                                c.
                                                                                                                                                o

                                                                                                                                                +J
                                                                                                                                                rt3
                                                                                                                                                cn
                                                                                                                                                QJ
                                                                                                                                               CC
                                                                                                                                         T3
                                                                                                                                         CU
 3
 CT
 CU
a:
•x
                                                                          148

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
0)




















CO
E:
o
3 ^T

•p-
*^~ r-
0 °
tj •p—
O 4j
g §•

"O
O r-
o: .JE
0 =

fO •*""*»
DC 0
C CO
O 1
•r- E

CO O1
LU

i— 13
(0 -C
O Ol
>
QJ ^— ^
CT)
(O O
S- •«-
> M-
^i fO
$-
h-
>,-—
E r-
O 'O
C CT
o -^
LU CL>
'ou E

U_


^—
r™~
rt)
CD
C CD
C «J
(O
3:
•o
03
O
Q^







LO ^O
| |
0 O
p— r—
X x

*fr 00
CTl O
r— CTl
V


o
O
O
*
O
LD








O *3-
O r—
CXJ C\J










0 0
CM r—





^ rv.
f>*. r*^.
O> CTi













)— —
'O
S-
3
a:







^J~ h«. >^ co
CTl O CO LD
VD *sj* OO CO

















10 CTI oo r—
CM oo LD r^
CM CM CM CM










CO LD LD LO
O O O O





CO CTl O r-
r~^ r>. co co
CTI CTI CTI CTI
-C
a.
E
0 1-
to ID O
CO 4J
aj u u
u  C CO
•i— (O -tO
E i— CD T-
•r- i > E
	 1 «^> ^ 1 1 1





t

00
ID
r-

















00
o
Pf\
\ t








LD
CM
0





CO
OO
CTl










^J"
ID
•


II



r*. VD i^ co
i ii i
o o o o
^M f^f, p«B f^^
X XX X

rsoo ^-r>-OCTtvooocor^
CMC»r— OCTi^-«3-
COCMCM r— -cJ-OOCMCMr— C0«*



O
O
O
ft
in









OCMCM CMCMi— r— CMLDCOLD
IDCM'3- LOyDr>.COCTiOOO e
QC CM 
-------































4-1
C
o
C_J
— '
to

cu

ra
1—





























cu
-P
.
CO CD
"i"""
UJ
>, !o
r— TD
rO -C
O Ol
>
0) • —
cn
ro O
S_ -i-
O) 4-
> 4-
to *^C f"0
C S-
0 h-
•1—
rO
s-
3 >~,-—~-
CD E i—
•i- O ro
M- C CD
C 0 ^
CD CJ (/)
o LU cu
>•> r— -r—
rO CU E

T3 U-
o
a:
s_ ^_^
0 i—
CD
(/) 4-) vSXi
CU C CD
ro -P
o: ecu
o c
Ct ^ •«-
S*J *l^
o •—
•r- -00
to ro to
to CU ro
O
S-
(O
CU





>^
ro
2
T3
ro
O
C£.







LT>
I
O
« —
X
cn co
CM IT)
CO i—



o
0
o
•»
00

r—











Tl <^-
3 CM
rM CM













o o
CM i—




•vt* r^

cn cn






•
>j
5
a.
X
UJ
ai
c c
ro ro
o ^™
S_ 1
=> vo

to r--.
i i
o o
i — i —
X. X
tOCOr— CO CO CM \O CO
i — cOLnoto«3-cnr-»
r- U3 tO IO CM r- CO' CO





















uncn«*ooi — in > — i —
CM CM CM CM CO CO ^" ^"












Lf) LT> Lf) ID
00 LO LO LO CM ^™ CO r""i
oooooooo




oo cn o r— co LO o LO
r**H r*^ co co co co cn cn
cncncncncncncncn
_c
a.
E
CO i-
^- o
•M
II 0
rO
^3 ' '
CU LO
CU C O
a. o «*•
00 •!-
to
• to
CD -r- II
> E
eC UJ

to r-. oo
i i i
0 O 0

XX X
cor^cn^Loocoi — ocn
CM cn cj~ oo co oo ^* oo o r*^
^•r— r— OOOOr^COi— LO**




o
o
o
A
00
CM











or^cocnt^.LO«d-LocMr—
OOr— r— CMCOLOI-^r— CM












LO LO LO LO
OOOOLOLOLOCMr— OO
CMr- OOOOOOOO




«=j-r-~cocnor— COLOOLO
r^ r^* r^ i^* oo co oo co cn cn
cncncncncncncncncncn

Q.
E
VO J-
•— o
-p
II U
-P (O
cu -00-
ai cu LO
5- cu c o
-P O. O r—
oo cu oo •!- •
c to
>> ro -to
-Pi— CD-i- II

O «tf- 
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I






























•£>
c
o
0


kO

O)
fwmm
JD
(O
i
r^


















































10
c
o
fO
S-
3
CD
•r-
4-
c
o
o
ro
5
•o
fO
o
o:
s_
o
4-

CJ
4-)
fO
o:

c

t/>
,w
E
UJ




















(D
-M
fO «*•—•»
o: u
0)
c to
O 1
to ^
1/1 CD
£=*"'
UJ
>> re"
•l— "^v,
03 -C
O O)
>
CL) s—- ^
CD
(O U
S- -1-
OJ 4-
> 4-
>-^~^*
E •—
O (O
C CD
0 ^-
O (/>
UJ 
O
S-
ro
Ol




>^
rO
2
T3
*t3
0
cc.






Ln 10 r~^
i i i
O 0 0
r™ * r— r—

XX X
OO O^ O*i O*^ *-Z? c^3 Oi ^3 ^3 ^J1
oo c^ oj co ^o oo oo ^3 r**^ *^D
CMr— OO^f^J-'d-r— r— CMCM

O
O
O
**
oc
o
r—











c «*-i^r^oooooooocncn
cncncncncncncncncncn
~ci

00 S_
co o
•M
II O
>> (O
S TJ U.
O- OJ
X rjj c O
UJ Q. O CO
OJ OO -r-
c c to
fO n3 • in
-Q i— CD -r- II
i- i > S
ID «3 
i — ^oLncocMCMi — vor— i —


o
o
o
"
00
CM











or~~.cocni--*.Q*d-LncMr—
OOr-r— CMCOinr^r—CM














in in in in
ooeomLnincMi— oo
CMr— OOOOOOOO





«3-r~~cocncDi — coLncDLn
i — . i — r^i-~.oococQcoencn
cncncncncncncncncncn

Q_
1— s-
u. o
-P
II U
•*-> fO
OJ -0 U_
QJ OJ LT)
S_ (U C VO
-t-> Q. O OO
OO QJ OO -r-
C l/>
>> (O "VI
4J r- CD'i- 11
•r- 1 > g
O 
-------































~~*
c
CJ>
10

0)
1 —
_Q
ra
1—































CU
+->
fO ' — **
a: o
cu
c to
0 I
*i"~ CS
to ->»
tO CD
UJ
>> ra
^— ~x3
•i — *"*^.

>—-
•r- O CO
M- C CD
C~ Q — ^
O O to
O UJ } r— 'I—
^ -^ > 	 ^
O
a:
s_ 	 ..
O r—
{j 	 ff^
CD
(/> -|-3 — ^
cu c: CD
4-> CU 	
ra 4->
o; ecu
0 C
c O -r—
O^__
'
•r- -O O
to ro t/)
to CU ro
•r- _J CO
E
1 1 1 tf_
0
S-
ro
CU





ro
5
"O
ro
O
Or*







^J-
1
0
(-—

X
o
CO
00

o
o
CO


rO
S_
CU
! 	
CD
O
0
•^







00

*""

0
o
0

o
00

II

.
>s
3
-i-


r^
r—
CM



1^ —
O












o
,_





r-^
cn











cu
c:
ra
t-~
\
r—







un
CO
1
















1JD
CM
CM



CO
—












CO
o




CO
r-.
cn


c~
a.
E
O
to
1
o

II

-a
cu
a>
a.
oo

40
1
O
r—

X
VO O
CT> LT>
r^ rv.
















cn co
OO LT>
CM CM



cTt r^
r— CM












LT) UO
0 O




cn o
i-~ oo
cn cri



5-
O
O
ra
1 1
0
c
O CD
•r— i—
to
to
•r- II
E
UJ

ID

o


X
i — r— CO 1O
O O VO LO
r-. co i — ^j-
















r— 00 O CM
r^ o LO CM
CM CO CO -=f



IO »* IT) CM
co LO r^. i —











LO LT) LT)
UO CM i — O
O O O 0




i— 00 LO O
CO co co cn
cn cn cn cn
"o.
E

O i.
to o
II CJ
rO
-D U_
O)
CU C
>> QL 0
ra CU 00 -i-
3 C to
.C rO -to
CD i — CD •!-
•r^ 1 > P^
2C CO ec UJ







O
OO
"*
















CM
.^j.
^J-



r—
CM











tn
O
o




tn
cn
cn









«^-
LO
•


II



























^— ^
•
21
*T~
s 	 ^


CL
fO
—





































152

-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
90+
80--
70i-
                                   Figure  1

                      Percentage of Burned Lead Exhausted

                           vs. Vehicle Cruise Speed
                Y Hirschler and Gilbert,  1957

                D Habibi ,  1970

                O Ter Haar, 1972

                f Sampson  and Springer, 1973
                  Ganley and Springer, 1974

                ® Bradow,  1976
                          20          30
                             VEHICLE CRUISE SPEED
                                                  (MILES/HOUR)50
                                         153

-------
9.
8.
7.

6.

5.

4.


3. --



2. --
  .3--
  .2--
                                       10
                                    20   30  40  50  60  70  80
                                                                             90
Lead Particle Size Distribution

EmfET6(i\from *\n Automobile Under

City-Type Driving, 1968
I          Figure  2

Particle  Size u vs. % Less Than
   Stated  Particle Size by
   Accumulated Mileage (Thousands
   of miles)
                                                                     Ref:  Habibi , Kamran.  "Characterization
                                                                           of Particulate Matter 1n Vehicle
                                                                           Exhaust."  March 1973.
                                                                              A  =  28,000 Accumulated Miles

                                                                              B  =  21,000 Accumulated Miles

                                                                              C  =  16,000 Accumulated Miles

                                                                              D  =   5,000 Accumulated Miles
                                          % Less Than Stated Particle Size
                                                       154

-------
 I
 I
 I
 I
 I
 I
 •                                       APPENDIX D
 •                          DEPOSITION OF PARTICLES AND GASES*
 I
 I
 I
 I
 I
               I*Excerpted from Slade, David H. (ed).  Meteorology and Atomic Energy
                1968.  Prepared by Air Resources Laboratories, Research Laboratories,
                Environmental Science Services Administration, U.S. Department of
                Commerce, for the Division of Reactor Development and Technology,
 I              U.S. Atomic Energy Commission, Oak Ridge, Tennessee.  July, 1968.
I
I
I
I
                Pp. 202-208.
155

-------

                        METEOROLOGY AND ATOMIC ENERGY— 1968
                                         §5-3
5-3  DEPOSITION OF PARTICLES

     AND GASES (Isaac Van  der Hoven)


List of Symbols

  Symbols used frequently in Sec. 5-3 on the de-
position of particles and gases are listed here.
(The dimensions  mass, length, time, and tem-
perature are  abbreviated as M, L, T,andD, re-
spectively. The equation number indicates the
first appearance of the symbol.)

CD       Drag coefficient (dimensionless),Eq.
             5.35
Cv, C2     Button's virtual diffusion coefficients
            (Ln''~), Eq.  5.38
g         Gravitational acceleration (LT~2),Eq.
            5.35
h         Height  of source above ground (L).
            Eq. 5.38
n         Sutton's parameter  associated with
            stability (dimensionless),  Eq. 5.38
Q', Q5    Initial  source strength (MT"1), Eq.
            5.38
Q'x       Depleted  source  strength at  a dis-
            tance, x, from the source (MT"1),
            Eq. 5.42
 r         Particle radius (L), Eq. 5.35
u         Average value of the wind component
             in the direction of the  mean hori-
             zontal vector wind (LT"1), Eq. 5.38
 vd       Deposition velocity  (LT~'), Eq.  5.41
 vg       Fall velocity (LT"1), Eq. 5.35
 A         Mean free path of air  molecules (L),
             Eq. 5.37
                                    157
M         Atmospheric  dynamic  viscosity
            (ML-'T"1), Eq. 5.36
p         Particle density (ML~3), Eq. 5.35
p.        Atmospheric density (MIT3), Eq. 5.35
OY, a,     Standard deviation of the distribution
            of material in a plume in the y and
            z directions (L), Eq. 5.44
 X         Average concentration (ML~3),  Eq.
            5.38
 u        Amount of  aerosol  removed per unit
            time  per  unit area (ML"2!1"1), Eq.
            5.39

 5-3.1  Gravitational  Settling

   The earth's  gravitational field plays  an im-
 portant  role  in  the  deposition  of  particulate
 matter  on  the earth's surface. The  rate  of
 descent  of the particle depends  upon a balance
 between the aerodynamic drag  force and the
 gravitational force exerted by the earth. For a
 smooth spherical particle, neglecting the effect
 of slip flow, this balance may be  expressed as
         ,      8
     Pa v< CD = ^ r g P
                                        (5.35)
 where  the  notation is as  given in the list of
 symbols. Equation 5.35 cannot be solved directly
 for the fall velocity because the drag coefficient
 is an empirical function of the Reynolds number,
 Re,  and  therefore also of  velocity. McDonald
 (1960)  has conveniently plotted  this empirical
 relation for Re > 1.0  from which values of fall
 velocity vs. particle size can be  computed.  For
 the Reynolds number range between 10~4andlO.
 the relation CD  ~ 24 /Re may be used, and, since
 Re = 2pjvgr/^,  Eq. 5.35 reduces  to the familiar
 Stokes equation
                    _ 2 r  g p
                    ~
                                        (5.36)
   The effect  of the slip  flow upon  the fall ve-
 locity is a function of the  ratio of the mean free
 path of the air molecules  to the particle size. It
 can be expressed by multiplying the fall velocity
 by a slip correction factor (Davis, 1945)
1 i A  1.26 i  0.4 exp ^ 1>lr
                                        (5.37)
 where X is primarily a function of altitude.

-------
 §5-3.1
                    PROCESSES AFFECTING EFFLUENT CONCENTRATIONS
  The effect of shape upon fall velocities is, on
the average, to reduce the velocity by about ? :i
from  that of a smooth sphere. Smooth ellipsoidal
particles theoretically  will vary in fall veloci-
ties by factors ranging from 0.5 to 1.04.
  At  fall velocities less than about 1 cm/sec,
the effect of  sedimentation  is negligible, and
vertical  movement of  the particle is largely
controlled by the  larger vertical turbulent and
mean air motions. Figure 5.4 (after Hage, 1964)
shows the fall velocity of smooth spheres with a
density of 5  g/cm3  as a function of the altitude
and particle  diameter,  with inertia terms and
slip flow corrections taken  into account where
significant. It  can be seen that the predominant
  30
  25
  20
    10-'
            10"
                    10'       10'
                FALL VELOCITY (cm/nc)
Fig.  5.4—Fall  velocity  of  smooth  spheres with a
density of 5 g/cm3 as a function of altitude and parti-
cle diameter (microns). (From Hage, 1964.)
factor affecting fall velocity is the particle size.
Similar fall-velocity  computations foraparticle
with  a  density of  2.5  g/cm3  are given  by
McDonald (1960a).
  In the range where the sedimentation rate is
significant,  the vertical transport of an initially
airborne  particle  (fall  velocity greater than
about 1 cm/sec) depends upon horizontal as well
as vertical  transport and diffusion.  For fall ve-
locities ranging from about 1 to 100 cm/sec, the
diffusion of a  cloud of  particles under homoge-
neous horizontal transport  (no wind shear with
height) can be  described by assuming that the
particles are diffused according to a statistical
diffusion model, such  as that of Sutton (1953),
and  at the  same time  that they will settle with
appropriate fall velocities.  For the case of an
elevated plume, the  effect is essentially that of
the downward tilt of the plume center line, which
can  be  expressed  by replacing  the  constant
height of the  plume  center  line in the Sutton
equation by a variable expression such that
X (x,y,0) =   **'  ... exp J-x»-» IX
                           (h-xvs/u)2
                               cT
(5.38)
With  the  assumption that the particles are re-
moved  (deposited) when they reach the ground-
air interface,  the deposition pattern can be de-
scribed by the expression
                                                                    ^ vsX (x,y,0)
                                         (5.39)
where w is the amount removed per unit time
per unit area and \ is the volumetric concen-
tration pattern of the air at the surface.
  Van der Hoven (1963)  used this tilted plume
model to describe the observed deposition pat-
tern  of radioactive effluents  and  included  a
cloud depletion factor (Csanady, 1955) of
                                                             1 -
                                                                (1-n 2)(h0u/xvg-l) * 2
                                         (5.40)
where n is one of Button's diffusion parameters.
In practice the tilted plume model is only appli-
cable  in a well-mixed atmospheric layer, such
as  is  typical  of daytime adiabatic conditions
within the lowest thousand meters.
  The effect of horizontal wind-direction shear
in the  vertical becomes important as a diffusion
mechanism if  there is an initial distribution of
particle sizes with height. With an initial cloud
dimension of about  1000  m and particle fall ve-
locities  greater than  1  m/sec,  the  effect of
turbulent  diffusion  on  the  ground deposition
pattern  can be neglected. The problem then be-
comes that of calculating particle trajectories
using  the appropriate fall velocity of each parti-
cle and the  resultant wind  vector of the atmo-
spheric layer  through which  the particle falls.
This  technique as  described  by Kellogg, Rapp,
and Greenfield (1957) has been applied primarily
to  the particle  cloud resulting from surface
nuclear-device detonations to calculate the fall-
out pattern for the  first few hours after detona-
tion.
                                          158

-------
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
                         METEOROLOGY AND ATOMIC ENEKGY —1968
                                                                  §5-3.2
5-3.2  Dry  Deposition

  5-3.2.1  Deposition IWon'/v.  The  observed  fart
that the  deposition rate of small particles onto
the ground can be greater than can be explained
by the appropriate gravitational fall velocity has
focused  attention on  nongravitational  and non-
precipitation mechanisms, such  as surface irn-
paction,  electrostatic attraction,  adsorption, and
chemical interaction. In analyzing the deposition
of  spores,  Gregory  (1945)  concluded  that the
deposition  rate was proportional to the immedi-
ate ground-level air concentration. Chamberlain
(1953) defined the ratio of the deposition rate to
the immediate ground-level air concentration as
the deposition velocity, which, analogous to Eq.
5.39, can be stated as
                        (x,y,0)
                 (5.41)
The interesting feature of such a formulation is
that by using it as an experimental tool to com-
pute vd through the field or laboratory measure-
ment of w and \, we  can apply  it to gases and
vapors as well as to small particles. It in no
way  explains  the  physics  of  the  deposition
mechanism, but  nevertheless it is a convenient
way to express  the  whole  complex  and  little-
understood dry-deposition phenomenon.
  5-3.2.2  Cloud Depletion.   To account  for the de-
pletion of an airborne cloud because of dry de-
position,  Chamberlain (1953) modified Button's
equation so that the original source term.  QJ,
was replaced  by an effective depleted-source
term,  Q'. Thus, for a  continuous  source  at
ground level, the depletion factor was expressed
by  Chamberlain as
             Qo'
               7= exp
                         4vdx
                              n/2
                  (5.42)
Using Eq. 5.41 and the modified Sutton equation,
we can express the deposition rate per unit time
per unit area for a source at ground level as
     -u CvC,x
             2-n
nu
                      x exp I-
                                         (5.43)
 Chamberlain  further  expressed  the depletion
 factor for the case of an elevated source.  Cul-
                                        159
                           kowski (1958) has presented graphical solutions
                           to these equations.
                             The  generalized Gaussian diffusion formula
                           (in  the  notation  used in Chap. 3) can also be
                           modified for cloud depletion. The depletion cor-
                           rection  for a continuous elevated source can be
                           derived as follows:
                            !(x,y) r: vd x(x,y,0)
                                                 y'   A2
                                                                   (5.44)
                           where w(x,y) is the  surface deposition at (x,y)
                           and Q!, is the residual source at x meters down-
                           wind. The  depletion of the source per unit dis-
                           tance is given by
     -/'
       »/-co
                                                             w(x,y) dy
                                   W   ua7  6XP - \2o?
                           which can be rearranged as
                                                                   (5.45)
 PCIQ;_   /2\
Jo  Q'     W
                           If QJ = Q5 at x = 0, then
                v,  r __ _dx
                u Jo °> exp  z
                                                        (hz/2a|
, Q; /2V1
lnQr-U)
and therefore
?1
" _ dx
CT; exp (h2/2cr?)
                           Q;
                           OS
     exp
 r	d
Jo  CT'  exP
                  dx      1
                                                      -(2/n) *.
                                  d/u
                                                                   (5.46)
                                                                   (5.47)
                                        (5.48)
  Since a, is  not  generally  available  as  an
analytical  function  of  x in  the generalized
Gaussian  form, the  integral expression in Eq.
5.48 was  evaluated  numerically using the ex-
perimentally  derived  values of az(x) given by
Hilsmeier and Gifford (1962). Figure 5.5 shows
the depletion  fraction, (Q'/Qj),  as a function of
distance from the source, diffusion category (in
terms  of  Pasquill types or  the corresponding
values of  afl  as given  in Chap.  4, Sec. 4-4.4),
height  of release, a deposition velocity of 1cm/
sec, and a mean wind speed of 1 m/sec. To ob-
tain depletion  fractions for other  deposition
velocities and wind speeds, we may use the ex-
pression

-------
§5-3.2
   PROCESSES AFFECTING EFFLUENT CONCENTRATIONS
    UJ -
    0. _
    -I -
   82°  I
    111
       n E
       O —
                  2    -
    i E
    o —
      in
      6
o  «">
—  o
                 bto
                                     160

-------
  I
  I
  I
  I
  I
  I
  I
  I
 I
  I
 I
 I
 I
 I
 I
 I
I
I
I
                        METEOROLOGY AND ATOMIC ENERGY— 1968
                                                                                        S5-3.2
                            "2vtll
                                                 vapor,  131I
                                       (5.40
where  the  subscript 1 refers to values found in
Fig. 5.5 and the  subscript  2  refers  to the de-
sired values. Thus, for example, to find the de-
pletion  fraction  at  a distance of  104 m for a
source  50  m high, a u2 of 1 m/sec, a v(P of 0.1
cm/sec, and a type F diffusion category, first
find, in  Fig. 5.5, the  value of  (QyQ'oh  for h =
50  m, x •  104 m, u =  1 m/sec,  and vd - 1 cm/
sec: (Q{/Qo)i = 0.50.  Now substitute  this value
in Eq.  5.49,
             -~f  - (0.50)0>1 = 0.93
  The Chamberlain and the generalized Gauss-
ian depletion  models assume that  the shape of
the concentration profile in the vertical is un-
altered by deposition. In an approach suggested
by  Calder (1961), K theory diffusion equations
(see  Chap. 3, Sec. 3-2.1.2) were  used to de-
scribe the effect  of deposition on  a plume. In
models of this type, the reduction of concentra-
tion  caused  by  removing  material from the
cloud is not distributed evenly through the depth
of the cloud but depends upon the profile of ver-
tical mixing. Therefore the shape of the vertical
profile of concentration will change as the com-
putation proceeds. Smith (1962a) schematically
illustrates the effect for a case  where h - 0,
vg - 0,  vt| > 0, and the exchange coefficient is
constant with height. The net  result is a more
rapid  depletion  of the  bottom portion of the
plume; so downwind from the source the height
of the  maximum concentration is above the sur-
face and  increases in the downwind direction.
Definitive  field measurements to  evaluate the
statistical depletion model and the K-theory de-
pletion model are not yet available. Computa-
tions  show that the statistical models (Eqs. 5.42
and 5.48) give higher depletion factors than the
K-theory  models under stable conditions.
  .i-.'J.2..'J  /)c/)(i.
-------
55-3.2
PROCESSES AFFECTING FrFLUENT CONCENTRATIONS
             Table 5.9a—SUMMARY OF nil DFI'nM, ION FIELD F.XPI-.R1MF.NT RESULTS'
Ilaruc'l 'i

Deposition velocity,
cm/sec
Grass
Soil
Snow
Carbon
Clover lea\es
Paper leaves
Filler paper
Sticky paper
Wind speed, in / sec
Friction velocity,
cm/sec
Roughness length,
cm
Downwind distance, in
Grass cover, g/nr
Stability
1 2


1 9 2.6





0.0 0.7

5.2 4.:i

48 .'!5

2.x 1.5
15 20
500 200
Lapse Neutral
:t


1.8



l.U
2 0
0.9

5.2

48

1.2
20
200
Lapse
I


8.7



0.9
1.5
0.0

4.1

.'(8

2. 1
20
420
Lapse
-Is
5


1.7



0.5
1.0
0.1!

1.6

15

2. 1
20
120
Neutral
Idaho tests
0 7


l.f 1.1



0.3
0.0


2.:! :i.9

20 20

5 II 1.0
100 lou
010 420
Lapse Neutral
I 11


0.6 1.0
0.8 0.4

0.6 0.7



0.2 0.1
7.1 9,,'i

01 69

.'! 1 1.5
'(Ml 1175
lo.'i 240
Lapse Lapse
Snow




0.2
0.9



0.6
6.0

50

2.1
:! 10
Snow co\ er
Stable
   *Chamberlain, 1960, and Hawley, Sill, Voelz, and Isliuer, 19G4.
                  Table S.'Jb —SUMMARY OF CONVA1R utl FIELD RELEASK TKSTS*

Deposition velocity.
cm/sec
Grass
Soil
Sticky paper
Water
Wind speed, in/sec
Downwind distance, m
Stability
E



0.5
1.1
1 .8 1"
(i>
1,600
Lapse
F



1.4
0.7
2.;>1
5..'i
1,000
Lapse
H



0.5
1.5
1.41
4.0
1,600
Stable
1


2. It

0.4

5.0
.'S2.000
Stable
2




(Mi

4.2
1.000
Stable
3




0.1

;).2
1,000
Stable
4




0.2

2 4
1 ,000
Stable
5




0.2

1 1
4.00(1
Stable
8




o.:i

2.7
i.ooo
Stable
10


1 .2t

0.2

2.6
4.000
Stable
11




0.6

4.4
16,000
Stable
    "Convair, 1959, 1960.
    TDownwind distance of 2000 m.
    tDownwind distance of 1000 m or less.
 zirconium,  cerium, niobium, and tellurium for
 which deposition-velocity calculations were
 made. Meteorological conditions included  both
 adiabatic and  stable  lapse  rates, and mea-
 surements  of ground  deposition and  air  con-
 centrations were  made to distances of 3200 m.
 Some  sticky-paper measurements were made
 out  to 3.2 < 104  m.  All  particles were  less
 than  10^ in diameter. Table 5.10  summarizes
 the  deposition-velocity calculations. Note  that
 these values are averages  and  that there is
 considerable scatter in the data which cannot be
 explained by the meteorological parameters that
 were  measured.  For example, the average de-
 position  velocity  of 0.2 cm/sec for l37Cs  on a
                              grass  surface was computed  from 21 values
                              ranging from 0.04 to 0.4 cm/sec.

                                     Table 5.10—SUMMARY OF CONVA1R
                                    RAD1ONUCL1DK FIKLI) RKJ.KASE TEST
                                          DISPOSITION VELOCITIES

                                              Deposition velocities, cm/sec
                                          Water
                                                   Soil
dr.tss  Stick\ paper
                              '•1:Cs       0.9(5)*  001(15)  02(21)   0.2(117)
                              lo'Ru       2.3 (9)   0.4 (16)   O.b (20)   0.4 (96)
                              95Zr, S5Nb   5.7 (6)   2.9 (6)             1.4 (10)
                              »'Ce                                 0.7
                              «2'Te, 129Te                            0.7 (8)

                                •Number  in parentheses indicates  the number of
                              determinations.
                                        162

-------
  I
  I
  I
  I
  I
  I
  I
 I
 1
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
                        METEOROLOGY AND ATOMIC ENERGY  - 1968                  S5-4.1
  Another  technique used  in  calculating de-
position  velocities is material-balance  me;i
surements  such as  performed by Islitzer aim
Dumbauld  (1963)  in their  fluorescent-particU
studies. The technique involves the determina-
tion of the mass flux of material through a ver-
tical  plane  perpendicular  to  the  mean wind
direction. The downwind decrease of this flux is
attributed to a material loss through deposition.
Using  uranin particles with a median diameter
of 1 p.  and  a fall velocity of less than 10~2 cm '
sec, Islitzer and Dumbauld computed deposition
velocities of 0.2,  2.4, and 7.1  cm/sec for  in-
version,  neutral, and lapse conditions, respec-
tively, for  the arid terrain in Idaho. From data
presented by Simpson (1961), these authors also
computed a value  of 0.5 cm 'sec for  four stable
cases  with  zinc  sulfide  submicron particles
over similar terrain at Hanford, Wash. These
and other deposition data are quoted in a sum-
mary article by Gifford and Pack (1962).
  Two  conclusions  seem apparent from  the
available field data  on the deposition of vapors
and submicron particles,  i.e.,  that  chemically
active  materials  such  as  131I  deposit  more
readily than inactive materials  such as 137Cs or
nonradioactive  fluorescent  particles  and that
vegetation surfaces such as grasses  and bushes
provide removal  rates that are greater than
bare surfaces. At present, however, what effects
atmospheric transport and  diffusion parameters
have upon  deposition or what effect  more com-
plex surfaces,  such as buildings and forests,
have upon deposition rates  is not clear.
                                               163

-------
 I
 I
 I
 I
 I
 I
 1                                        APPENDIX E
                            A SURFACE DEPLETION MODEL FOR DEPOSITION
 I                                  FROM A GAUSSIAN PLUME*
 I
 I
 I
 I
 I
 I
             I*Reprinted with permission from Atmospheric Environment, Vol. 11, No. 1,
               Horst, Thomas W., "A Surface Depletion Model for Deposition from a
               Gaussian Plume," 1977, Pergamon Press, Ltd.
 I
I
I
_                                              165

-------
 I
 I
 I
 1
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
 I
I
I
I
        Liuiionnu'iit Vul II, pp 41 46 P^rgamon Pu
                                            PrmlLd in Grx-.il HnUm
           A  SURFACE  DEPLETION  MODEL  FOR  DEPOSITION
                            FROM  A  GAUSSIAN  PLUME*|

                                           THOMAS W. HORSF
                  Atmospheric Sciences Department, Battelle, Pacific Northwest Laboratories,
                                       Richland, WA 99352,  USA

                         (First received 15 March 1976 and in final form 24 June 1976)

       Abstract- When the atmospheric diffusion  of a material is described by  the usual  Guassian plume
       model, dry  deposition  of the contaminant  onto the underlying surface is  commonly  accounted for
       by appropriately reducing the source strength, as originally proposed by Chamberlain. A more realistic
       model is developed which selectively  depletes the Gaussian plume in the  vicinity of the deposition
       surface rather than throughout the vertical extent of the plume as done in the source  depletion model
       This improved model is used to show  that the source  depletion model consistently  overpredicts the
       surface air  concentration and the deposition at downwind locations  close to the source and,  as  a
       consequence, is biased  in the opposite  sense for locations far from  the source. At all distances  from
       the source, the source depletion model overestimates the total deposition between source and receptor
       and consequently underpredicts the  amount of remaining airborne material. Quantitative comparisons
       are shown  to aid  the user  in choosing, for his particular circumstances, between  the  less  accurate
       source depletion model and the computationally more complex surface depletion model.
Dry deposition of an airborne material onto the un-
derlying surface is of interest from two different stand-
points. First  of all, it can  be an important  sink for
the material,  reducing air concentrations (and conse-
quent dry deposition) further  downwind. In  the case
of a noxious  substance this is  beneficial to downwind
receptors. However,  dry deposition  secondly  is  a
mechanism for accumulation  of the material on the
ground and  hence may  be detrimental at the point
of deposition. Since a high estimate  for the deposition
flux will  be  conservative at  the point  of deposition
and, at the same time, nonconservative downwind of
that point, modeling of the  dry deposition process
should be unbiased.  However, it will be shown here
that the  source  depletion  model currently used for
predicting deposition from the Gaussian plume is
biased. This will be done by developing a surface dep-
letion model which  eliminates  the  artificial bias of
the former model  and by comparing the predictions
of the two models. The biases of the source depletion
model will be quantitatively  delineated in  order to
provide a basis for choosing, in a  specific situation,
between the less  accurate source depletion model and
the computationally  more  complex surface depletion
model.

             THE DEPOSITION VELOCITY

  The deposition of airborne material  onto the un-
derlying surface  can be caused by  a combination of
mechanical processes (gravitational  settling, turbulent

  * This paper is based on work performed under  U S
Atomic Energy Commision Contract  No  AT(45-1)-1H30
  t This paper was presented at the Atmosphere-Surface
Exchange of  Particulate  and Gaseous Pollutants  1974
Symposium held in Richland, WA, 4-6 September.  1974
and  molecular diffusion,  inertial  impaction)  and  is
often further complicated by electrical and chemical
effects or the existence of heat and  moisture fluxes
normal to the deposition surface. In the absence of
detailed microphysical measurements, the  deposition
flux  Fd is  usually assumed to be directly proportional
to the local air concentration evaluated at a reference
height :d.
                Fd(x,y) = vdx(x,y,zd).              (I)

The  constant of proportionality vd has the dimensions
of a  velocity and has been  appropriately named the
deposition velocity. For common aerosols vd can vary
from about 10~4 to 10cms"1, depending on  specific
properties of the particles, of the atmospheric struc-
ture,  and of the deposition surface  (Sehmel, 1975).
Most of the results to  be presented here, however,
are parameterized only by the ratio of the  deposition
velocity to  the mean windspeed, vd/u, and this  will
be assumed to be independent of the horizontal coor-
dinates. Since u in the lowest 100m of the atmosphere
is commonly in the  range of 1 -10 m s~ ', vju can vary
from about 10~7-10-'.

            SOURCE DEPLETION MODEL
  The atmospheric  diffusion process  will be described
by the standard Gaussian plume model for a nonde-
positing  material,  incorporating "reflection"  of the
material at the ground to insure conservation of mass.
For  that  purpose  we  define  a diffusion function
     x 
-------
                                           THOMAS W. HORST
where  x(x, y,  z) is the  downwind  air concentration
due to a continuous source of constant strength QQ
located at the  point (0, 0, h).* The coordinates x, y, z
are oriented, respectively, in the direction of the mean
wind u, horizontal and normal to u, and vertical and
normal to u. The diffusion parameters ay and az are the
horizontal  and  vertical standard deviations  of the
assumed Gaussian plume.
  The source depletion model (Van  der Hoven, 1968),
then,  accounts for the loss of airborne material due
to deposition  by appropriately  reducing the source
strength as a function of downwind distance, i.e.
                    =  Q(x)-(x,y,z,h).
                           u
                                                (3)
Conservation of mass requires that:


      = -  f   »a(x, y, zd) dy = - -- Q(x) D"(x, zd,h),
           J-7                    U
 dx

where
                                                (4)
              2;: a.
                    exp


       + exP I —^IT-

Thus

              f    r«d
Q(x) = Q0exp<  -    -l)(t,zd,
              {   Jo  u
                                                (5)
                                                (6)
where Q0 is the undepleted source strength at x = 0.
Note that  although the source  depletion model cor-
rectly determines the deposition flux in terms of the
air concentration  near the surface  /(z = zd), it  also
instantaneously distributes the effect of the deposition
throughout the entire  vertical  extent of the plume.
Retaining the Gaussian shape of the plume thus artifi-
cially enhances the vertical diffusion. The  qualitative
and quantitative results  of this  effect will be  seen
below.
                                                     (i^,f),0) will effect a reduction of the air concentration
                                                     at the downwind point (x,y,z)  equal to:
      [ - vttit, n, zd) d£ di;] - (x -
                         u
                                                                                       >• - r,, z, 0) ,   (7)
where  the bracketed  quantity  is  an  areal  source
strength  due  to  deposition. The air concentration
at any point  can  then be calculated as the sum of
the nondepositing diffusion from the primary source
at (0, 0, h) plus the diffusion from all of the upwind
surface sources which account for deposition,

              D
X(.x, y, 2} = Qo — (x, y, z, h)
              u

     -  f   f  °-*-x(S,r,,zjD(x-t,y-t,,z,0)dtdt,.
       J_a, J0 M

                                                (8)

Given  then that  Q0D/u is the solution  to the con-
vective diffusion equation without deposition, equation
(8)  is  the exact  solution to the  same  differential
equation subject to the deposition velocity  boundary
condition,  equation   (1).  This  may  be verified  by
direct substitution into the diffusion equation.
  Equation (8) and  the subsequent calculations can
be considerably simplified if we consider  the  cross-
wind-integrated air concentration:
                                                                 D           f°°   f*'  f* v*
                                                        ,z) = eo-(X,2,/l)-             -X(U
                                                                 U           J-^J-^Jo U
                                                            x £>(x - £, y - r\, z, 0) d£ Ar\ dy.
                                                                                                     (9)
                                                     Performing, in succession, the integrations over y and
                                                     ^ yields:
                                                                 D
            D           C* V*
  x,z) = Q0-(x,z,h)~    -flfc
            u          Jo u

       x  D(x - £, z, 0) d£.
                                                                                       i)
                                                                                                    (10)
                                                     For  comparison, a similar equation may  be written
                                                     for the source depletion model:
            SURFACE DEPLETION MODEL

   A more realistic approach is provided by the sur-
 face depletion model.  In  the  development  of this
 model, advantage is taken of the fact that the linearity
 of the differential equation describing gradient diffu-
 sion allows  the  superposition  of  solutions such  as
 equation (2) to account for a number of sources  at
 different locations. The deposition  flux to  the surface
 is represented as a material  sink, i.e. a  source for
 downwind diffusion of  a material deficit from the
 point  of deposition. Thus deposition at  the  point
  * For a bouyant effluent, h is the effective source height.
                                                     X(x, z)
                                               (11)
                                                     This is equivalent to equations (3 and 6). Comparison
                                                     of the variables upon which D is functionally depen-
                                                     dent  in equations (10 and 11) again emphasizes the
                                                     artificial assumption of the  source depletion  model
                                                     that deposition is a loss  at the source (0,0,/j)  rather
                                                     than  at the surface where it is actually occurring. The
                                                     source depletion model, of course, is the easier  model
                                                     to apply since it retains the Gaussian distribution in
                                                     the vertical and thus all effects of deposition are sum-
                                                     marized in the single parameter  Q(x)/Q0. Further,
                                                     equations (6) or (11)  is  much more  economical  to
                                                     compute than equation (10) since the integrand of the
                                                 168

-------
 I
 I
 I
 I
 I
 I
 I
 I
 1
 I
 I
 I
 I
 I
 I
 I
I
I
I
                                     Deposition from a Gaussian plume
           i.o
           0.5
           0.2
       o  0.1
       i  1.0
	  SOURCE DEPLETION MODE
SURFACE DEPLETION MODEL
	Z-lm
	  SUSPENSION RATIO
                                                                    u - 10"2, h • 100 m
                                                                    PASQUILL A
           0.5
           0.2
           0.5
           0.2
           0.1
                                                            h • 10m
                                                        h -2m
                                                        I
                                  102
                                      103
                                 DISTANCE DOWNWIND, m
104
105
       Fig. 1. Comparison between source depletion and surface depletion models for vju = 10  2 and un-
                                        stable thermal stratification.
deposition integral is not  a function of the receptor
coordinate x.*

           COMPARISON OF THE MODELS

  Both models were applied to a variety of situations
to assess the accuracy of the source depletion model
relative to the surface depletion model and to provide
information  upon  which  to base a  decision  as to
which model to use in a given  situation.  Some of
the results of these calculations are shown in  Figs.
1-4 which display, as a function  of downwind dis-
tance from the primary source, the ratio between the
air concentrations with deposition and  without  depo-
sition  as  calculated  by both the source depletion
model and the surface depletion model. For the sur-
face depletion model three ratios are shown: that of
the air concentrations at a height of 1m, zd\ at the
source  height, h; and  vertically integrated  over  all
heights. The latter curve,  also called the  suspension
ratio, represents the  fraction of the material which

  * The surface  depletion  model  requires  3.84   CDC
CYBER 74-18 s to calculate  the three functions displayed
in the following figures for one combination of h, vju and
stability and  three decades of downwind distance. The
source depletion  model required only 0.17s  for the same
task and it could have been done even more economically.
                                     is still airborne or, subtracting from unity, the fraction
                                     deposited on the surface to that distance. As indicated
                                     in the model  derivation, all three of these ratios are
                                     predicted for  the source depletion model by the sus-
                                     pension ratio, Q(x)/Q0. Calculations were made for
                                     three source heights: 2, 10, and 100m, and for three
                                     atmospheric stability  categories:  unstable (Pasquill
                                     A), neutral (Pasquill D), and stable (Pasquill F), using
                                     the rural diffusion coefficients amalgamated by Briggs
                                     (1973). While these tr's,  shown in Table 1, rrfay  already
                                     include the effects of deposition, they are entirely ade-
                                     quate for comparing  the two deposition models. Note
                                     finally  that the  curves are parameterized with  the
                                     ratio va/u. It  can be shown from the equations pre-
                                     sented  above that the  ratio between the air concen-
                                     trations with  and without deposition is  a  function
                                     only of the ratio vju and  not vd and u independently.
                                       Figures 1-3 present  results for vd/u = 10"2, a case
                                     of moderately strong deposition, and for all combina-
                                     tions of thermal  stability  and  source  height.  As dis-
                                     cussed above, the assumptions of the source depletion
                                     model produce artificial vertical mixing of the  plume.
                                     Since the vertical mixing due to real atmospheric pro-
                                     cesses decreases  with  increasing stability, the effects
                                     of the artificial mixing would  be expected to become
                                     correspondingly  more noticeable as the thermal stabi-
                                     lity  increases.  It  is easily seen from  Figs. 1-3 that
                                                    169

-------
                             THOMAS W. HOKSI
10
     	 SOURCE DEPLETION MODE
     SURFACE DEPLETION MODEL
     	Z = lm
     	Z =h
     	SUSPENSION RATIO
                                                        yj u • 10 , h = 100 m
                                                          PA SOU ILL D
                                   DISTANCE DOWNWIND, m
       Fig 2. Comparison between source depletion and surface depletion models for
                                          thermal stratification
                                                                 = 10  2 and neutral
the differences  between the two models  do increase
with increasing stability. Further, it is expected that
the differences  would decrease as the deposition de-
creases since the  models are identical in  the  limit of
vd — 0. This is demonstrated by comparing Fig. 4 for
Vj/u = 10"3  and  three combinations  of  height and
stability with the  corresponding cases in Figs. 1-3 for
vju = 10"2. For the source depletion model it can
be shown (Van  der Hoven, 1968) that the  relationship
between  the suspension ratios for different values of
i'd/M is:

          [ CM/Co]!"'""" = [Q(x)/Qo]b'jM"-      ('2)
Figures 1-4 show that the functional dependence of
the surface depletion model on vju is equally strong.
   As  expected,  the surface  depletion model in general
predicts  smaller  air concentrations at zd than  does
the source depletion  model. This is  due  to the fact
that it realistically duplicates  the actual physical sit-
uation by  selectively  depleting  the  portion of the
plume adjacent to the surface rather  than the  entire
vertical extent of  the plume. Consequently, the surface
depletion model  also produced less deposition close
                                       to the source because  the material  deficit  near the
                                       surface insulates the bulk of the plume from the depo-
                                       sition surface. Thus, as shown by the suspension ratio,
                                       the surface depletion model always has a greater total
                                       quantity  of airborne  material.  At  large  distances
                                       downwind of the source this has the  effect of raising
                                       the near-surface air concentration of the surface dep-
                                       letion model until in some cases it eventually equals
                                       and surpasses that of the source depletion model. This
                                       occurs likewise for the  dry  deposition. These  effects
                                       are  best  seen in  Fig.  3,  which presents the  most
                                       extreme  surface model/source model differences.

                                       Table 1. Formulas for the determination  of a, from Briggs
                                                         (1973)(x  and atm)
Stability class a,
A
B
C
D
E
F
0.20x
0.1 2x
0.08x(l H
0.06x(l J
0.03x(l -
0.02x(l -


h2*10~*x
h 1.5*10-3
(- 3*10-* >
H3*10-*>


.j-l/2
x)"1/z
,\- 1
.\-l
                                    170

-------
  I
  I
  I
  I
  I
  I
  I
  I
 I
 I
  I
 I
 I
 I
 I
 I
I
I
I
                                  Deposition from a Gaussian plume
                       SOURCE DEPLETION MODEL
                SURFACE DEPLETION MODI

                	Z -1m

                	I -h

                	SUSPENSION RATIO
          01
          005
     2   0.5
     o   01
          005
          002
          10
                                                       103                   104                    105

                                                 DOWNWIND DISTANCE, m

Fig. 3. Comparison between source depletion and surface depletion models for vd/u = 10  2 and stable
                                        thermal stratification.
     i.o
     0.5
     0.2
P    0.1
     0.5
     0.2
     0.1
     1.0
     0.5
     0.2
     0.1
            	 SOURCE DEPLETION MODEL

            SURFACE DEPLETION MODEL

            	Z'lm
                   d/ U • 10 ,  h • 10 m

                     PASQUIU. D
                  • SUSPENSION RATIO
                                                              h -10m, PASQUILLF
                                                                   h -2m, PASQUILLF
                       j	I
j	I
                               102
        103
104
105
                                                 DISTANCE DOWNWIND, m
     Fig. 4. Comparison between source depletion and surface depletion models for vd/u = 10 ~3


                                                171

-------
                                           THOMAS W. HORST
  Figures 2 and 3 also show the ratio of the air con-
centration at source height as predicted by the surface
depletion model  to  that  predicted by the  Gaussian
plume model without deposition. This variable is not
shown in Figs.  1  and 4  due to the  tight  grouping
of the curves. The only consistent relationship to be
seen between this variable and the others  shown is
that  the  source  height  air  concentration  ratio  is
always greater than the reference height air concen-
tration ratio as also calculated by the surface deple-
tion model. This  is due to the vertical concentration
gradient  required to support  the deposition flux. In
most cases  it is also less  than the suspension ratio;
but  since circumstances  can always  be found, e.g.
large release height and strong stability, in which the
deposition  is not effectively  communicated to the
release height, it can also  be greater than the suspen-
sion ratio.

                  CONCLUSIONS

  Figure 3  graphically  displays  the biases of the
source depletion model:  an  overprediction of total
deposition and a consequent underprediction of the
remaining airborne material. For a case of moderately
strong deposition,  vd/u = 10 ~2,  and  stable thermal
stability  the source depletion model is in  error by
factors of  3-4 for some  parameters at a  downwind
distance  of 10km. These  factors  can become much
larger at greater distances and for stronger deposition.
Correspondingly, the differences between the models
are sharply attenuated by decreasing vt/u or decreas-
ing the thermal stability. Figure 4  shows the worst
cases calculated for vd/u = 10"3 from the set of three
release heights  and three stabilities. The errors of the
source depletion model  are only 10-20%. Similarly
for vt/u = 10"2 and Pasquill A stability, the worst
error is about 35%. Thus the source depletion model
can be entirely  adequate for a situation involving low
deposition  and grossly in error for a  case of high
deposition. The user must determine where the divid-
ing line lies between these two extremes by  carefully
considering his own situation and requirements.

Acknowledgement—This research benefited from  discus-
sions with C. E.  Elderkin who originally proposed the idea
of accounting for deposition by diffusion  of a material
deficit.

                    REFERENCES
Briggs G. A. (1973) Diffusion  estimation for  small emis-
  sions. ATDL  Contribution  79  (Draft), Air Resources
  Atmospheric Turbulence and Diffusion Laboratory, Oak
  Ridge, TN.
Sehmel G. A. (1975) Particle dry deposition  velocities. Pro-
   ceedings Atmosphere-Surface Exchange of Paniculate and
   Caseous Pollutants—1974 Symposium (Edited  by  Engel-
   mann  R. I. and Sehmel G.  A.), CONF-740921, AEC
   Symposium Series,  Oak Ridge, TN.
Van der Hoven  I. (1968) Deposition of particles and gases.
   Meteorology and Atomic Energy, 1968 (Edited by Slade
   D ), pp. 202-208, USAEC, TID-24190.
                                              172

-------
I
I
I
I
I
I
•                                         APPENDIX F
I                     CALCULATION OF CRITICAL AMBIENT LEAD CONCENTRATION
-                       BELOW WHICH THE NAAQS WILL BE ATTAINED BY 1982
™                          DUE TO MOBILE SOURCES IN URBANIZED AREAS
I
I
I
I
I
I
I
I
I
•                                              173

-------
I

I
                                       APPENDIX F
|                 CALCULATION OF CRITICAL AMBIENT LEAD CONCENTRATION
-                   BELOW WHICH THE NAAQS WILL BE ATTAINED BY 1982
•                       DUE TO MOBILE SOURCES IN URBANIZED AREAS
         •ASSUMPTIONS
                                              3
              1.  A NAAQS for lead of 1.5 jjg/m , maximum quarterly mean to be
|       attained by 1982.
              2.  No change in urbanized areas of traffic (VMT), average vehicle
•       speeds, or stationary source emissions between 1976 and 1982.
•            3.  No change in the level of control on stationary sources in
         urbanized areas between 1976 and 1982.
•            4.  No reduction in lead emissions between 1976 and 1982 from
         vehicles other than automobiles.
•            5.  Automobile lead emissions contributed 90 percent of total lead
•       emissions in urbanized areas in 1976.
              6.  A background lead concentration originating from outside the
•       urbanized area of 0.1 pg/m , quarterly mean, in 1982.
              7.  Air quality concentrations vary proportionally with emissions;
•       also, assumptions normally associated with proportional  modeling apply.
m       CALCULATION
              The proportional model  relating air quality concentrations to reductions
 •       needed to attain a standard  is given by the following equation:

 I

I

I
 -                                             175

-------
     R - A
       ~
         VB
     Where:  R = the proportional reduction in emissions needed (non-
                 dimensional decimal);
                                                                     3
             AB= the air quality concentration in the base year (j-ig/m );
                                                            3
             S = the level of the air quality standard (ug/m );
                                                                3
             B = the level of the background concentration (jjg/m ).
Equation (1) can be expressed in terms of the critical air quality concen-
tration in the base year as follows:
     .  _  S - B R                                         (1A)
     HB      1 - R
The emission reduction obtained in an urbanized area from 1976 to 1982 can
be expressed as follows:
     R = E76 " E82                                         (2)
           E76
     Where:  E7g = lead emissions in the area in 1976;
             E82 = ^ea(* em''sslons in tne area in 1982.
Lead emissions can be separated into two classes, automobile (E.) and
non-automobile (EN), where:
     E76 = EA,76 + EN, 76                                  (3)
and
     E82 = EA,82 + EN,76*'                                 (4)
From assumption No. 5,
     EA,76 = °'9E76
or
     E76 = 1J1 EA, 76.                                    (5)
*Since non-automobile source emissions are assumed to remain constant.
                                    176

-------
I
I
1
             Therefore,
|                EN,76 " 0-1 E76
•                      -0.1(1.11 E,^)
                  EN,76 = OJ1 EA,76
I           The ratio of automobile emissions in 1982 and 1976 is the same as the
             ratio of the automobile emission rates for the two years, or,
                  EA,82 = e82,s  ,                                        m
                          -                                              (/)
I                   A,76   e76,s
                  where eRo _ = emission rate for 1982 and speed s (g/day);
                         O£ ,S
£                      e7fi   = emission rate for 1976 and speed s (g/day);
™                The emission rate from automobile sources as area sources is
•           calculated by the following equation:
I
   fn,s
       V                                         (8)
                  where: e    = emission rate for calendar year n and speed s
                          n ,s
|                              (g/day);
                         ar   = percentage of lead burned that is exhausted
                          Is
                                (nondimensional; expressed as a decimal);
_                       Pb   = probable pooled average lead content of gasoline in
*                              year n  (g/gal);
V    = vehicle-miles travelled daily (vehicle-miles/day);
f    = average fleet fuel  economy for calendar year n
|                              and speed s (vehicle-miles/gal).
—           The terms a  and V are assumed to remain constant from 1976 to 1982.
™           Using values of Pb  and f    from the draft lead guideline, (interpolating
                               n      n 5 s
•           where necessary and using the average fleet speed), for 1976, Pb  =
             1.4 g/gal and fR s = 13.0 miles/gal; for 1982, Pbn = 0.34 g/gal and
I           f   =17.9 miles/gal.
              n ,s
                                                177

-------
Substituting these into equation (8), we obtain





      76 >s        13.0 mi/gal



           = 0.108 (a V) g/mile



and


     e     _ (0.34 g/gal) a V


     e82,s "	TTTiii—



           = 0.019 (a V) g/mile.
                     *>


Substituting these results into equation (7), we obtain



     E. 7, =  0.019 (acV) g/mile
      M • / D            S
     Eft 76 B  0.108 (asV) g/mile



           = 0.176,



or




     EA,82 = °'176 EA,76.



Substituting this expression and equation (6) into equation (4) yields



     EM = (0.176 E. 7,) + (0.11 Efl 7,).
      O£           M j / O           r\5/O


         = 0.286 EA>76



Substituting this result and equation (5) into equation (2) yields



     R = E76 " E82


            E76



       = (1.11 EA   )- (0.286 EA>?6)



                    1J1 EA,76



       = 0.74.



Substituting this into equation (1A) yields the critical  air quality



concentration:



     A76 =  1-5  ug/m3 - 0.1 ug/m3 (0.74)



         =  5.5 jjg/m .
                                   178

-------
I
I
I
I
I
I
_                                       APPENDIX G
                           ROLLBACK MODELING -- BASIC AND MODIFIED
I
I
I
I
I
I
           Reprinted from de Nevers, Noel, and Roger Morris,  Rollback Modeling ~
—         Basic and Modified.  Paper No.  73-139, presented at the Annual  Meeting
•         of the Air Pollution Control  Association, June 24-28,  1973, Chicago,
•         Illinois.
I
I
I
I

-------
I
I
                                                                        No. 73-139
                       Abstract of a Paper to be Presented to the APCA
                        Meeting, June 24-28, 1973, Chicago, Illinois
I
                          Rollback Modeling - Basic and Modified
I                                         by
•                           Noel de Nevers* and Roger Morris
                             Environmental Protection Agency

™             The body of information  presented in this paper is directed to those
•        interested in the formation and evaluation of emission control strategies
          for air quality standard achievement  and maintenance in urban areas.  The
J        paper will be of particular interest  to those charged with the selection
          and use of models for  relating urban  emissions to expected air quality.
•             The "rollback" or "proportional" model is widely used in pollution
•        control calculations and included in  the guidelines for preparing and
          evaluating state implementation plans.  Its basis and limitations are
£        not widely known or understood.   In this paper the basis and limitations
          of rollback are listed and discussed. In its simple form, it is only
™        truly applicable to a  very narrow range of pollution control situations.
•             Four modified forms of rollback  are derived and presented.  The first
          of these extends basic rollback to multiple categories of sources, which
j        may experience differing rates of growth and degrees of control.  The
          second modified form extends  this multiple-source version to include the
•        effects of average stack heights  for  the various categories.  The third
•        model includes the radial distance from source to receptor, and the
          fourth model adds wind direction  frequency.
I        _____
          *Present affiliation:   Department of  Chemical Engineering, University of Utah
I
_                                              181

-------
     All  of these rollback  models  including  simple rollback are shown
to be derivable from the same, simple  formula  which  is
          crb) future   ztkij  ej^future       .                     (1)
                           ij ej-lbase
where c.,- is the pollutant concentration at point i
       b is the background concentration
      e- is the emission rate of category j
       J
 and k^ is the contribution of a unit emission rate from category j to
         concentration at point i.
Each model differs from the others only in the number of categories into
which the area emissions are divided and in the method of estimation of
     The principal conclusion is that "rollback" modeling should  begin
with equation (1) and proceed to the form that is justified by plausible
assumptions and data availability.  Simple rollback would seldom  be used
if this procedure were adopted.   In most cases the  third or fourth model
would be more appropriate.  These latter models are significant steps
from simple rollback in the direction of the more advanced "diffusion
models" which are often used in  air quality modeling.  They are clearly
much less sophisticated than the large computer models now in use but
more sophisticated than the simple rollback formula used for the  state
implementation plans.  Thus, they may play a useful role as intermediate
tools between the very simple, hand-calculated rollback formula and the
more advanced and complex models.
                                     782

-------
I
•
                                                                            No.  73-139
             MODELS AND MODELING
•                A model  is an intellectual  construct, which  represents  reality, and
             can be manipulated to predict the consequences  of various  actions.  In
•           recent years  there has been some controversy  over what  is  a  model,  and
m           whether we ought to base air pollution  regulations,  which  have  the  force
             of law, on models.  According to the above definition,  simple rollback  (1)
•           is a model, and regulations based on it (e.g.,  State Implementation Plans)
             are ultimately based on modeling.  As this paper  shows, rollback  is a very
             simple model, probably the simplest air pollution model which can be used
m           to make quantitative predictions.  Considerable effort  has gone into more
             complex models (normally called "diffusion models"), (2-5) whose  function
I           is to do the same thing as rollback, but with greater detail and  accuracy,
             and with greater confidence in the result.
|                In any modeling effort one is constantly making a  tradeoff between
_           simplicity and accuracy.  The true physical world is complex; we  will not
             have models of total accuracy unless they are complex.   An accurate model
•           cannot be simple; a simple model cannot be accurate. We all strive to
             produce a modeling breakthrough like Copernicus did, when  his much  simpler
|           heliocentric model of the solar system replaced the  extremely complex
p           Ptolemaic geocentric model, and produced more accurate  results, i.e., more
™           accurate predictions of the observed positions  of the planets.  So  far  no one
ff           has made such a breakthrough in air-quality modeling, so our current choices
             are complex models of fair accuracy, and simple models  (like rollback)  with
|           lesser accuracy.

I

I
.                                            183

-------
SIMPLE ROLLBACK (OR PROPORTIONAL)  MODELING
     The simplest and most intuitively  obvious  air  pollution model is
the qualitive one which says  "If you  reduce  emissions,  the air will
become cleaner."  This is the intellectual basis  for  all air pollution
control regulations enacted pntil  the late 1950's.   It  fits logically
with the "maximum technology" approach, which simply  requires all
emitters to use "good engineering  practice"  in controlling pollutant
emissions.  By the late 1950's it  became clear that in  Southern
California emissions from automobiles would  have  to be  reduced beyond
what then constituted "good engineering practice."   To  provide a.basis
for setting numerical standards, those active in  that area developed
the next level of air pollution model, which in its current form is
called  "simple rollback" or  "proportional modeling."
     In its most basic form rollback assumes that the concentration
of any  long-lived pollutant at any point is  equal to the background
concentration of that pollutant plus some linear function of  the total
emission rate of that pollutant in the area which influences  the
concentration at that point,
     ci = b + ke                                                  (1)
where  c^ is the ambient  concentration of one specific pollutant at the
i-th point, normally expressed in yg/m3,
     b is the irreducible  background concentration of that pollutant
for  air uninfluenced by  those nearby emitters which influence the
concentration at point i,  normally in vg/m3,
      k 1s a  proportionality  factor, which takes  Into account the
meteorology,  location of all  emitters as seen  from point 1, and the
                                   184

-------
I
I
             other factors which influence the source-receptor interaction  at
             that point.  Its normal  dimensions are (yg/m3)/(gm/sec)  and
                  e is the total emission rate of all  emitters of that  pollutant
•
             within the geographical  area modeled, normally a city  or  metropolitan
I           area; its normal dimensions are gm/sec.
                  For standard-setting purposes one proceeds by solving  Eq.  (1)
|           for e, and defining the allowable emission rate,
•                Allowable = ('allowable "b)/k                               (2)
™           where the "allowable" subscript indicates the allowable emission rate
•           is that which produces the allowable concentration at  the point of
             interest.  If we further assume that ca-]-|owab-]e is the applicable
P           ambient air quality standard for that specific pollutant, which we
_           will call std, then we may write
*
                      Allowable = (std -
•           To solve this equation we need the value of k.  From the discussion of
             Eq.  (1) it is clear that k is not a single constant for a given city,
H           but  is a function of location within the city; it is higher for points
—           near major emission sources than those far from them.  In American air
™           pollution law the standards must be met at every point, so we need the
•           value of k corresponding to the highest value of c.  Solving Eq. (1)
             for  this value we find
I                    k - (Sax - b)/e                                         «>
_           Here cv is the highest pollutant concentration in the region of
^H            .     luGA
•           Interest.  Substituting the value of k from Eq. (4) into Eq. (3) we
•           find
                      Allowable a fi(std - b)/(cmax " b)                       (5)
I
•

-------
Figure 1  illustrates the relations  in  Eq.  (5).
     The  next manipulation commonly made  is  to  write
         eallowable = (population)(allowable emissions per
                                unit of population)                (6)
Here the  appropriate population may be a  population of residences  or
automobiles, or industries, etc.  Similarly  one replaces  the  e  in
Eq. (5) with
         e = (population at time of measuring c)(emission
                       per unit of population at time of
                       measuring cmax).
Dividing  both sides of Eq. (5) by e, and  making these substitutions,
we find
          population)(a11owable emissions per unit of population)
          population at time of measuring cm  ) (emissions  per
              unit of population at time  etc.^f
              - (std -b)/(cmax)                                   (8)
We then simplify this by defining
                               	(population)	
         gf = growth factor = (population at time of  measuring  cT  (9)
                                                                max
and
                                ^allowable emission per  unit  of population
         ef = emission factor Remissions per unit at time of measuring cmax)
                                                                  (10)
Substituting these in Eq.  (8) and solving for ef we find
         ef =  (std - b)/gf (c    - b)                              (11)
                      1 3  x max
Finally we define  the required  percent reduction in emissions per unit
of population as
         R = 100%  (1-ef)
           -100%
                                   186

-------
I
                           (gf.c     - std + b  (1 - gf))
                    =  100%      max _
                               gffC    - gfib                                (12)
                               3   max   3
I

1
I
           Eq.  (12)  (which  is  the  basic  result  of  the  linear assumption of Eq.  (1))
I         appeared  too complex  to the early  workers in air pollution, so they
           simplified it by setting the  b  (1  -  gf)  in  the  numerator  to zero and
•         changing  the gf.b in  the denominator to a simple b.   (This is equiva-
           lent to changing the  denominator of  Eq.  (11) from gf  (c    - b) to
•


I                             (gf.c     -  std)
                    R = 100%     max
           (gf.c    - b).   Making  these  changes  we  find
                max
J         which is the "simple rollback" or "proportional model"  equation  used  in
—         previous work into auto emission standard  setting,  and  which  is
*         specified in the guidelines  for preparing  State  Implementation Plans  (1).
•              One may most easily see the effect  of this simplification by
           constructing the ratio
                    0-R)Fn  t-i9\    (std -b)/gf(c   -b)    gf.c     -  b
                         ta.-  U2)  _              max    =     max             n/j)
                                        ^FT/Tgf.Cmax-b)   9f«cmax  -  9^

           If we divide both top and bottom of the  right  hand  side of Eq.  (14)  by
•         c  v we see that the ratio of (1-R) for  the  complete  equation to  (1-R)
            max
           •for the simplified equation depends on gf and  (b/cm,v). Figure  2  shows
                                                            IMG A
           the values of this function for several  values of gf.  From it  we see
•         that if gf is one the two equations give the same value,  and if b is
           zero the two equations give the same value.  For all  other values of gf
           and b/c     (l-R)for the complete  equation  is  larger  than  (1-R)  for  the
                  max
           simplified equation, Indicating  that  the  simplified equations  leads  to
           a more stringent set of standards  than the  complete  equation.   As
•         long as b/cmax 1s small the ratio  1s  small, but as b/c^,^  becomes

•                                               187

-------
greater than about 0.3, the ratio begins  to  grow  rapidly,  becoming
infinite for b/cmax equal  to one.  Thus,  for areas  in which growth is
low, or background is small compared to the  highest measured  concen-
tration, this simplification makes a small change,  and  always makes it
in favor of more restrictive standards.  For large  values  of  gf  and in
areas where the background is a large fraction of the  largest measured
concentration, this simplification makes  the standards  much more
restrictive than the complete equation.
FIRST NUMERICAL EXAMPLE
     To illustrate how this equation works  (and to  compare it with
subsequent equations) we will work a numerical example  here.  This
example will be expanded and continued in further parts of the paper.
The example  considers "pollutant x" in "hypothetical  city."  The data
used in this part of the example and the further parts  are all given  in
Table 1.
     For this example, c  „ = 200 yg/m3
                        ITlaX
                       b    =10 pg/m3
                       gf   =   1.30
                       std  = 100 yg/rn3
Thus, using Eq,  (13)  (simple rollback) we have
(1.3 «  200 -  100)           160
(1.3 .
         R = 100% (1.3 . 200 - 16)  • 100* . 25ZT « 64.0%
I.e., to meet the standard with this growth factor, all emitters must
reduce their emission rate by 64%.  If we had used Eq. (12) (the
complete form of rollback) we would have found,
                  (1.3 • 200 - 100 + 10(1 - 1.3))   100% (157)
         R - 100% (1.3 « 200 - 1.3 t 10J          •     247    • 63.6%
                                    188

-------
 I

             In this case, the simple rollback formula yields  a result 0.6% greater
 £           than the complete equation, i.e., the uniform emission reduction  required,
 —           as calculated by the simplified equation, is 0.6% greater than necessary.
 •           THE LIMITATIONS OF SIMPLE ROLLBACK
 •                 Simple rollback (Eq. (13)) is widely used because it is  simple
             and easily understood, and because it requires very little input  data.
 •           It has some severe limitations, which are discussed here.
                      1.  It is a purely theoretical model, for which no experimental
 •           verification has ever been attempted in a metropolitan area, and  which
 •           can probably never be subjected to experimental verification in a
             metropolitan area.  The reason that the experimental verification for a
 I           metropolitan area has not been attempted, and probably never will be,
             is that the relation between concentration and emission rate (Eq. (1))
 •           assumes that all other factors remain unchanged, including the spatial
 •           distribution of emissions.  Thus, to test the equation one would  have
             to reduce  the emission of each and every emission source in the area
 •           by the same percentage.  For practical reasons this does not seem
             possible in a metropolitan area.
 •                This  is not as severe a shortcoming as it might appear, because
 m           the theoretical basis of Eq. (1) is quite plausible.  However, it could
             be wrong in several ways.  If emissions influence climate (e.g.,  by
 I           changing turbidity of the atmosphere) then the linear assumption  in
             Eq.  (1) would probably prove false.  If pollutant disappearance  (e.g.,
 I           by agglomeration or photochemical reaction) is not a linear function

I

I
.                                              189

-------
of pollutant concentration (which it probably is not)  then we would
expect a non-linear relationship between emissions and concentration.
There are probably other causes which could lead to non-linearity in
this relation as well.
         2.  The application of the equation requires  that we know the
value of c   , the highest concentration of pollutant  in the area.  In
          liId A
general usage one substitutes the highest observed concentration
(c) ks for cmax.  These two will only be the same if one of the air
quality measuring stations is located at the point of  maximum concen-
tration.  This assumption is non-conservative, i.e., leads to less
stringent regulations than would be used if the true value of cmax were
known.
         3.  The growth factor  (gf) as used here assumes that all emission
rates will grow, without changes in other significant parameters  (e.g.,
distribution of emissions, city size, stack heights, etc.).  If the
value  used here is the projected incre.ase, for example, in vehicle miles
per  day per square mile of the downtown part of the same city, and there
Is  reason  to believe  that the percentage Increase in vehicle miles per day
per  square mile will  be the  same for each square mile of the area of
Interest,  then  this 1s a satisfactory way to use the growth factor.
If  the value 1s simply the projected increase in vehicle population or
vehicle miles per  day 1n the total metropolitan area then  there is no
reason for believing  that growth distribution will  be uniformly distributed,
and there  are reasonable grounds for assuming that  the model would give
misleading results.
                                 190

-------
 I
 I
 _                  If  the  air quality standard to be met is a short-term standard,
 *            which  is most  severely tested  by meteorological situations which mix
 fl            the  pollutants thoroughly  in a finite volume of air  (as for example
              under  inversion conditions in  a completely enclosed  valley) then
 |            the  growth factor  as  shown in  the  simple model is probably satisfactory
 —            if the boundaries  of  the area  considered are the same  as  the
 •            boundaries of  the  area whose emissions are trapped in  this body of
 •            air.  If, on the other hand, the standard to be used is an annual
              average  standard,  or  some  other standard which does  not represent
 |            this "thorough mixing of all emissions in a fixed, finite volume of
 —            air" then the  growth  factor used in simple rollback  should be very
 •            conservative,  leading to much  more restrictive standards  than would
 •            be needed for  a model which took into account growth in emissions
              per  unit area  in the  areas of  greatest interest, rather than total
 £            emissions in some  arbitrarily  defined metropolitan area.
 _                  The growth factor as  defined  in Eqs .  (12) and (13) is simply
 •            the  ratio of the population of emitters  (residences  or cars or
 •            factories etc.) at  the time when the standard is to be  met, divided
              by the population  of  emitters  at the time cm,v was measured.  There
                                                         IMG A
 •            has  been some  discussion over  whether this future population should
              be obtained  by linear or logarithmic extrapolation of  existing
 •            population  trends. This is really a question outside  of  the basic
 •             rollback model.   It asks for  the value of the population  on the
              appropriate  date;  it  is the responsibility of the demographers,
•            planners, etc. to  determine the most reliable way of estimating that  value.

i
i

-------
         4.  Simple rollback is  applicable  for short-lived  pollutants
(those whose "half-life" in the  atmosphere  is  comparable  to their
travel time across a metropolitan area at the  wind  velocity of
interest) only if the spatial distribution  of  emissions is  unchanged.
This means that any growth by "expanding into  the suburbs"  cannot  be
modeled for a short-lived pollutant by simple  rollback.   Short-lived
pollutants can be modeled by simple rollback for the "complete
mixing within a fixed finite volume of air" situation only  if there
is instantaneous horizontal mixing of pollutants over the entire area
being modeled,
         5.  Simple rollback is  applicable to  the problem of determining
the effect on air quality of a change in emission rate of one emitter
or one class of emitters, without equal percentage  changes  in the
emission rates of the other classes of emitters, only if  one of the
following  three situations exist:  either (1)  The class we  are consid-
ering is by far the largest contributor of the pollutant  in question,
so that we can ignore the effects of the others, or ignore  the
inaccuracy of making the assumption that their contribution to the
concentration at the worst point has the same  factor of proportionality
as the contribution of the source or groups of sources we are consider-
ing,  or  (2) The class we are considering has the same temporal, spatial
and vertical distribution of emissions as the average of  all the
other emissions in  the area  being modeled, so that a change in its
emission  rate has  the same effect on emission distribution in time
and space  as a properly-scaled reduction in all emissions rates would
have, or  (3) The standard for this pollutant 1s a short-term standard
which is  most severely tested in periods of excellent mixing within
                                 192

-------
I

I
_           a limited unchanging air volume, as might occur in a well-mixed layer
*           under an inversion in a completely closed valley with no exchange of
flj           air with the surrounding area.  Then we can ignore the difference be-
             tween the excellent mixing described here and the perfect mixing which
I           rollback would require.
_                If none of these three conditions can be shown or reasonably
™           assumed to exist, then the application of simple rollback or proportional
•           modeling to the question of the impact of changes in the emission of
             one class of emitters on ambient air quality is totally without
|           theoretical or experimental foundation.
_                    6.  Simple rollback assumes that the meteorological conditions
™           which existed when cmax and e were measured are those which will exist
•           on the date when the standards are to be met.  Climate does change
             without human intervention, and growth of cities and growth of energy
|           release does influence climate, so this is not necessarily a sound
_           assumption.  The more advanced models (2-5) generally also make this
•           assumption, so they have this limitation in common with rollback.
•           However, with them one can compute the effects of changes in meteorology,
             and thus estimate the sensitivity of the prediction to such changes;
|           with rollback such a sensitivity test does not appear possible.
             MORE ADVANCED MODELING SCHEMES
•                The previously listed limitations of simple rollback have led air
•           pollution workers to try to develop modeling schemes which do not have
             these limitations.  Most of these improved models begin with Eq. (15)
|                    c1 = b + Ek^ 6j                                         (15)

I
m                                              193

-------
where c^ is the concentration at receptor no i,
      e- is the emission rate for emitter j
       J
      and k-.  is the source-receptor-interaction for emitter j  and
            • w
receptor i.*  One customarily uses a "source inventory" to find the
individual values of the BJ (either source-by-source for large  "point
sources" like power plants, or in aggregates called "area sources"
for autos and home heaters and incinerators).  The values of k^ are
                                                              * J
based on meteorological calculations.  These meteorological calculations
generally fall into two categories, each of which seems applicable
under some meteorological circumstances.  These are:  (1) moving box
models, in which pollutants are assumed to be totally mixed within  certain
vertically-limited air parcels which travel  with the general wind velocity,
and may exchange matter with surrounding boxes as they move, and (2)
gaussian plume models which assume that the pollutants are dispersed
according to  "gaussian plume" formulae.
     Such models may be instantaneous, solving for the concentration
at a given  point at a given time, taking into account only the  current
and recent  past meteorology, or long term, sampling the various
meteorological conditions and assigning frequencies to each, and then
computing the concentrations for each meteorological condition,
multiplying this by the frequency and summing to compute the long-term
average  concentration.
      *The  k-jj  shown here is normally shown in the air pollution
 literature as  (x-j/Qj).  The simple form here 1s used for ease of
 typing  and clarity of presentation.
                                    194

-------
 I

 I
                      These more advanced models make more detailed predictions than
 •               rollback, which makes it possible to test their assumptions against
 •               experimental data.  They allow one to predict the spatial distribution
                 of various concentrations of pollutants on a given day, or for some
 I               long time period, which is not possible by simple rollback.  These com-
                 puted distributions can be compared with measured air quality, and the
 •               models modified to obtain superior agreement.  In addition they can
 •               make short-term predictions for specific meteorological conditions,
                 which can be compared with observed values.  Because these more advanced
 I               models have this testing potential, they have been widely studied and
                 tested.  As a result we have a much greater degree of experimental
 I               confirmation of them, and much more confidence in their predictions
 •               than we have for simple rollback.
                      The relationship between these more advanced models and simple
 I               rollback may be clarified if we consider the two circumstances in which
                 the more advanced models give practically the same answers as the simple
 I               rollback equation.  These are:
 •                        1.  The situation in which all polluters in the region make
                 proportional reductions, and the proportionate spatial, temporal and
 I               vertical distribution of these pollutants is unchanged.  In that case
                 we may write
 •                        ej s   (V original ' Pr°Portl'onaluy factor              (16)
 .               Substituting this into Eq. (15) we find
                          c< • b + z[k.. (e.)           • proportionality factor]   (17)
                           1          ij   j  original
 I               If and only if this proportionality factor 1s the same for each of the
                 j emitters  (or classes of emitters), then we can factor it out of the
 |               summation sign to get

i

-------
         Ci = b + prop,  factor •  spcjj  (e.)  original]              (18)
But if, as assumed, there is no change  in the relative spatial  and
temporal distribution of emissions then the  summation  term is  a constant,
because none of the factors which influence  the k^-  have changed and
the (e-j)  oriqina-| are constants.  Thus we have
         ^ = b + prop,  factor •  a constant                       (19)
which is merely another way of writing  Eq. (1).  Thus, in this  case both
the advanced models and simple rollback indicate that  at any point (not
only the most polluted point) the observed concentration for the appro-
priate time period is a linear function of the proportionality factor
applied to the baseline emission rate.
     The other situation in which rollback and more advanced models
coincide is the situation in whch the standard of interest is a short-
term one which is most severely tested during a period in which we have
perfect and instantaneous mixing of pollutants within  a finite air
volume.   In this  complete-mixing situation all the values of k.. are
the same,  so that Eq. (15) becomes
          Ci » b + kij zej                                         (20)
but
          z:ej = e
so that this also reduces to  Eq.  (1), with the further proviso that
the concentration is  the same at every point in the air mass of interest.
For this  complete-mixing situation both rollback and the more advanced
models  must  give  the  same results.
EXTENSIONS OF  SIMPLE  ROLLBACK
     As described above, more complicated modeling procedures are
theoretically  sounder and better experimentally verified than simple
                                  196

-------
 I

 I
                rollback.  Thus, one might well ask why bother with improved versions
 II             of rollback?  The justifications for so doing are:
 •                     1.  The required input data for rollback is much less than the
                required input data for the more complicated models.
 •                     2.  Only a few of the best-equipped air-pollution groups in the
                country are capable of performing the calculations for the more advanced
 I             models, while anyone with a pencil and a slide rule can do rollback
 •             calculations.
                        3.  Although the more advanced models have been experimentally
 I             tested for SO-.and total suspended particulates, they have not been
                developed and adequately tested for photochemical oxidants, N02 or
 I             hydrocarbons.  Similarly, the area source emission structure used in
 m             the  models for S02 and TSP is not necessarily applicable without further
                work to emissions from motor vehicles.  These should not be drawbacks
 I             to using the advanced models without this adequate verification, if
                one  intends to use instead the totally unverified rollback model in
 8             their place; but in many minds it apparently is.
 M                  For all of these reasons, there appears to be a need for improved
                and/or extended rollback models, to use until we have all the necessary
 I             data to use the more advanced models, and we have adequate verification
                of their predictive ability, and there exists widespread capability
 |              and/or willingness to use them.
 •              ROLLBACK WITH VARIOUS EMISSION CATEGORIES
 *                   The first obvious extension of rollback can be made if we wish to
•              study the effects of various classes or groups of emitters of a single
                pollutant.  We can treat this problem by returning to Eq. (15), and
i

i

-------
assuming that the k- •  for the various  classes  are  the  same.   As  dis-
                   ' J
cussed previously this has no experimental  basis and is  only theoreti-
cally defensible if all the classes  have the same  spatial, temporal
and vertical distribution or if we have a totally-mixed-limited-air-mass
situation.  If these  assumptions can be made then
         Cj - b + k. .  raj                                         (22)
We may determine this  value of k.. from the maximum value of c
measured in the region under study during some baseline  period,  and
the known emissions during that period.  We find
               'Wbaseline 'W                                (23)
Substituting this value in Eq. (22) we find
         cmax = b + (cniax-baseline -b) "j/^Aasellne          (24)
Eq. (24) can be arranged in several useful forms.  If we write
         ej • (e       ' Sf  ' cf                                 (25)
where ef and gf have the same meanings as before, and define fractional
contribution, fr, as
         <* d)base =- C        '                                   <26)
then Eq. (24) can be rewritten as
         Sax ' " * 
-------
1
1



1
1
•
M

1


1

1





1

1
1

1


1




Another useful form of this equation is
(std -b) -
^Snax-base ~ ' — IA j'base^ je jj ^ '
which may be used for control strategy analysis and synthesis. The
inequality has been introduced to indicate that concentrations less than
std are acceptable but greater concentrations are not. Note that the
left side of the inequality is determined by initial conditions while
the right side contains all the terms amenable to control. Various
control strategies may be explored by varying the gf and ef terms and
observing whether the inequality is satisfied.
SECOND NUMERICAL EXAMPLE
To apply Eqs. 27-29 we need additional data. In particular,
we need a logical (and available) disaggregation of emissions, and the
expected emission factors and growth factors for each emission category.
These data for our hypothetical example are from Table 1:
Category LDV OMS HI 01 A
j 1 2345

fr.j .45 .05 .20 .15 .15
gf, 1.28 1.34 1.40 1.22 1.28
u
ef, .4 .8 .5 .7 1.0
J
ancl cm*v k, >std, and b are the same as in the previous example. The
iuax~uase
category abbreviations stand for light duty vehicles, other mobile
sources, heavy industry (including power plants), other industry, and
area sources.
As a first step, we may apply the data to Eq. 27 to see what
concentration results from the expected growth and control.

199


-------
         Siax = 10 + (200-10)[(.45)(1.28)(.4)+(.05)(1.34)(.8)
                +(.20)(1.40)(.5)+(.15)(1.22)(.?)+(.15)(1.28)(1.0)]
         cmax = 10 + (190)(.230 + .053  + .140 + .128 +  .192)
         Snax = 10 + 090K.743) = 151  yg/m3
     The expected control  is  clearly inadequate to meet the standard.
In order to explore the possibilities for additional control,  we use Eq. 29.
100-10        ?  LDV   OMS     HI     01     A      z
200-10 = .474 >_ .230 +.053 + .140 + .128 + .192 = .743
The abbreviations have been added over the category contributions to
aid identification.  Although LDV is the most prominent contributor, it
can be seen by inspection that the inequality cannot be satisfied by
further control of LDV alone.  Let's assume that means  can be  found to
reduce the emission factors for OMS, 01, and A to .4, .5, and  .7
respectively and we wish to know what ef, (LDV) must be to achieve the
standard.  We use Eq. 28.
ef, - (1-tyiOO) = (>45){1>28) C.474-(.05)(1.34)(.4)-{.20)(1.40)(.5)
-(.15)(1.22)(.5)-(.15)(1.28)(.7)3 = .140
and R, = 86%.
Thus, even with the further reduction in emissions from OMS, 01, and
A, LDV emissions must be reduced by 86%, as opposed to the expected 60%,
to meet the standard using the simple  rollback formula.
     Other combinations of emission and/or growth control could be
explored using Eq. 27-29 but to continue the example seems pointless
since, as previously discussed, this extention of rollback only has
theoretical basis  for the case of equal spatial and temporal distribution
of emissions or perfect mixing, neither of which seems applicable to
the general case of  "pollutant x" 1n "hypothetical city."
                                 200

-------
1
1
1

1


1

1

1
1
IV
1

1
1
•V
1

1
1
w
1

1
••
1

1

A SEMI-DIFFUSION FORM OF ROLLBACK
What is proposed now is to take a few steps in the direction of
diffusion modeling. This will require additional complexity, more data,
and several more assumptions. The complexity will remain in the domain
of hand calculation, the data is generally available or soon will be,
and the assumptions, though rather crude, are much better than those
that have been required to support the models considered thus far.
The following development is rather lengthy and mathematical.
Our objectives will be previewed here so that you will know where the
development is. leading. Three successively more complex models will
be presented, all of which use some Gaussian diffusion concepts. The
first considers the effect of emission height only. The second includes
emitter-receptor (ij) distance. The third includes wind direction as
well. The first requires no additional emission categories, but does
require emission height by category and some meteorological data.
The second and third models require the subdivision of emissions by
location as well as by type, and the third model requires more meteoro-
logical data.
Each of the models is based on a determination of a value of k. .
unique to each emission source division.
Recall Eq. (15).
ci = b + z k^.e-j (15)
We may also write
(c.j) . • b + z[k. .(ej) ] (30)
When Eq. (15) 1s divided by Eq. (30) and rearranged, we get
( M
^i"D; - z[k..e.] (31)
ij j — * *
(W-ha<:p-b) sCk1l(Ok ]
max— Dase * ij J base
201


-------
If we substitute




         ei - (ej'base ' <*J  '  efj                                 t25>

into Eq.  (31) ,  we get



                                        '  8f.1  * efjJ               (32)
         
-------
I

I              denominator of Eq.  (33).   The absolute value of k..  is,  therefore,
                                                                 ' J
•              unimportant for our purposes; we only need values  that are proportional
                to kjj.   This fact  will  allow several shortcuts in the following
I              development.
                     Consider the formulation of k^-  which appears in the most widely
I              used meteorological diffusion equation, the Gaussian.
"                                                  r -H2x
                                                                                  (35)
•             where u is the wind velocity
                      H is stack height plus plume rise
I                    y is the receptor distance normal to the plume center line and
•                    ay and a  are the horizontal and vertical  dispersion coefficients.
                The dispersion coefficients vary with meteorological  conditions  and vith
•              the down wind source-receptor distance.
                     Values of of xu/Q have been plotted for y=0 (directly under tne plume
•              center!ine) by Turner6 for six atmospheric stability classes, several
•              values of H, and several inversion heights, L.  These curves are
                reproduced for C and E stability in Figures 3 and 4.
•                   In order to use these curves to generate k^. values for source
                categories, we must make simplifying assumptions for u and 
-------
should be.  The second assumption is that ay is proportional  to x,
the radial distance from the source along the plume centerline.
Actually the best current observations indicate that o  is proper-
                                                      J
tional to x'91 for all stability classes (from Figure 3-2, ref. 6).
The result of this assumption is to weight distant sources somewhat
more heavily than is their due.
     Now consider a category of sources—say heavy industry.   The
individual sources making up the category may be distributed  all over
the region in question in which case it may be impossible to  differen-
tiate the location of this category from the others.  The total
 emission height, however, is a characteristic of source types.  The
 s\ack height plus plume rise for power plants and heavy industry is
 clearly greater than that for light industry, commercial and apartment
 buildings, which, in turn, is greater than the emission height of
 mobile sources.  Our first effort, then, will be to develop k^ • factors
 based on emission height and independent of location.  To do this, we
 need  a measure of the relative impact of a source on the total region
 as the  source height is varied.
      It follows from our assumption that a  « x that emissions from a
                                          «/
 point source will effect a wedge-shaped area of the region for any one
 wind  direction.  This area actually has fuzzy edges.  The cross-wind
 concentration follows the normal distribution with the mode at the
 plume center!ine  and  standard deviation o  .  The two lines bounding
 the wedge are actually  the  loci  of  the cross-wind concentration standard
 deviation points  as  x varies.  Since  a  «  x, any characteristic
                                  204

-------
1
1
1

1


1

1


1

1

1
1

1

1
1

1

1
1
1


width of the cross-wind distribution we may choose will also be
proportional to x. We now define an element of this effected area
as shown in Figure 5. The area of the element is proportional to xdx.
The k,..- for the element is given by the appropriate (xu/Q) curve.
• j
The total source contribution to ground-level concentration in the region
is proportional to the integral of each elemental kin-. Thus
d 1J
kj <* /o (Xu/Q)x xdx (36)
where the i subscript has been dropped to indicate that the kj thus
derived is independent of receptor location. The limit of integration,
d, will depend on the size of the region in question. It should corres-
pond to the radius from the maximum concentration point to the boundary
of significant emission density. If the effect of distant sources is
neglected by this truncation, this is compensated by the assumption
that av « x.
y
The integration indicated in Eq. (36) has been carried out for
Class C and E stabilities in Figures 6 through 9 as a function of d
and H. Figures 6 and 7 show values of the product (xu/Q)(x), and
figures 8 and 9 show the integral of this product out to d. Please
note that the ceiling height, L, has been assumed infinite for these
calculations. This is not the correct estimate for many areas and
is most probably incorrect for high, short-term concentrations. Figures
6-11 should only be applied in areas and for time periods when C or E
stability and infinite mixing height are the best estimates of average
conditions.


205

-------
     To use these figures, or figures like these, each source  category
must be assigned a characteristic H value.  It is accepted practice
among meteorologists to use a minimum H of 10 meters.  Since the
minimum emission height will give the greatest  kj factor, and since
we are not interested in the absolute values of k-, we may make the
                                                 J
use of Figures 8 and 9 more convenient by dividing each kj value by
the corresponding k,- for H-10 meters thus producing a weighting factor
                   J
for emission height ranging from 0 to 1.0.  The resulting weighting
factor will be named h..  Values of h. are plotted for C and E
                      J              J
stability in Figures 10 and 11.  The appropriate hn- factors may be
                                                  J
entered in place of k-- in Eq. (33) thus
                     ' J
         ff\       hj(ejW    . hj(frj)base                  (37)
            J base = ^Wbase-1 " E[hj(frJ)base]
THIRD NUMERICAL EXAMPLE
     In order to apply the h. factors we must estimate the characteristic
H for each source category, the limit of significant emission density
for "hypothetical city," and the appropriate atmospheric stability class.
All three of these estimates involve "engineering judgment" rather than
precise measurement.  Nevertheless, such judgment is considerably more
accurate than the assumption that all emission heights or city sizes
are the same.
     The required estimates, from Table 1, are
         Stability Class - E, no ceiling
         Integration limit - d=20 km
                               206

-------
 I
 I
I
                      Category _ LDV      QMS      HI      01 _
                      H.  (meters)           10       10      100     30      20
                      J
 •                   h.  (From  Figure 11)    1.0      1.0      0.3      .76      .87
                 Stability class E has been selected because the standard for
 •           pollutant x  is assumed to be a  short-term standard, say 8 hours.   It
             has  been observed  that the stability  is usually class  E with a ceiling
 •           over 300 meters during the 8-hour periods of high concentration in
 •           "hypothetical city."  If the standard had been an annual  average,  then
             class  C stability  would be more appropriate.
 I               These data are combined below with that from the  previous example.
                             Values and Calculations for Example 3
            	LDV        QMS      HI        01      A
            j                    12345
             (frj)base             -45        .05      .20       .15     .15          1.0
|           hj                   1.0        1.0       .3        .76     .87
             hj(fr.j)base           .45        .05      .06       .11     .13           .80
I           (wfj)base  ^'  37>    '562       -063     -075      -137    -163         1-°
gf,                 1.28      1.34    1.40     1.22   1.28
  J
ef.                  .4        .8      .5       .7    1.0
  J
I           (wfj)basegfjefj       <288       -067     -053      -117    -209          -734
             Target sum (Eq.  34 and  second  example)  	 .474
             The  maximum concentration  given  the  expected  ef.,  and  gf_-  is
                                                            IJ        J
                      c^  =  10 + (200-10)  (.734) =  149 pg/m3
P           which  is  almost  the same as  in the previous example.   Observe, however,
•           that the  category contributions, (W^i)base9^je^j» are altered due  to the
             weighting by hj.   Eq.  (34) can now be balanced by further reduction  of

                                              207

-------
the LDV contribution alone.   If (wf.|)b    gf-^f,  is  reduced  from  .288
to .028, Eq. (34)  will  balance.  This requires
         ef] = .028/(. 562) (1.28)  = .039
      or R!  = 96%.
Since a reduction  of 96% in  LDV emissions in 5 years  is  rather heroic,
let's see what ef, would be  required if the additional reduction in
ef_, ef4, and efg  mentioned  in Example  2 were applied.   From Eq. (28)

efl=      1         [.474-(.063)(1.34)(.4)-(.053)-(. 137)(1. 22)(.5)
eV~7T9 (-474--317) = -218
and RI is 78%.  This is quite a bit different from the 86% found in
Example 2.
THE LOCATION FACTOR
     A glance at Figures 3 and 4 should convince the reader that source-
receptor distance is the most critical  factor in determining k...  If
we can disaggregate source categories by location, then we can include
consideration of source-receptor distance in the k^ . and simultaneously
relax the worst assumption we have been carrying thus far; namely that
k.. is the same for each source in each emission category.
  • J
     To see how this might be done, consider how it is done in one of
the better known "diffusion models," namely IPP (ref. 3).  That model
computes the concentration at any point i by
                                  208

-------
        cl=
               all
               sources
all wind
directions ,
/frequency\   ]_ .;
  of this     u I
\weather /
                                              (38)
                          velocities
                          and stability
                          classes
In the currently used form of IPP  there are 480 wind direction,  speed  and
stability classes, to each of which a frequency is  assigned from the
observed meteorology.  Then for each  receptor point the concentration
is obtained by a sum of the contributions from each source (individually
for point sources, or in groups called area sources), with each  such
contribution being the sum of 480  contributions for different meteoro-
logical conditions.  (These are normally annual average weather
frequencies, and the program is normally used to compute annual  averages.
From Eq. (38) it is clear that this is not necessary; one can use the
same basic approach over any time  period).
     To make a "rollback" type version of Eq. (38)  we proceed as follows.
First we assume that the joint wind velocity-wind direction-atmospheric
stability frequency distribution is symmetric about the worst polluted
point in the city.  This makes the summation over meteorological
conditions independent of source locations, so that Eq.(38) becomes
          max
an appropriate
meteorological
factor
            E
            all
        individual
        sources
                                                                 (39)
     Next we assume that we can subdivide the area of the city into
annular rings about the most polluted point, as shown in Figure 12,
                                 209

-------
and that in each ring, distribution of each category  of  emitter  is

uniform over the area.  In equation form,
         E Q.   =  E       E
         Air      All      All
         emitters  classes rings
("
J
                                    /e..A
'That
ring
                              dA .
                                nn9
                          (40)
where (e-/A) is the emission density of area element dA.
        J

When we substitute this into Eq.  (39) we obtain
          max
an appropriate
meteorological
factor
  z
  All
  classes
                                                             dA   (41)
                                             2   f(ei)  1  f
                                             Aii'V-  U
                                             rings

However, by our assumption that within each ring the emission  density
for each category is constant, we can take (e-/A) out of the integral
                                             j
sign.  The wind speed may be included in the "appropriate meteorological

factor" because it has been assumed to be the same for all parts of the

city and to be equally distributed in all directions.  We can also note

that if we take our area integral as a symmetrical one shown in Figure

12 that the limits of integration for each ring are from x-j to x2, and
the element of area is

         dA = ZTT x dx

so that
                                                 (42)
         cmax
an appropriate
meteorological
factor

2TT


E
All

E
All
r- x2 -i
/e.\ f /xuwdx
-~7 Xl Q
                                                                     (43)
                                     classes  rings
      The  integral on the right is just the integral we used in the

 previous  version of rollback, and as in all rollback type calculations,

 the factor  in the equation which we have labeled (appropriate
                                 210

-------
1
1

1
1

1

1
1

1

1




meteorological factor) is determined by some form of equation (1), in
whi
ch we relate present emissions to present worst observed air quality.
Thus we can, subject to these assumptions, write:
E £ i- ~l

All All e,) Xo /YU\ »i*
(cmav future -b) Classes rings 1 — fc l^Q'JxT^
max L A * -J (44^
M XT r..+nv~ \W)



All All F(e,) f2 [YU] vjv
. (c , -b) Classes rings v*- "v ^**QJxAU"^ D
max base LA x-j M J Base




The

When Eq. (44) is compared with Eq. (32), it can be seen that
1 x2
kjj « A r1 / M x dx (45)
4 x
value of the integral is found for each ring-source type in
Figure 8 or 9 (or similar curves for the appropriate average stability
1

I
^B
1

1

1
1
1
1
1


and

val
is

emi
ceiling height) by subtracting the value given at x, from the
22 '
ue at X£. The value of A is TT(X - x, ) but since only proportionality
needed, the n can be disregarded.
The k.. thus calculated will be named hr. to signify that both
i j j
ssion height and radial distance from the point of maximum concen-
tration have been considered.








„ 1 x2
hr. = (x£ - x?) / /xyj x dx (46)
J 2 ' xl V Q/X




211


-------
FOURTH NUMERICAL EXAMPLE
     "Hypothetical  City"  is  divided  into  three  annular areas by drawing
circular boundaries of radius  2,  10, and  20  kilometers about the
point of maximum concentration.   We  now need the  distribution of
emission between 15 categories instead  of 5. This  distribution is
taken from Table 1  and is presented  in  the first  column  of  Table 2.
The gf, H, and ef may not be the  same for source  types in different
location zones (e.g., suburbs  grow faster than  city centers).  This
is shown in the following three columns of Table  2.  The hr^ are
found using Eg. (46).  For example,  for heavy industry in the second
ri ng
         x  = 2 km
         x~ = 10 km
         x22 - x^ = 102 - 22 * 96
The integral, for H = 50 m and stability E,  is  0.1  at 2  km  and  0.7
at 10 km from Figure 9.  The value of hrg is
         hrg =  (0.7 - 0.1)/96 =  .00625
This  and the other similarly  calculated hr  factors have been multi-
plied by 100 in the table for convenience.  The other indicated
calculations are similar to the previous example.
      The maximum concentration, given the expected ef. and gfj  is
         cmax  B 10 +  (20° -.10)(.662) =  136 pg/m3
which is somewhat  less than that  obtained with the previous model.
 For  this example,  Eq.  (34)  can easily be  balanced  by further reduction
 of LDV  emissions.  The total  (wfj)base9fjefj for LDV 1s .106+.117+.028 « .251.
 This must  be  reduced to  .063  in  order  to balance Eq. (34)  by LDV control.
                                  212

-------
I
I
•         The required ef-|_3 is
                                                                         .063
                     Ief, o- .063/[(.250)(1.0)+(.250)(1.1) + (.041)(1.6)] = l^± = .107
                        '"J                                                .589
                   or R-[_3 = 89%
 J               If the additional reduction in ef, for OMS, 01, and A mentioned
            in the previous example are applied, and the required LDV reduction is
 B          calculated, the result is
I                                                (.4)                (.5)         (.7) -,
                     efl-3 =   1   [.474-(.033)tT8T - .055 - (.101)177) - (.162)UTO)J
                             .589
 I                   ef, * = —L (.474-.287) = .317
                        1-3   .583
 I          and the required R-|_3 is 68%.  Note that our rollback calculations have
             been  markedly  effected  by  the  inclusion  of  location  considerations.
 •           THE WIND DIRECTION FACTOR
 •                Our rollback  model  has  grown  considerably in  complexity  and  in its
             need  for data  with the  inclusion of each  additional  factor.   It is still
 •           well  within range  of hand  calculation  and available  data.  We will,
             then,  consider one more  factor:  wind  direction.   This will require the
 •           further subdivision of  the region  into sectors.   Four, 8,  or  16 sectors
 •           may be used depending on the accuracy  desired and  data availability.
             If the standard to be achieved is  an annual  standard, the  annual  wind
 •           direction distribution must be known.   If the standard is  for an
             averaging time less than annual, the wind direction  distribution  should


I

I

I

-------
be obtained for those periods  when the standard has  been  exceeded.   A
weighting factor is assigned to each sector such that the sum  of  all
sector weighting factors is equal  to (1.0)  and each  sector weight is
proportional to the time the wind  blows from that sector.  The figure
below shows one possible set of weights for an 8 sector subdivision.
     When the hrn- factor for an emission subdivision is multiplied by
                J
the wind direction weighting factor, a kj,- factor including height,
and location (hi.:) is obtained.  This factor is used in the same way
                vl
as were the h^ and hr^ factors.  This addition to our model is relatively
             j       j
simple, but the subdivision of "hypothetical city" into 8 or 16 times 15
categories is tedious so no numerical example will be given.
     The inclusion of wind direction in the rollback model requires
about an order of magnitude increase in the emission inventory sub-
division.  As this will create as much more effort in performing the
required calculations, it is reasonable to ask under what conditions
the increased accuracy is worth the effort.  Wind direction will be
important if its distribution is decidedly uneven and if the distribution
of emission density and/or growth is also unevenly distributed around
the annular rings.  The latter conditions will almost always be the case,
so the  decision whether to subdivide by sector depends primarily on the
accentricity of  the wind direction distribution and, of course, on the
availability of  the required data.
                                  214

-------
SUMMARY OF THE ALTERNATIVE MODELS
     We may recapitulate by summarizing the results of the several
examples:
     Example 1:  By simple rollback we find that if all sources reduce
their emissions by 64%, then the standard will  be met.  No information
can be obtained from this model about the required rollbacks if all
sources do not make the same proportional rollback.
     Example 2:  By rollback with various emission categories, we find
that the expected growth and emission factors will reduce c max to only
151 yg/m3 (the standard is 100 yg/m3).  Attainment of the standard
through additional LDV control is impossible.  When additional control
is applied to OMS, 01, and A, then the standard can be attained if LDV
emissions are rolled back 86%.  Spatial distribution (three dimensions) of
emissions is not considered in this example.
     Example 3:  By rollback with emission categories and emission
height considered, it is found that expected growth and emission
reduction will reduce c max to 149 yg/m3 which is about the same as
in Example 2.  It is possible to achieve the standard by a rollback of
LDV of 96% because LDV contributions are weighted more heavily than
others due to their low emission height.  When additional controls on
OMS, 01, and A are applied, the LDV rollback required drops to 78%.
     Example 4:  By rollback with emission categories, emission height,
and distance from the point of c max considered, we find that the
expected c max is 136 yg/m3.  The LDV control needed to achieve
standard air quality is 89% without additional control of other source
categories, and 68% with the additional control.
                                  275

-------
     If a fifth example had been presented which  used  a  rollback model
incorporating wind direction as well  as the other factors,  and  if  the
wind direction frequency distribution and emission density  distribution
varied among the several sector subdivisions of "hypothetical city,"
then we could expect results somewhat different from the previous
examples.
CONCLUSIONS AND RECOMMENDATIONS
     1.  Simple rollback is widely used.  However, it has no experimental
basis, and its theoretical basis is restricted to very unusual  situations;
i.e., situations in which we either have perfect atmospheric mixing in
the area of interest, or all emitters make the same percentage  reduction.
     2.  It is recommended that we continue to use simple rollback for
situations in which we can reasonably assume perfect mixing or  equal
percentage emission reduction.  But if these conditions are not satisfied,
then it is recommended that we not use simple rollback, and that calcu-
lations based on simple rollback be understood as having no theoretical
or experimental basis.
     3.  It is recommended that Eq. (34) be used as the basic form of
rollback.  The number of categories, j, and the treatment of wf, should
be the most detailed that the available data will allow.  The assumptions
leading to the particular form of Eq.  (34) used should be explicitly
stated and justified.
     4.  In situations  in which we cannot assume perfect mixing, It 1s
recommended that a  rollback model with emission type, height, and
location considerations be used.
                                 2)6

-------
     5.  In the situations described above we have every reason to
believe that full diffusion models (2) will give more reliable
predictions of the consequences of changes in emission rates and
patterns than any of the rollback models presented herein.  Therefore,
any result obtained by simple or semidiffusion rollback must be con-
sidered a rapid and inexpensive approximation to the more accurate
and reliable results which we could obtain with additional time and
money by a diffusion modeling effort.
ACKNOWLEDGEMENT
             »
     We would like to acknowledge the constructive criticisms of the
preliminary drafts by Dr. Edwin L. Meyer, Jr.
                                 217

-------
                              REFERENCES

1.  Federal  Register, 36,  No.  158,  P.  15490,  Aug.  14,  1971.

2.  Fan, L.  T.,  and Y. Horie,  "Review  of  Atmospheric Dispersion and
Urban Air Pollution Models."  CRC Critical  Reviews  in  Environmental
Control, Oct. 1971.  (This article  contains an  extensive  bibliography
on diffusion modeling.)

3.  Air Quality  Implementation Planning Program.   Developed  for EPA  by
TRW, Inc. under  Contract No. PH 22-68-60, November 1970.

4.  Johnson, W.  B., et al.  Field Study for Initial  Evaluation of an
Urban Diffusion  Model for Carbon Dioxide, prepared by  SRI for EPA
Contract CAPA-3-68(l-69), June 1971.

5.  Sklarew, R.  C., A. J. Fabrick,  J.  E.  Prager,  "Mathematical Modeling
of Photochemical Smog Using the PICK Method."  APCA Journal, Vol. 22,
No. 11, November 1972.

6.  Turner, D.  B., Workbook of Atmospheric Dispersion  Estimates.   Office
of Air Programs  Publication AP-26, 1970.
                                  218

-------
                              NOMENCLATURE
A
b
c
e
ef
gf
H
(hj)
(hr )
   J
(hi.,)
R
std
u
X
y
(x/Q)
                                                     yg/m3
                                                     yg/m3
                                                     gm/sec
area
background concentration
ambient air concentration
emission rate
emission factor (allowable emission rate per unit
of population)/(current emission rate per unit of
population)
fraction of total emissions in class j
growth rate
growth factor (future population) /(current
populations) - Population may be households,
cars, industries, etc.
stack height plus plume rise ("effective plume rise")  m
                                                                  %/yr
a k factor based on emission height only
a k factor based on emission height and source
receptor distance
a k factor based on emission height, distance, and
angular location
constant of proportionality between c and e
see (x/Q)
Rollback percentage
applicable ambient air quality standard
wind velocity
same as (fr.) but weighted by a k factor
downwind distance between source and receptor
cross-wind distance between source and receptor
source receptor interaction coefficient, same as k
horizontal and vertical dispersion parameters
                                                     (wg/m3)/(gm/sec)
                                                     vg/m3
                                                     m/sec
                                                     m
                                                     m
                                                     (yg/m3)/
                                                     (gm/sec)
                                                                  m
                                 219

-------
Subscripts
i
j
max
allowable
baseline
base
b
x
y
z
receptor
source
maximum
allowable to meet standards
corresponding to emission rate from which rollback is to
be calculated
value at distance x
in the crosswind direction
vertically
                                  220

-------
Table 1.  Data Used 1n the Example "Pollutant x," "Hypothetical  City"
Parameter                                       Value (for 8 hrs.)
Sax-base                                       20° *9/m3
b                                                10 yg/m3
std                                             100 vg/m3
Average growth rate (gr)                        5.4%/year
Average growth factor (gf) over 5 years         1.30
Atmospheric Stability Class (for periods above standard)  E (ceiling >300m)
Ring boundary radii                             2 km, 10 km, 20 km
                        LDV        OMS         HI         01         A
fr city
Ring 1
Ring 2
Ring 3
gf city
Ring 1
Ring 2
Ring 3
ef city
Ring 1
Ring 2
Ring 3
.45
.02
.25
.18
1.28
1.0
1.1
1.6
.4
.4
.4
.4
.050
.005
.025
.020
1.34
1.1
1.3
1.5
.8 (.4)
.8 (.4)
.8 (.4)
.8 (.4)
.20
.00
.10
.10
1.40
-
1.1
1.7
.5
-
.7
.3
.15
.003
.097
.05
1.22
1.1
1.2
1.3
.7 (.5)
.7 (.5)
.7 (.5)
.7 (.5)
.15
.01
.09
.05
1.28
1.1
1.2
1.5
1.0 (
1.0 (
1.0 (
1.0 (








.7)
.7)
.7)
.7)
 (ef  in  parentheses  are  the  limit of further possible reduction in the
 following  5 years in  "Hypothetical City.")
H city
Ring 1
Ring 2
Ring 3
10
10
10
10
10
10
10
10
100
-
50
200
30
50
30
30
30
50
30
20
                                 221

-------
Table 1 (Cont'd)
"Hypothetical City" is just that.  The values for all the parameters
were selected to conform generally to observed and expected values of
the 1970-75 period but do not represent any particular pollutant or place,
                                 222

-------
Table 2.   Data  and Calculations for  the  Third Numerical Example
                                                    

.106 1
.117 >
.028 J
.059 ")
.028 >
.006 )
.052 C
.003 J
.007 •)
.081 >
.013 )

,030 •)
.109 >
.023 J

.662
A7A



.251


.093

.055

.101



.162



                             223

-------
    \
      std
    o>
    OS
                       I
                    Allowable

                  EMISSION RATE, g/fcec
Figure 1.   Graphical representation of Equation 1  and the
            computation of eallowable.
                          224

-------
                           max
Figure 2.  Comparison of complete and simplified forms of
           simple rollback, for various values of gf and
           b/c,
              'max
                      225

-------
                                               100
             DISTANCE, km
xu/Q for Stability Class C (from reference
              226

-------
(U
  Figure 4.
        1                     10                    100
            DISTANCE, km
yu/Q for Stability Class E (from reference 6).
               227

-------
                  BOUNDARY OF EFFECTED AREA
                    dx-
   SOURCE
                 (Xu/QUx)dx
Figure 5.   Calculation of  K,.
                            228

-------
 I
 I
 I
 I
 1
 I
 I
 1
 I
 I
 I
 I
 I
 I
I
 I
I
I
I
    0.16


    0.14


    0.12


~.   0.10


£  0.08



    0.06



    0.04


    0.02
                                  i      i      i      i     r~i      i      i     r
'0    0.2    0.4    0.6    0.8     1.0
                                                          9  10  20     40     60    80   100
                  Figure 6.  Product of \u/Q and x for Stability Class  C
                             and celling over 2000 m.  (Note scale  change
                             at x » 1 km and x » 10 km).
                                           229

-------
                 I      I      I      I     I   I  I     I      I
                                         9   10 20    40    60
0    0.2    0.4    0.6    0.8    1.0
Figure 7.  Product of x"/Q and x for Stability Class E
           and celling over 300 m.  (Note scale change
           at x  »  1 km and x » 10 km).
                          230

-------
                               10    20    30     40     50    60    70    80    90    100
0X5
             Figure 8.  Integral  of (xu/Q)(x) from zero to x * d for
                        Stability Class C and ceiling over 2000 m.
                        (Note  scale change at d « 10 km).
                                       231

-------
                                        50    60
70
80
90    100
                        d, km
Figure 9.  Integral of (\u/Q)(x) from  zero  to x * d for
           Stability Class E and celling  over 300 m.
           (Note scale change at d « 10 km).
                          232

-------
            8     10    20    30    40    50     60    70    80     90    100
Figure 10.  Values  of  h  for Stability Class C and
            ceiling over 2000 m.   (Note scale change
            at d -  10  km).
                           233

-------
                                                                     I
                                                                     I
                                                                     I
Figure 11.   Values of h for Stability Class  E and
            celling over 300 m.   (Note  scale change
            at d - 10 km).
                   234
I
I
I
I
I
1

-------
Figure 12.  Annular subdivision and calculation  of hr.,
                         235

-------
                                     TECHNICAL REPORT DATA
                              (Please read Instructions on the reverse before completing)
  1. REPORT NO.
     EPA-450/2-78-038
                                                             3. RECIPIENT'S ACCESSION-NO.
  4. TITLE AND SUBTITLE

   Supplementary Guidelines for  Lead Implementation Plans
                                                             5. REPORT DATE
               July 1978
                                                             6. PERFORMING ORGANIZATION CODE
  7. AUTHOR(S)
             8. PERFORMING ORGANIZATION REPORT NO
               OAQPS No.  1.2-104
  9. PERFORMING ORGANIZATION NAME AND ADDRESS
   U.S.  Environmental Protection Agency
   Office  of Air and Waste Management
   Office  of Air Quality Planning  and Standards
   Research  Triangle Park, NC  27711
                                                             10. PROGRAM ELEMENT NO.
              11. CONTRACT/GRANT NO.
  12. SPONSORING AGENCY NAME AND ADDRESS
                                                             13. TYPE OF REPORT AND PERIOD COVERED

                                                                Final	
                                                             14. SPONSORING AGENCY CODE
  15. SUPPLEMENTARY NOTES
  16. ABSTRACT
        This  guideline presents  information on the development of implementation  plans
   for lead that are not contained  in  EPA's regulations for preparation, adoption,  and
   submission of implementation  plans,  found in Part 51 of  Title 40 of the Code of
   Federal Regulations.  In several  cases, the guidance presented herein is referenced
   in those regulations; EPA will use  this guidance in determining the acceptability
   of a plan.

        The guideline covers the following topics:  general  implementation plan
   development,  reporting requirements,  analysis and control  strategy development,
   siting of  urban area ambient  air  quality monitors for  lead,  new source review,
   and the determination of the  lead point source definition.   In addition, appendices
   cover the  following topics:   procedures for determining  inorganic and organic
   lead emissions from stationary sources, projection of  automotive lead emissions,
   deposition  of particles and gases,  and  rollback modeling.
  17.
                                  KEY WORDS AND DOCUMENT ANALYSIS
                    DESCRIPTORS
                                                b.IDENTIFIERS/OPEN ENDED TERMS
                           c. cos AT I Field/Group
   Air pollution


   Atmosphere contamination control


   Lead
 State implementation
  plan

 National ambient  air
  quality standard
  13-B
  18. DISTRIBUTION STATEMENT


   Release unlimited
19. SECURITY CLASS (ThisReport)
 Unclassified
21. NO. OF PAGES
    238
20. SECURITY CLASS (Thispage)
 Unclassified
                           22. PRICE
  EPA Form 2220-1 (9-73)
                                               237
[U.S. GOVERNMENT PRINTING OFFICE: 1978-640-01?  416 5REGION NO. 4

-------
                                                                                                                                           83  ^

                                                                                                                                           5L 9
                                                                                                                                                        o  o^ o
                                                                                                                                              TJ

                                                                                                                                              >
SJ.K ~£ -
       o =
O. T

5m
U)  O
   m o j» o
   „" o -i
   Do a S
  5
                                                                                                                                m> Tjm-n-o
                                                                                                                                T)(£ ^3^O

                                                                                                                                >§l|Sg
                                                                                                                                UJo o o -n
-------