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 \
) CM
I/) CU 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
)
o
•!->
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
------- |