EPA-450/2-79-002
(OAQPS No. 1.2-123)
Development of an Example
Control Strategy for Lead
by
Albert E. Smith, Marshall R. Monarch,
Byung S. Cho, and Danna M. Hediger
Energy and Environmental Systems Division
Argonne National Laboratory
Argonne, Illinois 60439
Interagency Agreement No.: EPA-79-D-F0502
EPA Project Officers: John Silvasi and Daniel deRoeck
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
April 1979
-------
OAQPS GUIDELINE SERIES
The guideline series of reports is being issued by the Office of Air Quality Planning and Stajidjardjs (OAQP.SitoK
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 Virqinia
22161.
Publication No. EPA-450/2-79-002
(OAQPS No. 1.2-123)
-------
TABLE OF CONTENTS .
. . Page
LIST OF FIGURES • V
LIST OF TABLES . . . . . . . . .... . . . . . . . .
ACKNOWLEDGMENT
NOTE .....'. ...
FEDERAL REGISTER INDEX .
1 INTRODUCTION 1
1.1 Scope and Objectives ........ 1
1.2 Overview . 2
1.3 Description of Example Area ..... 2
2 DEVELOPMENT OF BASELINE DATA 5
2.1 Emission Inventory 5
2.1.1 Point Source Inventory ..... 5
2.1.2 Area Source Inventory 10
2.1.3 Example Area Lead Emission Inventory 19
2.2 Air Quality Data . . 19
2.2.1 SIP Requirements 19
2.2.2 Example Area Lead Air Duality Data .25
2.3 Meteorological Data 31
3 PROJECTING FUTURE EMISSIONS .............. 33
3.1 Sources Other Than Highway Vehicles 33
3.1.1 General ................ 33
3.1.2 Illustrations from Example Area ......... 35
3.2 Highway Vehicle Related Sources . .-...-. ... ... . . 48
3.3 Projected Future Example Area Lead Emission Inventory ... 52
4 ALLOCATION OF EMISSIONS 59
4.1 Procedures 59
4.2 Illustration from Example Area ........... 62
-------
TABLE OF CONTENTS (Cont'd).
5 MODELING PROCEDURES .... ' ,e
.••«.«• 65
5.1 Requirements fir
5.2 Applicable Models . . . . . . ." •[" ] ) j \ ' ' ' ^
5.3 Example Area Modeling Results !!!!*' 69
6 ANALYSIS OF MODELING RESULTS ........... 77
6.1 General
6.2 Illustration from Example Area ....!!!![** 79
7 SELECTION, TESTING, AND EVALUATION OF STRATEGIES ....... 83
7.1 Available Strategies ....... 83
7.2 Selecting and Testing Strategies ...*!] 85
7.3 Strategy Selection •-•......!!*].*]* 86
APPENDIX A Base-Year Activity Levels for Area, Freeway, and
Arterial Sources ........ 89
APPENDIX B Emission Projection Model 93
APPENDIX C HATREMS Emission Factors 99
APPENDIX D Guidelines for Economic, Impact Analysis for
Lead SIP 107
REFERENCES .....
-------
No.
1.1
2.1
2.2
2.3
5.1
5.2
5.3
5.4
B.I
LIST OF- FIGURES
Title
Map of Example Area ,. . . . ; .
Location of Lead Point Sources in Example Area .
Freeway Links for Example Area Line Sources . .
Locations with Available Lead Ambient Air Quality Data
Scatterplot of Baseyear Lead Concentrations at
Calibration Receptors
Modeled Example Area Lead Air Quality in Base Year 1975 .
Projected Example Area Lead Air Quality in 1982 . . .
Air Quality Model Receptor Network
Emission Projection Model
Page
4
12
18
28
- 70
71
72
73
95
-------
-------
LIST OF TABLES
No. Title page
2.1 Base-Year Point Source Lead Emissions .......... 11
2.2 Example Area Baseline Lead Emissions (1975) . . 20
2.3 Base-Year (1975) Lead Air Quality Data ........ 30
3.1 Lead Analysis Time Period Requirements 33
3.2 Employment Growth Sates for Projecting Point Source Emissions . 37
3.3 Growth Sates for Projecting Area Source Emissions 38
3.4 Replacement Rates 40
3.5 Projected Example Area Lead Emissions (1982) 53
4.1 Lead Emission Allocation Parameters , 63
4.2 Sample Listing of 1982 Allocated Area Source Lead Emissions . 64
5.1 Base-Year (1975) Modeled Lead Concentrations 74
5.2 Projected (1982) Modeled Lead Concentrations 75
6.1 Sample Source Contribution Analysis for Example Area .... 80
A.I Base-Year (1975) Area Source Activity Levels, ...... 90
A.2 Example Area Freeway Link Lead Emissions for Base Year 1975 . 91
A.3 Base-Year (1975) Tailpipe Lead Emissions from Vehicle
Activity on Arterial Roads ... 92
B.I Primary Symbols Used in Emission Projection Model 94
C."1 HATREMS Lead Emission Factors for Point Sources 100
C.2 HATREMS Lead Emission Factors for Area Sources ...... 105
-------
-------
ACKNOWLEDGMENT
The authors would like to express their appreciation to Susan Karash
of Energy and .Environmental Analysis, Inc. and Allen C. Basala of the Economic
Analysis Branch of the U.S. Environmental. Protection Agency who provided the
text for Appendix D on economic impact analysis.
-------
-------
- NOTE
Throughout the text, references to 40 CFR 51, the Federal Register
regulations for the preparation, adoption, and submittal of implementation
plans for the lead National Ambient Air Quality Standard (Ref. 18), are
frequently given in abbreviated form in square brackets. For example,
"[51.83(a)]" indicates that the material being presented in the text relates
to the requirements of 40 CFR 51.83(a). The following index should .facili-
tate the location of most of the information in the text related to those
sections of 40 CFR 51 specific to lead.
Users of this guidance should also note that on July 16, 1979, EPA
revised its recommended procedure for projecting automotive lead emissions
found in Section 4.3 of the "Supplementary Guidelines for Lead Implemen-
tation Plans" (EPA-450/2-78-038). As a result of this revision, a major
portion of the discussion pertaining to automotive lead emissions contained
in this example control strategy guideline is no longer applicable.
Specifically, the guidance found on pages 14-19 and 49-51 is based upon
the old EPA estimating procedure. Copies of the revised Section 4.3
(Projecting Automotive Lead Emissions) may be obtained by contacting any
EPA Regional Office, or by writing to the EPA Library Services (MD-35),
Research Triangle Park, N.C. 27711, and requesting "Revised Section 4.3 -
Projecting Automotive Lead Emissions," EPA-450/2-78-038a.
-------
-------
FEDERAL REGISTER INDEX
Section of
40 CFR 51
*•) . . . . 6
51,17(b) . . . , . ... . . 26
51.80(a)(l) . . . . . . . . . .. 6
51.81(a) . . ... . . . . .6>io
51.81(b) .......... 33
51.82(a) ...... 19
51.82(b) ... . 25
51.83 ......... .65,78
51.84 ......... .59,65
51.84(a) 6
51.84(b) .......... 6
51.85 . . . . 59,65,78
51.86(c) 19
-------
-------
1 INTRODUCTION
1.1 SCOPE AND OBJECTIVES
Current and possible future air quality problems must be analyzed in
order to develop an air pollution control strategy designed to attain the
National Ambient Air Quality Standard (NAAQS) for lead. Previous publica-
1-15 '
tions by the EPA office of Air Quality Planning and Standards have been
designed to give guidance in plan preparation. Reference 16 illustrates the
application of some of the procedures described in these guidelines to the
development of a control strategy emphasizing: particulates. Recentlv
17 '
guidance designed specifically to aid state, regional, and local air
pollution control agencies in meeting the current regulatory requirements
•to
for lead implementation plans has been published by EPA. The present work
illustrates, for a ficticious three-county example area, some of the quanti-
tative and qualitative procedures used in developing a control strategy for
lead. Since it is not possible to cover all the different situations with
which states must deal in strategy development, the scope of this example
control strategy is necessarily limited. The general situation is handled
by reference to appropriate guidance or by providing such guidance explicitly.
Development of a control strategy for lead follows lines parallel to
those followed in developing a control strategy for particulates. Most of
the guidance already available for particulates is applicable to lead and
will not be repeated here; this work is an incremental document, particular
to lead, which goes beyond the illustrations provided in Ref. 16. The major
areas of difference between lead and particulate control strategy development
relate to:
• The use of different emission factors for lead;
• The specified significant lead point sources to
which dispersion models must be applied;
The large fraction of lead emissions caused by
highway vehicles and the consequent importance
of this source category for lead; and
The lead-specific methodology to be used in
estimating projected highway vehicle lead
emissions.
-------
This work focuses on illustrating these and closely related areas where the
development of a lead strategy might differ from the development of a partic-
ulate strategy. In general, the procedures and analyses with which states
are already familiar, based on their experience with particulates, are equal-
ly applicable to lead.
This document does not address issues of intergovernmental cooperation,
public hearings, or procedural and administrative matters related to plan
submission and review. The emphasis is, rather, on the analytical procedures
useful in developing and choosing a control strategy to meet the lead NAAQS,
It is not the purpose of this work to present new policy statements
nor to issue new guidelines for analytical procedures. Rather, it is design-
ed to illustrate existing guidelines and to clarify the application of cur-
rent policy. This document is specifically directed at providing working-
level state and local air pollution control engineers and planners with this
guidance.
1.2 OVERVIEW
The analysis procedure reviewed here consists of the following steps:16
1. Develop baseline data bases,
2. Project future lead emissions,
3. Allocate emissions for modeling,
4. Estimate lead air quality by modeling,
Analyze the modeled results,
Develop and test alternative control strategies, and
Evaluate and select strategy for implementation based
on relative effectiveness, ease of implementation,
and, if desired, economic impacts.
These steps are discussed in Sees. 2-7, and are illustrated by application to
the example area.
1.3 DESCRIPTION OF EXAMPLE AREA
In order to achieve a measure of realism in the example strategy, the
data for the ficticious three-county example area have been based on Cook,
DuPage, and lake Counties, Illinois. This choice was made because a large
-------
amount of data had already been compiled for this area and because there are
a large number of lead sources and heavy traffic flows in the area.
A large lake forms most of the eastern border of the example area,
shown in Fig. 1.1.. The topography is relatively flat throughout the area.
The City of C, located in County C on the shore of the lake, is the major
metropolitan and commercial center in the area. The shores of the lake have
been kept relatively free of industrial development which has tended to
concentrate in the southeastern part of County C, along navigable waterways
to the west and south of the lake and around the satellite City ¥. Much of
the area between cities C and W along the lake is residential and commercial
in character. The area from the City of C to beyond the eastern half of
County D contains many residential suburbs and light industrial parks.
-------
CITY W
LAKE M
CBD
AREAS WITH HIGH EMPLOYMENT IN
POLLUTING HEAVY INDUSTRIES
Fig. 1.1. Map of Example Area
-------
2 DEVELOPMENT OF BASELINE DATA • • '
The set of information required as a baseline for the development of
a lead air quality analysis is made up of three parts: an emission inventory,
meteorological data, and lead air quality data. The information should all
be for the same base year and must correspond to the study area in question.
In the example strategy, 1975 was chosen as the base year, because a full set
of baseline data was available for that year.
2.1 EMISSION INVENTORY
The emission inventory forms the basis for making an assessment of air
quality management problems. Detailed guidance in the development of an
emission inventory has already been published. ' ' ' Specific guidance
in the development of a lead emission inventory is also available. '''
When developing the emission inventory, consideration should be given
to temporal variations in source emissions where possible. Since the air
quality standard for lead is based on a calendar quarter, source emissions
by calendar quarter in the base year would be desirable. In this example
strategy, lead emissions by a calendar quarter in the base year were not
available. Emissions were assumed constant throughout the year.
The example area has it? eastern border in common with another state.
In developing the example strategy, emissions from the other state were not
considered. In such a situation where an actual evaluation is being per-
formed, states may need to consult the bordering state or states regarding
their emission inventories and account for emission sources from the border-
ing states in the development of the control strategy.
2.1.1 Point Source Inventory
EPA regulations define a point source of lead as any stationary source
whose actual emissions exceed 5 tons per year of lead or lead compounds meas-
*Some of the information in Ref. 19 relating to area sources has been
updated.
-------
ured as elemental lead [51.1(k)(2)].* All such point sources must be con-
tained in the summary of the baseline lead emission inventory contained in
the plan [51.81(a)]. In the vicinity of some of these point sources, called
significant lead point sources, the plan must demonstrate attainment and
maintenance through the use of an atmospheric dispersion model [51.80(a)(l)
and 51.84(a),(b)]. It is convenient to assemble the data needed for modeling
while the sources are being inventoried. The significant lead'point sources
are [51.80(a)(1)]: .
• Primary lead smelters,
• Secondary lead smelters,
• Primary copper smelters,
• Lead gasoline additive plants,
• Lead-acid storage battery manufacturing plants
that produce 2,000 or more batteries per day,'and
• Any other stationary source emitting 25 or more
tons per year of lead.
The example point source lead emission inventory was based on a 1975
state particulate emission inventory. Although an existing particulate
emission inventory provides an attractive basis for building the lead emis-
sion inventory, states should be aware that some stationary sources may be
point sources of lead but not point sources of particulates. Thus, in com-
piling the list of lead point sources, states may need to consider sources
not in the existing particulate point source inventory. The inventory
used here was updated by including a number of additional point sources
obtained from a local agency lead emission inventory.
The lead point source emission inventory should include fugitive emis-
sions as well as stack emissions. Special attention should be given to the
fugitive emissions as they may be the dominant lead source for some facili-
ties, m the example area, fugitives were, in many instances, the more sig-
nificant stationary source lead emission. Not many sources in the example
area had annual emissions in excess of 5 tons of lead. Therefore, for demon-
40 OTR 51 iJvS?r f \ f /^ material bei*g discussed relates to
sL I 51;.1(k>(2). °* the Federal Register requirements for lead SIPs.
See pp. «.* and *M^ for additional details and an index to pages dis-
cussing the requirements of 40 CRF 51 specific to lead.
-------
stration purposes, stationary sources emitting more than 0.1 tons/yr of lead
(fugitive plus stack emissions) were included in the point source inventory.
Determination of Point Source Lead Emissions
The determination of point source lead emissions should include both
stack and fugitive emissions where fugitive emission determinations are pos-
sible. EPA indicates that it is preferable that the baseline lead emission
1 Pt
inventory be based upon measured emissions. Stack test procedures have
been published by EPA for determination of lead emissions. EPA has also
published procedures for measurement of fugitive emissions from industrial
21 22 23
sources ' ' and states are not precluded from developing their own
emission factors based on field studies using the published measurement
-I Q '
procedures. In this example strategy, point source stack and fugitive
lead emissions were determined solely by -the application of documented lead
emission factors.
24,25
The first step in the construction of the lead point source emission
inventory involves a compilation of lead emission factors (both stack and
f
fugitive) for stationary source categories. EPA has published documents^
that will be useful In this task. Emission factors keyed to specific Source
Classification Codes (SCCs) are available in EPA's Hazardous and Trace Emis-
sions System (HATREMS). A listing of these factors and associated default
multipliers is given in Appendix C. The nonfugitive component of lead emis-
sions from point sources calculated by using the HATREMS emission factors
and the.operating rates and control efficiency data in NEDS should be avail-
able to the state through EPA Regional Offices. Such an inventory may serve
as a useful basis for the nonfugitive base-year emissions from a majority of
point sources.
The example inventory presented here was developed primarily using the
HATREMS factors as described below. At the time of this work, HATREMS could
not be used exclusively for point source lead emissions factors, as the emis-
sion factor file did not contain fugitive lead emission factors or emission
factors for all point source categories. The HATREMS factors were supple-
mented by factors based on References 24 and 25.
-------
Having compiled a list of point source lead emission factors, the next
step involves application of the emission factors to compute point source
lead emissions. Reference 17 gives the following formula for the determi-
nation of lead point source emissions:
MS = (1/2000) x OR x EF x DM x [1 .-.(W x Y/100)/100] (2.1)
where:
s
OR
EF
DM
W
-• stack lead emissions (tons'/yr) ,
*•Operating rate from NEDS or state emission
inventory (SCC- units/yr) •;.
• Emission factor from HATREMS or other
reference (Ib/SCC unit),
• Default multiplier from HATREMS (dimensionless
HATREMS units),
' Particulate control efficiency from NEDS or
state emission inventory (percent), and
Y - Control efficiency multiplier from- HATREMS.*
Except for DM and Y, all these quantities are similar to the corres-
ponding quantities for particulates. The quantity DM simply gives the lead
content of the process material, for example, ppm by weight of lead in coal.
HATREMS contains default values but the use of locally representative or
source-specific values is recommended if available. Care should be taken to
ensure that ;the units of DM are those appropriate to the corresponding emis-
sion factor EF. The quantity Y gives the percentage of the particulate con-
trol efficiency that can.be applied to lead. Lacking any other data, T *
100 was assumed in this work.
Illustrations from Example Area
The following example calculation illustrates the computation of the
lead emissions from a grey iron foundry cupola (SCC 30400301) using Eq. 2.1.
*SCC (Source Classification Code) units are the operating rate units
used in NEDS and HATREMS and are consistent with the emission factors
used in these systems. For example, the SCC units for external
combustion boilers are "tons burned" and the SCC units for primary
lead smelters are "tons of concentrated ore." Appendix C of Ref. 26
(AP-42) lists the SCC units for each SCC.
-------
According to the state inventory, the cupola produces 3,035 tons/yr of metal
and the associated particulate control equipment is 95% efficient. Thus,
OR = 3035 tons metal/yr,
EF = 0.51 Ib lead/ton metal from HATREMS,
DM = 1.00 from HATREMS,
W = 95 from state inventory, and
Y = 100 from HATREMS.
M = (3035)(0.51)[l-(95 x 1QO/100)/10Q]
s 2000
or
M = 0.039 tons lead/yr.
S
This source was too small to be included in the baseline point source inven-
tory.
At the time of this work, HATREMS did not provide point source lead
fugitive emission factors. Fugitive emission factors were obtained from Ref.
25 and lead fugitive emissions were calculated as above.
The following is an example calculation of fugitive lead emissions
from a reverberatory furnace in a secondary lead smelting operation:
OR =14,327 tons metal/year from state emission inventory,
EFf = 1.73 Ib fugitive lead/ton metal from Ref. 25,
DM = 1.00 (i.e., default multiplier not applicable),
W = 0 control efficiency (i.e., no control), and
Y = 100 so that
= 14,327 x 1.73
nf
2000
or
M = 12 tons of fugitive lead/yr.
In the above calculation there was no control of fugitive emissions. Hence,
W = 0. If, for instance, there were a 50% control efficiency on fugitive
particulate emissions and if it were assumed or determined that the control
efficiency was applicable to lead emissions as well, then the calculation
would have had Y = 100 but W = 50.
-------
10
Table 2.1 contains a listing of each inventoried point source along
with the stack and fugitive lead emissions in the base year (1975). The
locations of these point sources are indicated in Fig. 2.1.
2.1.2 Area Source Inventory
Sources not inventoried as points are inventoried as areas. For lead,
these area sources are of three types:
1. Particulate point sources whose size does not qualify
them for inclusion in the lead point source inventory,
2. The traditional area source categories such as resi-
dential fuel combustion and nonprocess fugitives, or
3. Highway vehicle sources including both specific high-
ways (line sources) and nonspecifically located roads
(area sources).
Determination of Nonhighway Source Lead Emissions
General. For particulate point sources not qualifying as lead point
sources, the same procedures as described in Sec. 2.1.1, including Eq. 2.1
on p. 8, can be used to determine lead emissions. In reporting emissions
in a format similar to that of Appendix D of 40 CER 51 [51.81(a)], the
emissions from these sources would be included as area sources. However,
depending on the air quality model used and the level of detail of the anal-
ysis, it may be desirable to treat these sources as hybrid point/area sources.
In this approach the sources would be inventoried as part of the area source
emissions. The information regarding their locations would be retained,
however, for use in allocating area source emissions as described in Sec. 4.
This approach utilizes the maximum degree of spatial resolution available in
the baseline data without requiring complete treatment and modeling of even
small sources as points. . ' '
Standard methods, as described in Refs. 7 and 16, can be used to esti-
mate emissions from the traditional area source emission categories. Emis-
sions from these categories are generally estimated and projected at the
county level and then allocated to subcounty areas for modeling purposes.
The basic working equation is17
M
(1/2000) x OR x EF x DM
(2.2)
-------
11
Table 2.1 Base-Year Point Source
Lead Emissions
Plant
A
- B
C
D
E
F
G
H
I
J
K
L
M
N
Source
Classification
Code (SCC)
50100101
• -: 30400303
30300802
10200402
30300901
50100101
50100101
50100101
50100101
30400303
30902099
30400401
30300801
30300801
30400402
30400401
10100201
10100201
10100203
10100202
10100202
30300903
30300802
10100203
10100201
30400402
30400402
Source
Description
Municipal Incinerator
Gray Iron-Electric Inductions
Iron -Blast Furnace
Boiler-Residual Oil
Steel-Open Hearth
Municipal Incinerator
ii H
ii ii
ii ii
Gray Iron-Electric Induction
Can Making
Secondary Lead-Smelting
Iron-Blast Furnace
ii it H
Secondary Lead-Reverberatory Furnace
Secondary Lead— Smelting
Boiler-Bituminous Coal
ii it H
ti it ii
ii it ii
ii it ii
Steel-BOF
Iron-Blast Furnace
Boiler-Bituminous Coal
ii -M ii
Secondary Lead-Reverberatory Furnace
\
ii ii ii H .
Lead Emissions (tons/yr)
Stack
29.76
9.02 -
.87
.03
10.57"
0.83
0.83
0.83
0.83
1.41
1.44
1.95
.45
.74
.73
.16
.06
.23
.49
.19
.08
.29
.06
.34
.05
.80
.67
Fugitive
0.02
9.02 '
0.02
b
3.85
neg
neg
neg
neg
1.41
b
1.20
neg
neg
12.39
.80
b
b
b
b
b
2.94
.27
b
b
b
b
Total
29.78
.'1R.04
.89
.03
14.42
15.34
0.83
0.83
0.83
0.83
3.32
2.82
1.44
3.15
.45
.74
1.19
13.12
.96
14.08
.06
.23
.49
.19
.08
1.05
3.23
.33
3.56
.34
.05
.39
.80
.67
Includes all lead sources in plants whose total emissions exceeded 0.1 tons lead/yr and
all plants on the significant source list.
No fugitive emission factor available.
-------
12
COUNTY L
COUNTY C
COUNTY D
N
E M
Fig. 2.1. Location of Lead Point Sources in Example Area
-------
13 .
where the various factors have the same meanings as in Eq. 2.1 and it has
teen assumed that emissions are uncontrolled. HATREMS provides the most
comprehensive source of emission factors and default multipliers for area
sources. Reference 24 may also be useful in determining default multipliers.
As for point sources, values of the default multipliers based on local data
are preferable to the national values supplied by HATREMS.
Reentrained road dust is one category that might not have received
close attention in previous particulate inventories. Reference 29 provides
specific recommendations for reentrained road dust lead emission factors
under steady state conditions. Use of these emission factors requires that
VMTs be known. The factor is 0.03 grams/vehicle-mile in 1975-76 and is ex-
pected to be reduced by approximately a factor of three to under 0.01 grams/
vehicle-mile by 1980. For projection to 1982 and beyond, it should be noted
that this reduction is almost the same as the reduction" in the probable-puol
ed average lead content of gasoline (from ^ 1.5 to 0.5 grams/gal) projected
in Ref. 17. Further proportional reductions in the reentrainment emission
factor should occur after 1982 as the lead content of gasoline is decreased
further.
Illustrations from Example Area. Since the calculation of point source
emissions has already been illustrated, no further example calculations are
presented here for the particu"".ate point sources not included in the lead
point source inventory.
The following is an example of how Eq. 2.2 and the HATREMS area source
lead emission factors were used to calculate traditional area source lead
emissions for commercial and institutional burning of distillate fuel oil.
As shown in the baseline data in Appendix A, 1975 distillate fuel oil con-
sumption in County C amounted to 140,160 SCC units/yr. Thus,
OR = 140,160 SCC units/yr,
EF = .004 Ib lead/SCC unit from HATREMS, and
DM = .100 from HATREMS.
Using Eq. 2.2.
M = (1/2000) x (140,160) x (.004) x (.100)
or M •= .028 tons lead/yr.
-------
14
The calculation of reentrained lead emissions is also relatively
straightforward, as shown by the following example calculation. From
Appendix A, the arterial VMT in 1975 in County D was 9,429,900 vehicle-miles/
day. On a yearly basis, emissions fr6m reentrained lead dust, Q , would
be given approximately by
Qrld = 9>429>900 (VMT/day) x 365 (day/yr)
x 0.03 (g/VMT) x (1/907,200)(tons/g)
= 114 tons/yr of lead from reentrained road dust on arterial roads,
The emission factor 0.03 g/VMT comes from Ref, 29. Similar calculations can
be applied to the areterial VMTs for other counties and.also to each free-
way link in the baseline traffic data. As noted above, when, emissions are
projected beyond 1975, the emission factor must be reduced to reflect the
expected reduction in the lead content of gasoline.
Determination of Highway Source Lead Emissions.
This example plan used the values listed in the Supplementary Guide-
line (Ref. 17) for the parameters needed to calculate lead emissions from
hJThway sources. Other values may be developed. It is suggested that states
check with the EPA Regional Office to determine whether any updates of the
parameter values given in Ref. 17 have been issued.
General. There is more than one way that a highway source emission
can be constructed. Alternative procedures have been documented
by EPA. »~ However, the supplementary lead guidline17 recommends a method
based on VMT data and this method should be used whenever possible. Lead
emissions from arterials and freeways can be determined using the formulas,
data tables, and graphs contained in this reference. The basic working
equaticn is: .
(a x Pb x T)/f
n
n.s
(2.3)
where
6n s ~ emission rate for calendar year n and speed s
' (g/road mile-day),
as = percentage of lead burned that is exhausted
(expressed c,s a decimal) »
-------
15
Pb *= probable pooled average lead content of gasoline
n in year n (g/gal),
T — average daily traffic (vehicles/day),
f = average fleet fuel economy for calendar year n
' and speed s (vehicle - road mile/gal).
This form of the equation is satisfactory for calculating emission densities.
However, for most emission inventory work, emission rates are desired. By
multiplying each side by the number of road miles traveled, Equation 2.3
can be modified slightly to give the following from which uses VMT data:
E = (a
n,s s
x VMT )/f
n n n,s
(2.4)
where
~ ~ E = emission rate for calendar year n and speed s (g/day-)-j~and—
n,s
VMT = average daily vehicle miles traveled in year n (VMT/day).
In both Eqs. 2.3 and 2.4, the factor a depends on vehicle speed. EPA has
S
recommended, however, that Ref. 17 be amended regarding the percentage of
lead burned that is exhausted for stop-and-go traffic (i.e., variable driv-
42
ing conditions). For such conditions (e.g., on most arterial roadways),
a value of a = 0.70 has been recommended for use rather than a value cal-
s
culated from the Fig. 4.3-1 in Ref. 17. This value (0.70) will be used
regardless of the average route speed. For freeways, steady-state conditions
on particular links, and acceleration lanes, the speed or acceleration-
dependent values of a in Ref. 17 would still be used. Thus, there will be
s
two equations corresponding to Eq. 2.4:
and
E = (a x pb x VMT )/f
n,s s n n n,s
E = (0.70 x pb x VMT )/f
n,s n n n,s
(steady-state conditions
on individual roadways
or acceleration lanes)
(stop-and-go traffic or
long-term, varied driv-
ing conditions).
(2.5)
A similar set of equations would correspond to Eq. 2.3.
These equations are basically the same for all vehicle types. However,
for heavy-duty gasoline-powered trucks, the factor f is assumed to be in—
17
n,s
dependent of both year and speed. Thus, average daily traffic or VMT data
-------
16
should be available for heavy-duty gasoline-powered trucks as a separate item
in the basic data set. Light-duty truck and automotive emissions are cal-
culated in the same way, requiring a calculation of f for different years
and speeds (for freeways). Motorcycles and diesel-powered vehicles may be
assumed negligible lead emitters.17
In reporting emissions in a format similar to that of Appendix D, of
40 CFR 51, emissions from both individual freeway links and arterials can be
included as area sources. (The format might also be altered as was do.ne here
to include the freeway emissions as a separate line item). However, it may
be desirable to retain the locational information for those links fqr which
individual emissions rates have been calculated, thus inventorying them as
line sources. Emissions from such links could then be allocated to the
source grid for modeling with more accuracy than if they had been aggregated
with other highway emissions and allocated by using an-allocation parameter.
Such a procedure was followed in this example strategy for illustrative pur-
poses but is not necessarily required under the current regulations. Some
states may not have access to transportation data in a format suitable for
developing such an inventory. Even if data is available, states may find
that the increase in spatial resolution of highway lead emissions is not
warranted in their areas.
Illustrations from Example Area. VMT data by vehicle type were dis-
aggregated from the county-level data presented in Appendix A using figures
on percentage VMT by vehicle type obtained from transportation planning
agencies. This disaggregation was necessary because the methodology for
calculating emissions depends on vehicle type.
The following is an example of how arterial lead emissions from heavy-
duty gasoline-powered trucks were calculated for County C in the three-county
area. Since the available traffic data were presented in terms .of VMT, the
form of Eq. 2.5 with ag = 0.70 appropriate to the varied driving conditions
characterizing arterials was used. To use the equation, Pb_,5, VMT , and
f75,s mUSt also be found- P*>75 can be found from Table 4.3-8 in Ref. 17;
Pb ' - '
75 = !-9 gram/gal. The required VMT figure was disaggregated from the
data presented in Appendix A, as noted above. It was found that VMT =
1,055,690 vehicle-miles/day for heavy-duty trucks in County C. Finally,
-------
17
75,s
Then,
5.7 vehicle-miles/gal as noted in Ref. 17 for this source category.
E^,. (County C; heavy-duty gasoline-powered trucks)
= 0.70 X 1.9 (gram/gal) x 1,055,690 (vehicle-miles/day)/
5.7 (vehicle-miles/gal)
= 246,330"grams lead/day
= 99 tons lead/yr.
The example calculation demonstrates how lead emissions were calcu-
lated for heavy-duty gasoline-powered trucks, using VMT data instead of
vehicles/day. The calculation of lead emissions from passenger and light-
duty gasoline-powered vehicles is slightly more complicated although the
VMT would be handled in exactly the same manner. The difference in the .
calculation for heavy-duty trucks and the calculation for light-duty trucks
and automobiles is in the factor f and the value of Pb . The calculation
n,s n
of the term f^ g for passenger and.light-duty vehicles is rather lengthy and
is fully explained and illustrated in the EPA lead guideline document (Ref.
17). Values of Pbn for passenger and light-duty vehicles are taken from
Table 4.3-1 in Ref. 17 rather than from Table 4.3-8.
Freeways in the example area were approximated by a series of straight
line segments, or links, on a transportation map. The VMT for each link was
computed and the UTM coordinates of the end points of each link determined
. for subsequent allocation. This level of detail is not required and may not
be warranted in particular areas. Lead emissions were calculated for each
link based on VMTs and vehicle type distribution, as discussed above for
arterials. In calculating emissions from freeways, account must be taken of
the speed dependence of the factor a in Eq. 2.5. The value of a can be
s s
determined as a function of vehicle cruise speed using Fig. 4.3-1 in the EPA
guideline document (Ref. 17). Where a link was not entirely within the
example area, the lead emissions applicable to the area were determined as
that percent of the link which was in the example area. For instance, if the
total link measured 4" on the transportation map and 1" was within the example
area, then 25% of the total link lead emissions would be attributed to the
portion of the link within the example area. The freeway links for which
individual emission rates were calculated are shown in Fig. 2.2. Since
-------
18
Fig. 2.2. Freeway Links for Example Area
Line Sources
-------
19,
the calculation is rather lengthy and has been described in detail in Ref. 17,
no example calculations are presented here.
2.1.3 Example Area Lead Emission Inventory
Using the methods presented and illustrated above, an emission inven-
tory was developed for the example area. Table 2.2 summarizes the base-year
emission inventory for the entire three-county example area in the National
Emissions Report (NER) format specified in Ref. 7. This format is similar
to that specified in Appendix D of 40 CFR 51. Both Ref. 7 and Appendix D
specify reporting emissions for each county separately and this specification
should be followed in plan development and submission. However, for illus-
trative purposes, only regional totals have been presented here. The format
has been modified slightly to reflect better the emissions of lead. For
example, emissions from freeways (line sources) and arterials (area sources)
have been listed separately to give greater detail for the dominant source
category. This additional detail is not required by the current SIP regu-
lations. Table 2.2 shows the importance of vehicular emissions in the ex-
ample area. Land vehicles contribute almost 78% of the total lead emissions
directly and are responsible for another 15% through dust from reentrainment
and unpaved roads. Lead point sources contribute only 2.1% of the total.
2.2 AIR QUALITY DATA
2.2.1 SIP Requirements
EPA regulations require states to submit a summary of all lead air
quality data measured since January 1, 1974 [51.82(a)]. In addition^ all
lead air quality data measured since January 1, 1974 must be submitted to the
appropriate'Regional Office with the SIP, but not as part of the SIP
[51.86(c)]. .Data reporting procedures and formats have also been published
17 27
by EPA. ' States are required [51.82(a)] to evaluate lead air quality
data for
• Reliability,
• Suitability for calibrating dispersion models, and
Representativeness.
-------
20
o\
1
O H-
O.t"
LU Z
3 O
LU O
-------
21
4J
(3
O
O
CM
0)
t-i
I
o
•
UJ
<
_J
5
i
Z
2
Z
2
OTHER
CM
-t
O
H
H
"^
-------
22
I I
H
O
O.
-
O oi
O
O
Z'CC
O ut
O
UJ ".
s o
ii
2 SS
-------
23
4J
a
o
CM
I
a
rs
i
LU
ac
D
°
•
.
m
f^»
Csl
(-
O
1-
I
CJ
VO
«M
•
OS
fSj
i— 1
HEAVY DUT
Os
0
•
in
vO
>-
OFF HIGHW/
.
o
rH
OS
es
TOTAL
GASOLINE
ARTERIAL
m
§
o
iH
V
HEAVY DUT
-------
24
4J
O
U
O
g
00
£2 2 u.
<0
o
O CO
> <
UJ <
o
C) 0)
?! 5
4-'
1
.
Kl
CO C
"0
-------
.25
EPA has promulgated a reference method (atomic absorption spectrophotometry)
for the determination of ambient .lead concentrations.20 If a state analyzes
previously collected filters from high volume samplers, for lead, EPA will
accept data based on analysis by x-ray fluorescence and has stipulated that
other analysis methods may also qualify if approved by the Regional Adminis-
trator [51.82(b)]. EPA has also proposed other equivalent sampling proce-
dures. (See Ref. 28~fbr the detailed proposal.) States which have not used
the promulgated procedures to collect and analyze base year air quality data
should consult with their EPA Regional Office as to the appropriateness of
using such data for a lead strategy.analysis.
2-2.2 Example Area Lead Air Quality Data
Lead air quality data for the example area were obtained from three
different sources: a municipal agency, a county agency, and the EPA Regional
Office. The UTM coordinates of the sampling stations were provided along
with a description of the sampling and analysis procedures. Air quality data
from the EPA Regional Office were obtained from EPA's Storage and Retrieval
of Aerometric Data (SAROAD) system. In all cases, lead sampling was indi-
cated as being done by the high volume air sampling procedure utilized for
sampling total suspended partlculate matter. The sample analysis procedure
was EPA's promulgated reference method.
Data Reliability
An in-depth determination of data reliability was beyond the scope of
this work. No inspection cf sampling sites in the example area or sample
analysis laboratories was performed. Discussion with the cognizant agencies
indicated, however, that there was substantial compliance with EPA sampling
and analysis procedures and that appropriate quality assurance measures were
taken. States are, however, expected to report deviations from promulgated
sampling and analysis procedures to the EPA Regional Office.
Suitability for Calibrating Dispersion Model
A limited discussion of the suitability of air quality data for cali-
brating dispersion models is given in Ref. 16. When possible, the air quality
data must be from the same base year as the emission inventory and must also
-------
26
be for the same averaging time as the standard. The lead ambient air quality
standard is 1.5 yg/m3 arithmetic mean averaged over a calendar quarter.
Monthly arithmetic mean air quality-data were received from the local agency
and averaged to give calendar quarter arithmetic means.
The spatial distribution of sites with air quality data is also im-
portant. Since states are required to calculate and locate, using dispersion
models, maximum lead concentrations in the vicinity of specified significant
point sources, it is desirable to have measured lead air quality data in the
vicinity of such sources. Measured air quality representative of significant
vehicular activity such as major freeways which have high VMT is also desir-
able. States, of course, will have to rely on data from their existing mon-
itoring network. The lack of suitable lead sampling coverage in areas deter-
mined to be potential hot spots as a result of lead emission analysis fre-
quently cannot be remedied unless previously collected high-volume filters
can be analyzed for lead.
Representativeness of Lead Air Quality Data
The measured lead air quality data should be representative of recep-
tor exposure. Measures of representativeness include:
• Height of the sampling station^, and
• Placement of the sampling station with respect
to both receptors and sources.
Guidance is available in Reference 17 to aid in evaluating monitoring site
location. EPA is also developing guidance describing network design and
siting criteria that will eventually appear in regulatory form. As currently
required, stations used for ambient lead monitoring should include at least
one roadway site and one neighborhood site [51.17(b)]. Roadway sites are
expected to show the highest ambient concentrations. Neighborhood sites, in
neighborhoods with poor lead air quality would give a picture of population
exposure in high population density areas. Reference 17 also provides
special treatment for the location of monitors at street canyon sites in
downtown areas. Street canyon sites usually combine high traffic and high
population density. Additional details are discussed in Ref. 17.
-------
27
The sites of the lead sampling stations- in the example area were not
inspected. Conversations with local agencies indicated that many of the
sampling stations are located on the roofs of schools. Lead sampling station
data obtained from the Regional EPA Office were in SAROAD format and the
elevations of the stations above ground level were specified. Of the two
SAROAD sampling stations, one station was 23 feet (7 meters) and the other
105 ft. (32 meters) above ground level. Reference 17 specifies that a.lead
sampler be placed at a height no greater than 5 meters above ground level.
If height criteria are exceeded, the air quality values near ground level
can be expected to be higher than the measured values. Although a 5 meter
maximum lead sampler height has been proposed, it is not applicable to base-
year sampling monitors because all the base-year data is needed for SIP
development. The height specification is indicative, however, of a deter-
mination by EPA regarding the collection of representative lead air quality
data as well as for the establishment of new sampling sites.
Although many of the rooftop sites probably exceed the proposed height
criterion, data from all these sites and the data from both SAROAD sites were
included in the baseline data set. States should consider the height of their
base-year lead sampling stations, especially for stations in potential hot
spots and, in cases where doubt exists, discuss with the EPA Regional Office
the possibility of poor representation of lead air quality as a result of
sampling station height.
Even though the height criterion was not applied in this example
strategy, not all lead sampling station data were used. A check of the
sampling station location by plotting the station UTM coordinates revealed
that some stations had erroneous UTM coordinates. Where UTM coordinate
corrections could not be made in a timely manner, the air quality data were
not used.
In the example area, lead air quality data were available from 39
verified sampling stations. Unfortunately 38 of these stations were in
County C. One sampling station was in County L, and there were no base-
year lead sampling stations in County D. Figure 2.3 shows the location of
each lead sampling station in the example area. This figure clearly illus-
trates the lack of spatial coverage.
-------
28
Fig. 2.3. Locations with Available Lead
Ambient Air Quality Data
-------
29
Determination of Background Lead Concentration
A discussion of background lead concentrations can be found in Ref. 17,
EPA recommends that states assume a background equivalent to concentrations
of airborne lead in a "representative nonurban area that is not significantly
influenced by stationary or mobile lead sources." After consultation with
EPA, measured nonurban^lead air quality data from outside the example area
were evaluated to estimate a background concentration. Since the range of
the data was only 0.08 yg/m3, it was decided to use an arithematic average of
all the outlying nonurban data as a reasonable approximation to background
lead air quality. The background ambient lead air quality level was deter-
mined to be 0.15 yg/m3. States may wish to consult with the EPA Regional
Office regarding the choice of a background ambient concentration for lead
in cases where doubt exists concerning validity of a determination or where
suitable data is unavailable.
Selection of Calendar-Quarter Base-Year Lead Air Quality Data
For lead, a calendar quarter of air quality data must be chosen as the
air quality baseline for strategy development and modeling rather than a full
year as is the case for particulates. Little specific guidance is available
as to how to choose the most appropriate quarterly lead air quality data.
For the example strategy, the choice was made based on inspection of the base
year measured data. There wei> no gross violations of the lead standard
measured in 1975 in the example area. The second calendar quarter data was
chosen as the baseline quarter because it contained the highest average lead
concentration (1.600 yg/m3} among all the sampling stations as well as the
greatest number of stations that measured lead concentrations greater than
1.0 yg/m3. Table 2.3 lists the 1975 air quality data for lead by calendar
quarter for the example area.
In general, the considerations involved in selecting the baseline
quarterly data should include, but not necessarily be limited to:
• The quarter in which the maximum quarterly average
lead concentration was observed;
• The existence of particularly large violations of
the standard in one quarter, say by a factor of
more than about two or three; and
-------
30
Table 2.3 Base-Year (1975) Lead Air Quality Data
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
.340
. .709
.540
.700
.826
1.075
.672
.470
.695
.521
.543
.601
.360
.577
.436
.625
.833
.800
1.067
1.200
1.133
.867
.733
.633
.733
.767
.867
.667
.633
,667
.733
.867
.767
.400
.667
.767
.501
1.089
.519
.859
1.093
1.206
.784
.617
.769
.669
.774
.627
.563
.518
.729
.875
1.100
.867
1.000
1.233
.933
1.067
.733
.733
.700
.667
.967
.500
.367
.733
.700
.400
1.167
1.600
.367
.700
.633
.433
.822
.562
1.250
. 1.287
.865
.566
.777
.379 •
.884
.685
.487
.401
1.059
1.007
.367
.833
.700
1.367
.500
.333
.700
.633
.500
.667
.433
.667
.633
.600
.300
.967
.733
.667
333. .
.203
.701
.47.7
.666
.700
.679
.486
.310
.536
.298
.467
.412
.345
.256
.507
.853
.667
.733
.633
.700
1.067
.933
.800
.467
.767
.533
.533
.533
.367
.400
.567
.300
.400
.867
.800
.467
,.900
Based on average of monthly means obtained
from local agencies and SAROAD.
-------
31
• The occurrence of the largest number of standard
violations and/or the largest number of exceedences
of some other relatively high concentration (such
as the 1.0 yg/m3 value used in this example strategy).
Choice of the quarter in which the maximum quarterly average was observed is
consistent with current EPA policy for other pollutants. The other two con-
siderations might thus assume less weight in choosing a baseline quarter and
serve primarily to confirm the choice based on the observed maximum. In the
abscence of specific guidelines, consultation with the EPA Regional Office
is advised in cases of doubt.
2.3 METEOROLOGICAL DATA
A discussion of meteorological data in general can be found in Ref. 16,
References 12 and 31 discuss the meteorological data requirements of dis-
persion models; Reference 30 discusses those of the various modified roll-
back models. In general, the development of baseline meteorological data
for lead is subject to the same considerations as the development of such
data for particulates. The major difference is that quarterly rather than
annual data is appropriate for lead.
Selection of. Baseline Quarterly Meteorological Data
Once the baseline quarterly air quality data have been selected, the
baseline meteorological data must correspond to the same year and quarter to
retain consistency between the two data sets. Use of data from .different
calendar quarters would not be appropriate.
Selection of Quarterly Meteorological Data for Air Quality Projections
Little guidance exists concerning the choice of worst case meteoro-
logical data for projection purposes. Reference 16 discusses some of the
important considerations. If time and resources permit, a sensitivity anal-
ysis comparing modeled air quality concentrations for meteorological con-
ditions representing data from several quarters while keeping the emissions
inventory constant may aid in determining worst case meteorology. No- more
than five years of data would need to be examined to make a reasonable de-
31
termination. However, it is unlikely that this amount of effort will be
-------
32
feasible in many situations. Also, data may not be available for such a long
time span. Beference 31 provides some guidance in attempting to determine
worst case meteorology for estimating future pollutant concentrations. In
any event, an air pollution meteorologist should be consulted to aid in the
determination of the worst case conditions to be used in estimating future
lead air quality.
One year (1975) of data was readily available for the example area in
the form of monthly (STAR) stability wind roses obtained from the National
Climatic Center (NCC), Federal Building, Asheville, North Carolina'28801.
As discussed above, the second quarter of 1975 appeared to represent a worst
case from an air quality perspective. A quarterly average stability wind
rose for the second calendar quarter was 'obtained by averaging the monthly
wind roses for April, May, and June. This averaging procedure will probably
be necessary if NCC data is used because calendar quarter stability wind
roses-are apparently unavailable; only seasonal quarter (for example March,
April, May) wind rose_s appear to be available. Users should contact NCC if
questions arise concerning STAR summaries.
-------
33
3 PROJECTING FUTURE EMISSIONS
Current regulations [51.81(b)] require that projections of lead emis-
sions be made for at least three years beyond the date by which EPA must
approve or disapprove the lead SIP (five years if an extension under 1-10(e)
of the Clean Air Act has been granted). As shown in Table 3.1. the schedule
requires that lead emissions be projected for the fourth quarter of 1982
(1984 if an extension has been granted). States, of course, are encouraged
to demonstrate attainment earlier if possible. Also, unless quarterly vari-
ations are being taken into account, the projection to the attainment year
can be based on annual rather than quarterly estimates in the same way as
can be done for the baseline inventory.
3.1 SOURCES OTHER THAN HIGHWAY VEHICLES
3.1.1 General •
This section describes the projection of emissions from point sources,
including industrial fugitives, and traditional area sources. Projection of
emissions associated with vehicular sources, including reentrained road dust,
is described in Sec. 3.2.
Table 3.1. Lead Analysis Time Period Requirements
. Date
Action
Time Interval
Oct. 5, 1978
July 5, 1979
Nov. 5, .1979
Nov. 5, 1982(84)
Lead NAAQS promulgated.. -
Lead SIP due. 9 mos.
Approval/Disapproval of SIP by EPA. 4 mos.
Demonstration of atfafiiment without 3 (5) yrs.
(with) extension.
-------
34
The techniques used to project lead emissions from point and area
sources are the same as would be used to project particulate emissions from
those source categories and states should already be familiar with them.
References 7 and 16 discuss the available methodologies and data sources that
can be used in making such projections. In general, projections of variables
such as process weight which are closely related to emissions are unavailable
and, as described in the references, recourse must be had to surrogate vari-
ables such as population o~r employment to project future emissions. Of
course, in cases where source-specific information is available as with a
scheduled startup or shutdown, this information should be used rather than
the projections based on surrogate variables. In making the projections,
care murt be taken to use the appropriate lead, not particulate, emission
factors and to account for any known difference in control efficiencies for
lead and particulates just as was done when developing the baseline lead
inventory.
In projecting future emissions, one source of difficulty is. disag-
gregating the projected capacity or emissions between new/modified and
existing sources. Disaggregation is necessary because different regulations
frequently apply to sources in the two categories. Reference 16 presents a
method for projecting emissions from new/modified and existing sources in the
absence of source retirement and assuming full capacity utilization. Refer-
ence 15 presents a method which accounts for both retirement and fractional
capacity utilization. Both references 'present equations based on compound
growth. For linear growth with retirement but assuming full capacity utili-
zation in the base year, the model presented in Appendix B of this work may
be used.
Projections of future emissions should account for the following
federal programs:
• New Source Performance Standards (NSPS),
• Fuel switching required under the '-Energy Supply and
Environmental Coordination-Act of 1974 (ESECA).,
• Particulate SIP revisions in nonattainment areas and
air quality maintenance areas (AQMAs), and
• Prevention of Significant Deterioration (PSD).
-------
35
Each of these programs has the potential to affect lead emissions. NSPSs
generally require more stringent controls on new and modified (growth and re-
placement) sources than are required under current state regulations. If
state regulations for new/modified sources are more stringent, then they
should be used rather than NSPS. Switches from gaseous or liquid to solid
fuels will usually be_ accompanied by' an increase in both lead and particulate
emission rates. In nonattainment areas, reasonably available control technol-
ogy (RACT) for particulates should be operating on existing sources by 1982
with a concommitant reduction jm lead emissions. In addition, new and modified
sources in these areas may be subject to the stringent lowest achievable
emission rate (LAER) control requirements. Some jurisdictions may also have
implemented AQMA strategies for particulates by 1982; the impact of these
strategies on lead emissions should be assessed when estimating future emis-
sions. Finally, PSD regulations require best available control technology
(BACT) on certain new sources as well as review of the sources' incremental
particulate air quality impact. Insofar as these regulations affect future
particulate emissions, they may also affect projected lead emissions. The
projection of future lead emissions should be based on actual expected emis-
sion rates rather than compliance rates if it is expected that the two will
differ. Thus, the potential impact of delayed compliance orders (DCOs)
should also be considered when emissions are projected.
3.1.2 Illustrations from Example Area
As a first strategy, point source projections for the example area
were made without considering the emission impact of nonattainment SIP
revisions. Promulgated NSPS were applied to growth and replacement activity.
These projections are presented in this section, because, as discussed in
Sections 5 and 6, the indications were that additional stringency was not
required to demonstrate attainment of the lead NAAQS. In developing plans,
however, states should estimate future'emissions as accurately as practi-
cable and take credit for all control programs likely to affect lead emis-
sions. Such credit may be particularly important for significant point
sources where modeling is required. For example, Plant A, the largest point
source in the inventory, was out of compliance with existing regulations in
1975 and is expected to come into compliance before 1982 with a significant
reduction in lead emissions.
-------
Mission proj ections for sources, other, than highway, vehicles used the.
following data:
• Employment: projections,
• Population projections, and
• Estimates of'replacement rates.
Employment, projections were available.at the two-digit Standard Industrial
Classification,(SIC)32 leyei ftom, Ref. 33; Thfese estimates were county-
specific. Industrial process emissions projections were thus done at a
level' between the: level' 11 and' level 2 described in Refs. 7 and l,6v States,
of course, should pick the-level,appropriate, to, the severity of their lead'
problems, the available data, and resource constraints, in consultation, with
the, EPA: Regional Office. Some,, plant-specif ic, data were' available for^ local
power plants and was used,, in a. level'. 3 projection. Emissions; fromvo-ther
power plants were handled, in the,-same way as industrial process-emissions.
In deciding which, if any, plants to interview in level 3 projections/ states
should be aware--that significant lead sources may not always be those which
are large particulate sources. Hence, some change in the source categories
traditionally treated at a high level of detail may be in order. Table 3.2
presents the employment growth rates used to make emission projections in
the example area. These rates were,obtained by assuming linear growth'be-
tween the 1975 and 1995.employment levels projected in Ref. 33,
Population'projections were available at the county level from Ref. 35.
Table 3.3 presents the-population growth rates for the example area. These
data assume.linear growth'.between the 1975 and 1990 population levels-pro^
jected in the reference. The area source categories to which the population
surrogate was. applied are also Indicated. With.locally available, county-
specific data, the level of projection corresponds, to about level 2. States
should make the same considerations, as noted: above in dealing, with these
categories.
Table 3.3 also presents the growth rates used in projecting future
emissions from other area source categories. These rates are averages of
the rates for industrial, commercial/institutional, and agricultural SICs,
from Ref. 33. In plan development, states will need to use methods tailored
to the available data as described in.Refs. 7 and 16.
-------
37
Table 3.2 Employment Growth Rates for
Projecting Point Source
Emissions3
Annual Percentage
Employment Growth
(%/yr)
2-Digit
SIC
99 .
07
14
15
24
25
32
33
34
35
36
37
38
39
20
. 22
23
26
27
28
29
30
31
40,46,42
49
50,51
53,58
65
86,69
72
73,75
80,82,83
92
Based on
— ^— •
Industry
Nonclassifiable
Agricultural Services
Mining •
General Building
Wood Products
Furniture
Stone, Clay, Glass Products
Primary Metals
Fabricated Metal Products
Machinery
Electrical Equipment
Transportation Equipment
Instruments
Miscellaneous Manufacturing
Food
Textile Mill Products
Clothing
Paper
Printing and Publishing
Chemicals
Petroleum and Coal Products
Rubber and Plastic Products
Leather Products
Railroads, Trucks, & Pipelines
Electric, Gas, Sanitary Services
Wholesale
Stores, Restaurants
Real Estate
Organizations & Miscellaneous
Personal Services
Business and Auto Services
Schools, Health, & Social Services
Public Safety
data in Ref. 33.
C
0.91
0.74
^-2. 09
0.44 .
0.76
0.83
1.28
0.35
1.44
0.83
0.90
0.72
1.96
0.61
0.08
0.09
-1.27
1.06
0.61
1.10
0.83
2.46
-0.60
6.44
O.R2
0.67
. 1.04
1.15
1.80
0.59
1.65
2.36
-0.45
County
D
0.90
0.54
-2.08
0.44
0.77
0.85
0.95
0.30
1.44
0.63
0.72
-0.11
1.81
0.60
0.12
0.23
• -1.11
1.12
0.60
1.20
0.37
2.60
-0.65
0.85
0.59
0.67
1.04
1.16
1.56
0.59
1.61
2.06
-0.45
L
1.06
0.59
-2.10
0.44
0.72
0.72
1.38
0.28
1.39
1.04
1.30
0.39
2.68
0.61
0.10
0.91
-1.30
1.18
0.56
1.15
0
2.32
-0.71
0.84
1.03
0.64
1.04
1.15
1.81
0.59
1.62
2.46
-0.04
-------
38
Table 3.3 Growth Rates for Projecting Area Source Emissions'
Source Category
Projection
Parameter
Growth Rate
(%/yr)
D
Residential fuel combustion
Off-highway vehicles
Residential incineration
Commercial/Institutional fuel
combustion
Commercial/Institutional
incineration
Industrial fuel combustion
Industrial incineration
Unpaved roads
Land tilling
Population
0,15
C/Iemp.,
0.95
Mfg. emp.
0,83
VMT 1,00
Agri. emp. 0.74
2.79
0.96
0.31
ti
1.00
0.54
3,63
1.08
1.09
1.00
0.59
aBased on data in Refs. 33 and 35.
Commercial/institutional employment rates are averages of rates
for SICs related to transportation, commerce, and public
utilities; trade; finance, insurance, and real estate; services-
and civil governments. '
Manufacturing employment rates are averages for SICs in manu-
facturing categories 20-39.
Based on average predicted VMT growth in 8-c.ounty transportation
study area which includes example area.
-------
39
Finally, .the replacement rates used in projecting point source emis-
sions are given in Table 3.4. These rates were developed for specific SICs
from data in Refs. 36 and 37.
Illustrative Point Source Lead Emission Projection
Note: The details of the emission projections depend upon the emis-
sions projection model chosen. The example presented here makes use of the
equations derived in Appendix B. Review.of these equations is suggested
prior to reading this subsection. States may, of course, use another appro-
priate model for projecting emissions. :
The projection of lead point source emissions is illustrated here for
Plant E in Table 2.1. The lead emitting source category in this plant is
SCC 30400303, electric induction furnaces. Figure 2.1 shows that Plant E
is located in County C. The inventory indicates that the plant's SIC is
3322, a malleable iron foundry. The projections is illustrated for two
cases:
1. The case actually modeled in the example strategy
for which no additional controls are applied, and
2. A hypothetical case in which additional controls
are assumed to be required on new and modified
electric arc induction furnaces and retrofits are
required on existing furnaces.
The second case is included to illustrate the manner in which the phase-in
of control programs is handled.
Case 1' In thls case source activity associated with existing, mod-
ified, and new capacity or activity would all be subject to the same 1975-
level control requirements. Using the growth model presented in Appendix B,
this condition can be expressed mathematically as n = n =n = n
eo 'en 'gi \f
Since the regulations for new/modified sources do not change during the pro-
jection period, T can be set equal to either tQ or t^ Under these condi-
tions Eq. B.10 on p. 98 reduces to
-------
40
Table 3.4 Replacement Rates2
Standard
Industrial
Classification
(SIC)
0723
0759
1421
1422
1441
, 1442
1541
2011
2013
2024
2032
2033
20'4i
2042
2043
2046
2048
2071
2083
2085
2086
2092
2096
2098
2099
2231
2261
2291
2295
2311
2399
2431
2434
2499
2511
2512
2514
2522
2542
2621
2631
2641
2642
2643
2644
2645
2647
2649
2651
2652
2654
2655
2661
2699
2751
2752
2757
2793
2799
2813
2816
2818
2819
2821
2824
Fractional
Replacement
Rate Cyr"1)
.037
.039
.055
.05
.05
.05
.025
.03
.04
.03
.03
.03
.037
.037
.037
.037
.03
.03
.04
.032
.04
.008
.03
.03
.03
.042
.042
.042
.042
.042
.042
.031
.031
.031
.039
.039
.039
.039
.039
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.031
.045
.045
.045
.045 '
.045
.045
0
.045
.045
.021
.035
Gravel Handling & processing average
Pathological incinerators
Sand, gravel & stone quarrying & processing
Lime processing
Sand and gravel mining
Sand and gravel mining
Cement plants
Average of all food related values
Meat smoke houses
Average of all food related categories
Average of all food related categories
Average of all food related categories
Grain -handling and processing
Grain handling and processing
Grain handling and processing
Grain handling and processing
Average of all food-related categories
Average of all food related categories
Beer processing
Whiskey processing
Beer processing
Fish processing
Average of all food related categories
Average of all food related categories
Average of all food related categories
Textile processing
Textile processing
Textile processing
Textile processing
Textile processing
Textile processing
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Average value for selected manufacturing categories
Average value for selected manufacturing categories
Average value for selected manufacturing categories
Average value for selected manufacturing categories
Average value for selected manufacturing categories
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Wood processing - pulpboard
Graphic arts
Graphic arts
Graphic arts
Graphic arts
Graphic arts
Average for chemicals & products
Lead pigment
Averaged over all chemicals & products
Averaged over all chemicals & products
Averaged over all resinous chemicals
Averaged over all synthetic fibers
-------
41
Table 3.4 (Cont'd)
Standard
Industrial
Classification
(SIC)
2833
2834
2841
2843
2844
2845
2851
2861
2865
2871
• 2873
2891
2892
2893
2899
2911
2951
2952
2992
2999
3069
3079
3111
3132
3141
3211
3221
3229
3241
3250
3251
3271
3272
3273
3274
3275
3281
3292
3293
3295
3297
3310
3312
3316
3317
3321
3322
3323
3325
3334
3339
3341
3351
3352
3356
3361
3362
3369
3391
3394
3399
3411
Fractional
Replacement
Rate (yr"1)
.045
.045
.022
.045
.045
.045
.065
.045
.033
.032
.031
.045
.045
.045
.045
.031
.05
.042
.031
.031
.003
.022
.039
.039
.039
.033
.033
.033
.025
.033
.03
.03
.03
.03
.05
.006
.06
.033
.033
.033
.045
.028
.028
.028
.(28
.028
.028
.028
.028
.035
.028
.032
.035
.032
.033
.036
.035
.035
.03
.03
.03
0
Comments
Fharamaceuticals
Pharamaceut icals
Average of soap and detergents
Average for chemicals & products
Average for chemical manufacturing
Average for chemical manufacturing
Paint and varnish
Average for chemical manufacturing
Average between malalc and phthalic anhydride
Average of all agricultural chemicals
Average of all fertilizer chemicals
Average for all chemical manufacturing
Explosives
Printing ink
Average for all chemical manufacturing
Petroleum refining
Asphalt roofing
Asphalt roof saturating & blowing
Miscellaneous petroleum industry point sources
Miscellaneous petroleum industry point sources
Rubber process industry
Average of all synthetics
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Glass processing
Glass processing
Glass processing
Cement plants
Brick & related clay products
Average of concrete & brick industries
Average of concrete & brick industries
Average of concrete processes
Average of concrete processes
Lime processing
Gypsum processing
Stone quarrying and processing
Average of all mining categories
Average of all mining categories
Average of all raining categories
Average for chemicals & products
Iron & steel plants
Iron & steel/blast furnaces
Iron & steel plants
Iron & steel plants
Iron & steel plants
Iron & steel plants
Iron & steel plants
Iron & steel plants
Primary aluminum smelters
Ferro alloy
Average of nonferrous secondary metals
Average of copper & brass handling, refining,
and smelting
Average of non-ferrous secondary metals
Average of lead, brass, magnesium & zinc
secondary processes
Aluminum production
Average of brass, bronze, all copper operations
(secondary metals)
Average of brass, bronze, all copper operations
(secondary metals)
Primary iron & steel
Primary iron & steel
Primary iron & steel
Can Manufacturing
-------
42
Table 3.4 (Cont'd)
Standard
Industrial
Classification
(SIC)
3429
3433
3441
3452
3461
3462
3466
3469
3471
, 3479
3481
3489
3491
3493
3496
3499
3519
3522
3531
3536
3537
3541
3542
3544
3545
3553
3555
3558
3561
3562
3563
3571
3579
3584
3585
3594
3599
3612
3613
3621
3631
3621
3631
3634
3634
3635
3641
3642
3651
3661
3662
3679
3694
3711
3714
3731
3741
3742
3743
3751
3799
3822
3823
3825
3842
3843
3861
Fractional
Replacement
Rate (yr""1)
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.045
.045
.039
.02
.039
.039
.04
.03
.04
.04
.04
.039
.039
.039
.039
.04
.045
.045
.039
.039
.039
.039
.039
.04
.04
.04
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.04
.04
.04
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
.039
Comments
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Average of selected manufacturing categories
Explosives
Explosives
Average of selected manufacturing categories
Internal combustion engines (spark-ignition)
Average of selected manufacturing categories
Average of selected manufacturing categories
Assembly plants
Average of food related processes
Assembly plants
Assembly plants
Assembly plants
Average of select manufacturing industries
Average of select manufacturing industries
Average of select manufacturing industries
Average of select manufacturing Industries
Assembly plants
Printing ink
Printing ink
Average manufacturing
Average manufacturing
Average manufacturing
Average manufacturing
Average manufacturing
Assembly plants
Assembly plants
Assembly plants
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
Assembly plants
Assembly plants
Assembly plants
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
General manufacturing average
-------
43
Table 3.4 (Cont'd)
Standard
Industrial
Classification
(SIC)
3931
3941
3943
3949
3983
3988
3993
4013
4226
4612
4613
4911
4923
4925
4931
4941
4953
5074
5092
5093
5099
5153
5171
5191
5399
5812
6512
6513
7211
7216
7391
7399
7534
7535
8062
8211
8331
8321
8661
8911
9223
9999
Fractional
Replacement
Rate (yr"1)
.039 ' •
.039
.039
.039
.039
.039
;039
.039
.02
.031
.031
.06
.04
.04
.037
.04
. .038
.039
.039
.039
.039
.037
.042
.03
.039
.028
.055
.055
.055
.05
.039
.036
.039
.042
.039
.037
.037
0
0
0
.037
.039
Comments
General manufacturing average
General manufacturing a erage
General manufacturing a erage
General, manufacturing a erage
General manufacturing a erage
General manufacturing a erage
General manufacturing average
General manufacturing average
Small boilers
Miscellaneous petroleum industry point sources
Miscellaneous petroleum industry point sources
Large fossil fuel boilers
Small to medium size boilers
Small to medium size boilers
Small boilers
Small to medium size boilers
Average for various types incineration
Miscellaneous manufacturing average
Miscellaneous manufacturing average
Miscellaneous manufacturing average
Miscellaneous manufacturing average
Grain handling & processing
Nonpipeline petroleum transfer
Average for food industry
Average miscellaneous manufacturing
Deep fat frying
Sand and stone mining
Sand and stone mining
Sand and stone mining
Dry cleaning
Average of miscellaneous manufacturing categories
Average of miscellaneous manufacturing categories
Average of miscellaneous manufacturing categories
Industrial surface coating
Pathological incinerators
Small to medium size boilers
Small to medium size boilers
No information available for minor category
No information available for minor category
No information available for minor category
Small to medium size boilers
Average of all manufacturing categories
on data in Refs. 9 and 10.
-------
44
QQ x [1-RR x
- t )] = Emissions from sources
existing at t and re-
maining at t^°(tons/yr)
and
n
Q^ = Q^ x (t^ - tQ) (RR + GR) = Emissions from sources
that were new or modi-
fied between t and t
(tons/yr) ° n
where GR is the growth rate and RR is the rate of replacement. Table 3.4
gives the replacement rate for SIC 3322 as RR = 0.028; Table 3.2 gives the
growth rate as GR = 0.0035. For. the projection period, t = 1975 and t =
O Tl
1982. Table 2.1 gives QQ = 2.8 tons/yr where the fugitive and stack com-
ponents can be added together because .the same conditions apply to each
component. Using these values in Eq. 3.1,
(3.1)
Q82 (E) = 2-82
- -028 x (1982 - 1975)] =2.27 tons Pb/yr
Total plant emissions in 1982 due to existing capacity
and Qg2 (E) = 2.82 x (1932 - 1975) x (.028 + .0035) = 0.62 tons Pb/yr
= Total plant emissions in 1982 due to new/modified
capacity.
Thus, the emissions for the year 1982 from Plant E (existing plus new and
modified capacity) are projected to be 2.89 tons of lead/yr.
Case 2- This case was not actually used in developing the example
strategy but illustrates how emissions might be projected when retrofit con-
trols are required and more stringent controls are required on new/modified
sources. The available inventory shows that the sources in question were
totally uncontrolled (n^ = 0.00) in the base year 1975. It is assumed here
that the strategy being simulated would call for the retrofit of all exist-
ing sources with 93% controls on stack emissions and 90% controls on fugitive
emissions by 1982. Furthermore, new/modified sources would be required to
control stack emissions by 97% beginning in 1980 and fugitive emissions by
95% beginning in 1981. Prior to these years, new/modified sources have been
subject to the proposed retrofit regulations.
Since the control requirements and effective dates for the stack and
fugitive, emissions are different, the projections for each type of emission
-------
45
must be made separately. The calculation is summarized below using the
notation of Appendix B. For stack emissions :
t = 1975, T = 1980, t = 1982;
n
eo
0, nen = 93, ng± = 93, and r,gf = 97
where r)eQ, nen> n ., and n are the control efficiencies required, respec-
tively, on sources existing in the base year 1975, existing sources not re-
placed by 1982 but required to retrofit, new/modified sources constructed
between 1975 and 1980, and new/modified sources constructed between 1980
and 1982. From Eq. B.12 (p. 98) ,
_ 1 - 93/100
Re ~ 1 - 0/100 " °'°7'
R± = 0.07, and R
0.03.
Re is t1ie ratio °f emissions per unit capacity of retrofitted-sources—in—1982
to emissions per unit capacity of sources under base year controls. R. and
Rf are similar ratios for new/modified sources constructed between 1975 and
1980 and those constructed between 1980 and 1982, respectively. Substituting
in Eq. B.10 (p. 98),
Qg2 (E-stack) = 1.41 x 0.07 x (1 _ .028 x 7) = Q.08 tons Pb/yr
= Total plant stack emissions in 1982 due to
existing capacity.
Qnm (E-stack) = 1.41 x 0.07 x 5 x (.028 + .0035)
+ 1.41 x 0.03 x 2 x (.028 + .0035)
= O.C2 tons Pb/yr
= Total plant stack emissions in 1982 due to
new/modified capacity.
For fugitive emissions:
t = 1975, T = 1981, t = 1982;
n
n
eo
= 0, nen = 90, ng± = 90, ngf = 95; and
Re = 0.10, R± = 0.10, Rf = 0.05.
-------
^82
(E-fu'gitive) = 1..41 x 0.1.0' x (l -
= 0.11 tons Eb/yr
.,0'28 x 7)
Total plant fugitive emissions; in 1982
due to existing capacity.
_nm
(E-fugitive) =--1.41 x 0...10, x 5 x C.02& +
..0035.)
+ 1.41 x o.05 x l x (.028 + .0035)
= 0.03' tons Pb/yr
= Total plant fugitive emissions to; 19.8.2
due to new/modified capacity.
In total, Qg2 (E) = Q® (E-stack) + qj (E-fUg) =0.19 tons Eb/yr, and
rim
^82
0>05 to'ns pb/yr- The total projected emission in this
case for
Plaril: E is 0.24 tons/yr, substantially less than the 2.89 tons Eh/yr pro-
jected in the example plan (Case 1). Although the projections are made on
a source-by-source basis, a given source would either be completely re-
placed or a new one would be added to the inventory. The projection pro-
cedure gives & reasonable estimate of total future emissions
arid for a particular source such as Plant E, the 1982 emissi
substantially different from those projected by the
or the methodologies referenced above. Lacking source-specific data, these
methods do, however, give a reasonable way of projecting total future emis-
from all
ns could
methodology us
sources
be
here
sions.
Illustrative Point Source Fuel Conversion
No instances of fuel switching under ESECA were identified in the
example area. The potential impact of fuel switching is illustrated by the
following hypothetical example.
Plant J in County L is an electric utility with large oil-fired boil-
ers (SCC 10100401). Interview results34 indicated that Plant J should burn
about 69,806,000 gal/yr of residual oil with a heating value of 150,000 Btu/
gal to 1982. The emission inventory shows an average of 94% control on par-
ticulate emissions from these boilers. Emissions can be calculated using
Eq. 3.1. The SCC units are 103 gal so
-------
47
OR = 69,806 SCC units/yr,
EF = 0.004 Ib/SCC unit from HATREMS,
DM = 1.00, the default multiplier from HATREMS, and
W = 94.
Thus, QgJ1 = (1/2000) x 69,806 x .004 x l x [1 - (94 x 100/100)/100]
= 0.0084 tons Pb/yr
= Total lead emissions from oil combustion
at Plant J in 1982.
Particulate and lead controls have been assumed equal (Y = 100) , information
to the contrary being unavailable. It should be noted that the emissions
from these boilers are too small to have been included in the point source
inventory given for Plant .J in Table .2.1.
However, conversion to coal would lead to an increase in lead emis-
sions. The bituminous coal currently being burned in Plant J has a heating
value of 24.5 x lo6 Btu/ton. The amount of coal required to produce the
same amount of heat as supplied by the oil gives the annual operating rate.
This value OR is given by
OR = 69.806.000 gal/yr x 150.000 Btu/gal /.jT
c 24.5 x 10b Btu/ton X l.~
= 5.13 x 105 tons coal/yr in 1982.
The factor (.6/.5) accounts for an expected decrease in the thermal effi-
ciency of the boilers from 60% to 50% as a result of the fuel conversion.
The appropriate SCC for large bituminous coal-fired boilers is assumed to be
10100201. Applying Eq. 3.1 again to calculate emissions and assuming no
change in control efficiency,
..coal
^82
(1/2000) x 5.13 x lo5 x .0018 x-9.00 x [l - (94 x loo/loo/loo]
..= 0.249 tons Pb/yr.
Lead emissions at Plant J would thus increase by a factor of about 30 due to
the coal conversion. Although'the lead emissions are still fairly small on
an absolute scale, the conversion would increase Plant J's total lead emis-
sions by over 20%. Analysis of similar situations should be of aid in iden-
-------
48
tifying potential problem areas where large increases in lead emissions
might cause source-specific NAAQS violations.
Illustrative Area Source Lead Emission Projection
This illustration uses the same example begun in Sec. 2.1.2, that of
commercial/institutional distillate fuel use in County C. No controls are
anticipated to be required on commercial/institutional fuel combustion given
the small size and dispersed nature of the sources. Replacement need not be
considered, as old and new units will probably have almost identical emission
characteristics and emissions can be projected considering only growth. In
cases where different conditions exist as when many units with higher thermal
efficiencies are expected to be installed such conditions should be accounted
for in projecting future emissions. In the example considered here, Eq. B.10
can be put in the form -
QQ x [1 + GR x
- tQ)] and
82
.028 x (1 + .0095 x 7) = :0.030 ton Pb/yr
Where the growth ^rate GR was taken from Table 3.3 and 'QQ came from the ex-
ample calculation in Sec. 2.1.2.
3.2 HIGHWAY VEHICLE RELATED SOURCES
General
Two categories of sources are considered here: the tailpipe emissions
from highway vehicles and lead from reentrained road dust. In projecting
emissions from these categories the following programs must be taken into
account:
• Federal programs for the reduction of the lead content
of gasoline and the requirements for the use of lead-
free gasoline in catalyst equipped vehicles,
• Federal requirements for improved fuel economy,
• Any transportation system management (TSM) elements
such as employer carpool incentive programs or traffic
flow improvement programs that will be instituted be-
fore the attainment date for the lead NAAQS, and
-------
49
. • Any other transportation control plan (TCP) measures
which reduce VMT's or improve traffic flow so that
vehicular lead emissions would be reduced;
In many areas lead emissions caused by highway vehicles are likely to
be the largest part of the inventory and hence it is reasonable to place ,
.considerable emphasis on estimating future emissions from these categories.
Projection of Highway Vehicle Lead Emissions
The emissions under steady-state, acceleration, and stop-and-go con-
ditions can be projected by the same methodology used for estimating the base-
year emissions as discussed in Sec. 2.1.2. The factors-e , Pb , and f
. . n,s n' n,s
in Eqns. 2.3, 2.4, and 2.5 have been developed to reflect existing federal
requirements for the reduction of the lead content of gasoline, use of lead-
free gasoline, and improved fuel economy. The same equations can be used to
estimate future lead emissions once the projected VMTs are known. The imple-
mentation of TSM measures or TCPs would affect projected VMT and/or average
speed figures for the projection year. The transportation agency and air
quality planning agency are required to integrate their respective trans-
protation and air quality planning efforts under Sec. 108 (e) of the Clean
38 39 40
Air Act and subsequent guidelines. ' ' Appropriate estimates of future
VMTs and speeds should be available as a result of these integrated planning
efforts and, of course, the states should already be aware of the progress of
these efforts. In cases where the VMT estimates will not be available in
time for the lead SIP, the Regional Office should be consulted. In such cir-
cumstances, it may be desirable to use VMT projections which do not account
for any transportation measures as a first estimate. If attainment can still
be demonstrated under such circumstances, a future upper bound to lead air
quality could be estimated. In areas where integrated planning is not re-
quired, Ref. 16 suggests several sources of transportation data. The local
transportation, planning agency is likely to be the primary source for such
information.
Projection of Lead Emissions from Reentrained Road Dust
The estimation of lead from reentrained road dust was discussed in
Sec. 2.1.2. The same method can be used in estimating future emissions if
projected VMTs are known. However, as noted in Sec. 2.1.2, Ref. 4 indicates
-------
50
that the emission factor for lead from reentrained road dust is expected to
decrease over time. It is suggested here that the emission factor for this
source category be calculated for the year n from
or
where
EFLERRD(n)
EFLERRD(n)
Pb
0.03 x
1.65
0.0182 Pb grams/vehicle-mile
(3.2)
= emission factor for lead emissions from
reentrained road dust in the year n, and
Pt>n = probable pooled average lead content of
gasoline in year n (g/gal).
The factor 0.03 is the emission factor given in Ref. 4 for 1975-76 and 1.65
g/gal is the average of Pb?5 (=1.7 g/gal) and Pb?6 (=1.4 g/gal). Equation
3.2 thus simply assumes that the lead emissions from reentrained road dust
will diminish in proportion to the prohable pooled average lead content of
gasoline. For 1980 when Pbg =0.5 grams/gal, Eq. 3.2 gives EF
LERRD
(80)
- .0091 grams/vehicle-mile in good agreement with the expected value of about
0.01 grams/vehicle-mile given in Ref. 4. It should be stressed that Eq. 3.2
does not reflect any definite guidance from U.S. EPA and this suggested pro-
cedure should be cleared with the EPA Regional Office prior to use.
Illustrations from Example Area
The following continues the example of arterial lead emissions from
heavy-duty gasoline-powered trucks in County C for which the base-year emis-
sions calculation was illustrated in Sec. 2.1.2. The arterial form of Eq. 2.5
is used again and Pbg2, VMTg2, and fg2 g must be found. Pbg2 again comes
from Table 4.3-8 in Ref. 17 which shows Pbg2 = 2.0 grams/gal. Users should
take care in using Eq. 2.5 to use the proper table for determining Pb . Table
4.3-1 in Ref. 17 is used for light duty vehicles and trucks while Table 4.3-8
is used for heavy-duty gasoline-powered trucks only. VMT00 was calculated for
oz
arterials using the area-wide growth rate of 1.00% yr"1 given in Table 3.3.
Thus,
VMTg2 (County C) = VMT?5(County C) x (l + 7 x .01) = 1,055,690 x 1.07
= 1,129,590 vehicle-miles/day.
-------
51
And as noted in Ref. 17, f = 5.7 vehicle-miles/gal independent of n and
n»s
s for this source category. Using Eq. 2.5,
E82^C°Unty C' neav7-duty gasoline-powered trucks)
= 0.70 x 2.0(g/gal) x l,129,590(vehicle-miles/dav)
5.7(vehicle-miles/gal)
.= 277,440 grams lead/day
= 112 tons lead/yr.
In developing the example plan, freeway calculations made use of link-
specific VMT and speed projections. These projected data were available for
1990 from the local- transportation planning agency. Linear interpolation was
used between the values for 1990 and values presented for 1975 in Table A.2
to estimate values appropriate to the projection year 1982. Because the 1990
data has the same form as that already presented in Table A.-2 ;it has not been'
reproduced here. Thus two levels of detail were used in projecting vehicular-
related lead emissions for the example area: a regional level for arterial
emissions and a link-specific level for freeway emissions. States should use
the level of detail appropriate to their resources and the extent and severity
of their lead problems.
A further simple example projects the County D lead emissions from
reentrained road dust on arterials. The base-year calculation has been pre-
sented in Sec. 2.1.2. VMTg2 was calculated for arterials using the area-wide
growth rate of 1.00% yr"1 given in Table 3.3:
VMTg2(County D) = VMTy2(County D) x (l + 7 x .01)
= 9,429,900 x 1.07
= 10,089,990 vehicle-miles/day.
Emissions can now be calculated using Eq 2.2 with DM = 1 and an emission fac-
tor calculated from Eq. 3.2:
82
Qrld = 10>089,990(VMT/day) x 365(day/yr)
x .0091(grams Pb/VMT) x (1/907,200)(tons/gram)
= 37 tons lead/yr
= Emissions from reentrained lead dust (rid)
in County D in 1982
-------
52
where additional factors have been -used to give, the result in tons/yr. This
is substantially, less than the 114 tons/yr of lead from reentrained dust from
arterial roads in County D in the base year. The,reduction in the emission
factor from 0.03 gm/VMT-in 1975-76-to ,.0091 gm/VMT in 1982 is more than suf-
ficient to counteract, the expected,increase in;VMT, The sensitivity of this
calculation to the value of the emission factor also indicates'that care
must be taken in calculations where the emission factor is a function of time.
3.3 PROJECTED'FUTURE EXAMPLE AREA.- LEAJ}, EMISSION^ INVENTORY
Using the methods presented and Illustrated above, an emission inven-
„ O 5 , „ t,, . „ X s. s f „ .„ ^m , _ - .• n »_ *.. 't _ , T, • .^ _ f ,^ , . • -j ^ »'.„»''•
tory was.pr.ojected for the .example area_. Table, 3.5 summarizes the projected
1982 emission inventory for lead in the modified NER format. Comparison with
Table 3.2 indicates a 68%-reduction in overall .lead emission, between 1975 and
1982. Reductions in land vehicle emissions,account for over 87%;of the"total
reduction; mo.st of the remainder, 12.7%, is accounted^ for by_Treduced emissions
from reentrainment and unpaved.roads. Reductions at .lead.:point.sources con-
'" -IS' "' ' ' ' - ', '"'**J J/"':.v -\ '^-• 'i>' '' "^-'-r.1- ,,>•*'.-*''» -irt:'-'... A:- '.'.B^' , .. '•-'.,
tribute less, than 1% *of -the.-, total-overall reduction^
-------
53
03
00
0\
CO
J
CO
CO
•H
T3
cd
01
_, CM
CO OO
cu o>
01 a
w
CU
1 \
u
cu
o
p-l
IO
CO
cu .
H
cd
cc
UJ
£
UJ
a.
in
Z
o
00
O
05
5
aj
u
C.
CC
C
0"
a
3
CN
C
c
ANTHRACITE COAL
_i
5
i—
UJ
Q
U4
CC
3
u_
C
C-
f
BITUMINOUS COAL
UJ
3
u-
z
0
-
J
m
S
D
vO
O
o
5
u
_l
5
UJ
cc
_t
JC
UJ
-
X
JJ
1
RESIDUAL OIL
!
NATURAL GAS
1
3
5
o
o
CM
O".
-J
3
ANTHRACITE COAL
y
cc
5
UJ
J
UJ
00
0)
01
*"<
1
BITUMINOUS COAL
Z
o
-
3.
LJJ
Z
LU
0
§
a
0
<
i
LIGNITE
I
n
0)
01
<
i
RESIDUAL OIL
§
cd
a
<
i
5
u
n
Q
NATURAL GAS
P-IOCESS GAS
i-H
4J
5
o
\J
b
4J
•H
pa
SOLID WASTE/COAL
N
00<
i-H •
O
4J
Z
2
o
OO
1
UJ
cc
o
u
_J
n
Q
Z
2
UJ
E
NATURAL GAS
Z
2
UJ 2
cc r
< Q
PROCESS GAS
Z
2
LU
8
UJ
cc
o
D
Z
2
z
D
LIQUID PETROL GAS
Z
2
BAGASSE
Z
2
OTHER
n
N
vO
X
<
5
CM
n
•
•z
2
-------
54
co
o
10
111
U
P <
u. o uj s.
I*
[ 03 °£
J i ^
=> Q 7
u. o 5
-------
55
•o
4-1
(3
O
CJ
Q
S
0
CC
1
o
£
^
CHEMICAL MANUFA
<
n
h
2
UJ
FOOD/AGRICULTUR
1
• co
• v
li
1
u
S
i
|
z
1
o
c
«
J
c
a
3
j
MINERAL PRODUCT
£
PETROLEUM INDUS!
WOOD PRODUCTS
PROCESS EVAPORA1
^
•-"
z
o
1
E
c
u.
t
LJ
n
LEATHER PRODUCT
o
CC
TEXTILE MANUFAC1
INPROCESS FUEL
CO
t>
ft
fl
> - Secon
!
a
«*
iH
CO
en
4-J
o
_j
5
^
•*
z'
CJ
Z
u
s
GOVERNMENT
§;
s
Z
7
CC
CO
z
UJ
1.
i
g
u
—
1
OTHER
_)
4.
S
is
5
3
«*
<
5
a
?
z
ON-SITE II
<
j
in
LU-
CE
U
;»
OPEN BUR
jj
X
K
^
en
TOTAL
LU
CC
±
o
ON-SITE IN
COMMERCIAL-
z
D
2
c
LU
_j
<
5
3
/5
:
UJ
CC
O
^
r
to
jj
i.
1-
g
£
g
—
APARTMEC
UJ
CC
OTHER
z
g
1
uj
CC
TOTAL
1
'
H
Z
°
LU
CC
Z
CJ
ON-SITE IN
INDUSTRIAL
Z
g
ERATION
UJ
CC
7
CC
m
jj
3.
1—
Z
g
z
g
z
u
^_
AUTO BOD
t-
z
g
OTHER
1
[
UJ
CC
TOTAL
1
I
i-
z
O
CC
1-1
en
UJ
CC
TOTAL SOLID WASTE
3
r-
Z
g
§
55
a
-------
56
•P
o
TJT
ON
v£>
*
m
o a)
* S 2
CO
o
.£
o £2
Z I
< 01
_J >
1C
o
o
-------
57
4-J
§
LO
en
01
i-H
•3
g
a
_J
LU
O
K
3
K
CO
O
r
INDUSTRIAL SOU
DRY CLEANING
SOLVENT
EVAPORATION
STRUCTURAL
[FROST CONTROL
SLASH BURNING
WILD FOREST .
AGRICULTURAL
-
CO
LU
CC
LU,
CM
oo
0
»
UNPAVED ROADS
fP
UNPAVED AIRSTR
e*
<
O 00 K
STORAGE PILES
TILLED LAND
UNTILLED LAND
AIRBORNE DUST
CAUSED BY
NATURAL WINDS
O
LU
Z 1
_J
UJ '<
O iJJ
»ff
5 -
in
en
en
i-H
LU
CC
_J
CM
cn
*
vo
Z
2
^
o
Q
•z
cc.
o
CO
CO
a>
o
(!)
C
CO 4J
e §
o e
CO -H
-------
-------
59
4 ALLOCATION OF EMISSIONS
The allocation of lead emissions employs essentially the same tech-
niques as outlined for particulates in Refs. 8, 13, and 16. States should
already be familiar with these techniques and in many cases may already have
operating allocation procedures available. However, a higher level of detail
may be warranted in allocating lead emissions from highway vehicles than is
warranted in allocating particulate emissions from this category in areas
where lead emissions from highway vehicles comprise the major portion of the
emission inventory. States must decide whether such additional detail is
warranted based on an assessment of their resources, the level of detail in
the data available for the allocation, and the severity of the lead problem.
The allocation of emissions is performed in order to put the emission
inventory in the form required by the chosen air quality model. Thus, the
model chosen affects the level of detail at which the allocation needs to be
performed. A dispersion model must be used to calculate the magnitude and
location of lead air quality concentration maxima resulting from significant
lead point sources [51.84]. Such a calculation would require the allocation
of emissions to an appropriate emissions grid. On the other hand, modified
rollback must be used as a minimum in demonstrating attainment and maintenance
in the vicinity of an air quality monitor recording violations [51.85]. If
one of the more complex forms of modified rollback is employed in which dis-
tance from the monitor and, perhaps, wind direction and frequency are taken
into account, emissions would need to be allocated to another appropriate
emissions grid. However, a dispersion model may also be used [51.85]. Thus,
if there are violations of the lead NAAQS in the vicinity of significant lead
point sources, states may want to consider using a dispersion model to satisfy
the requirements of both Sections 51.84 and 51.85 in order to avoid allocating
to two different emissions grids. Modified rollback could still be used in
the vicinity of other violations not located, in the vicinity of significant
lead point sources.
4.1 PROCEDURES
The emissions from point, area, and line sources must be allocated
prior to modeling. As indicated above, the allocation of lead emissions for
point and area sources is essentially the same as the allocation of partic-
-------
60
ulate emissions, with the possible exception of area source lead emissions
from highway vehicles. Reference 13 describes three orders of allocation
procedures applicable where county-wide totals are available. Reference 8
details the Computer-Assisted Area Source Emissions Gridding Procedure (CAASE)
which uses demographic data available from the U.S. Bureau of the Census to
develop subcounty allocation parameters. The entire procedure, including the
data requirements, the considerations involved in developing an emission grid,
and allocation procedures, is summarized in Ref. 16.
Sources to be modeled as points are located at their true locations
by UTM coordinates. This happens almost automatically when regional disper-
sion models are used. However, when modified rollback is used, the source
emissions must be allocated by UTMs to the appropriate emission grid sector.
It should 'be noted that allocation 'to these sectors must be treated somewhat
differently than the more familiar allocation to'dispersion model emission
grids, because the modified rollback grid cells are not squares.30* The al-
location of projected future point source emissions also requires a decision
as to how the emissions from new/modified Sources will be allocated. Refer-
ence 16 lists three possible procedures:
1. Assume all new source emissions occur at the same
location as existing sources (i.e., growth-in-place),
2. Assume all new source emissions occur at new sites
and allocate according to an allocation parameter or
land-use distribution function, or
3. Determine the mix between existing and new source
activity and allocate accordingly.
Procedure ~two or three is generally better than procedure one. teen growth
is assumed to occur in place, emissions from new/modified sources are allo-.-i
cated by UTMs to the locations of individual plants at which the associated
activity changes are projected to Tiave occurred. . Of course,, if precise lo-
cational information is available as 'the result of announced expansions, per-.
mit applications, or interviews, it should be used to allocate new/modified
emissions. If either of the latter two procedures is employed, emissions
from new/modified sources are totaled at the county level and allocated using
an appropriate parameter and the procedures described in Refs. 3, 8, and 16.
*A reprint of Ref. 30 is contained in Ref. 17.
-------
61
Except for emissions from new/modified sources which must be allo-
cated when their future locations are unknown, the allocation,procedures de-
scribed in these references are used principally for allocating county-wide
. area source totals to emission grids. If area source emissions are available
.based on activity levels at the subcounty level, this additional detail should
be retained, if possible, rather than relying upon an allocation procedure.
Such additional detail may be available for'arterial VMTs which are frequently
available for subcounty traffic zones, Given the likely.importance of this
category In the lead emission inventory, states with sufficient resources may
want to consider retaining the subcqunty level of detail when it is available
or developing an allocation parameter based on the subcounty VMT data.
For lead, one further point should be noted: additional detail for
allocation may be available for that part of the lead area source emissions '
associated with large particulate sources which are nevertheless small lead "";
emitters. As noted in the discussion of the emission inventory in Sec. 2,
the location of these sources by .UTMs can be retained so that they can be
allocated to the emission grid cell in which they actually occur. The alter-
native is to place- the emission from these "residual" point sources in the
appropriate area source category and allocate them to the subcounty grid cells.
States should use the method appropriate to their resources and the extent of
their lead air quality problem.. Regardless of the procedure used for the
residual points, their emissions will be part of the area source emission rate
in the grid cell to which they are allocated.
The allocation of emissions inventoried as line sources is not dis-
cussed in great detail in the references noted heretofore. If, however, lead
• emissions from freeways or other roadway sources have been inventoried as
lines and the chosen model does not simulate line sources, then these emis-
sions must be allocated to the subcounty emission grid. The emissions could,
of course, be totaled at the county level and allocated by the appropriate
allocation parameter. This procedure would lose the spatial resolution gain-
ed by inventorying the sources as lines in the first place. It is suggested
here that emissions from uniform line sources be allocated to emission grid
cells, according to the fraction of the length of the line in the grid cell.
That is,
-------
62
Contribution of Line
Source i to Total
Emissions in Grid
[Cell j
Fraction of Length
of Line Source i
That Lies in Grid
[Cell j
'Total Kmis-1
sions from
Line Source
i
(4.1)
This procedure provides a reasonable compromise between a fully detailed model-
ing of line sources and the loss of available information occassioned when
emissions are aggregated at the county level and then reallocated. Allocation
of nonuniform line emissions would require a direct determination of the emis-
sions from the segment of the line lying.in a particular grid cell.
4.2 ILLUSTRATION FROM EXAMPLE AREA
In developing the example lead strategy, the entire example area was
modeled using the ANLCDM dispersion model (see Sec. 5). This model treats
point and area but not line sources.
The area source grid was an array of squares 1.6093 km (1 mi) on an
edge. This array of small cells was chosen because population and housing
unit data derived from U.S« Bureau of the Census census tract data by the
methods described in Refs. 13 and 16 were available on this scale. There were
1847 grid cells covering the example area. Normally, these small cells would
be aggregated in areas where a high degree of spatial resolution was felt to
be unnecessary as described in Refs. 8 and 16. This aggregation would reduce
the number of area sources to be modeled and thus reduce the tiwe and resources
required for modeling. For the example area, however, the full set of 1847
grid cells was modeled, because the capability of handling a large number of
area sources was already available; states are not expected to use this high
level of detail in developing lead strategies. Since the grid cell siz.e
utilized is too small to show clearly on a map of convenient size, no graphic
presentation of the area source grid has been included here.
Depending on source type, the emission allocation was done by one of
four methods. First, fugitive emissions and the emissions from "residual"
point sources were allocated by UTMs.; that is, they were added to the emis-
sions rate of the area source in which they were located.* The second method
was applied to area source categories for which emissions were available at
*A preferable method of treating the fugitive emissions associated with
point sources may be to use the pseudostack method described in Sec. 5.2.
-------
63
the countywide level. This method was the traditional method of using. an
allocation parameter as described in Refs. 8, 13, and 16. Table 4.1 gives
the allocation parameters used for each of the lead emission categories.
These allocation parameters are consistent with those recommended in Ref. 8.
The third method was applied to freeway emissions which were allocated by
a method based on Eq. 4.1. Finally, the conservative "growth-in-place"
assumption was used to allocate emissions from new/modified point sources.
Emissions allocated by the first three methods were summed to obtain
a total area source emission rate for each of the 1847 grid cells. Table
4.2 presents a partial listing of the final area source emission rates. The
fugitive component of the point source emissions has been included in the
appropriate grid cells in this listing. It should also be noted that the
number of each grid cell can be placed in a one-to-correspondence with the
UTM coordinates of the grid cell's southwestern corner for proper location
during modeling.
Table 4.1 Lead Emission Allocation Parameters
Source,
Category
Fuel Combustion
Residential
Commercial/Institutional
Industrial
Off-Highway Vehicles
Unpaved Roads
Arterial Roads
Incineration"
Land Tilling
Residual Points
Subcounty Allocation
Parameter3
Housing units
Population
Population
Inverse population density
Inverse population density
Population
Population
Inverse population density
UTMs,
These parameters are generally based on those used in U.S.
EPA's CAASE Procedure (see Ref. 8).
Residual points are those lead sources too small to be
inventoried as points but for which UTMs exist because of
their inclusion .in the particulate point source inventory.
-------
64
to
e
•3
CO
CO
1
0)
O
H
O
(/I
§
•a
a)
•w
cd
o
O
t-l
CO
a\
o
00
M
•H
a)
tH
o.
CO
CM
•
si-
01
UJ >— jr
'
^™™!r!^ss^
.• c. *-! < • trj'c-i'~t:.'
r-1- r ; ^ "o r- °°rjs <=» *~ c'*to **tri ^»-- ^ *><=>- «f0 ^ to o r^ ei e. o
£SayjiJid£^y^
5«!8SPo&^S2^r«!2•Z2^2!iS^p^^JBSl^«^{^•^^^^;o?M
^^^^^^^u^siaiD^^^^^io^^^MM^^ln^^^^^
-------
65
5 MODELING PROCEDURES
5.1 REQUIREMENTS
Regulations require that:
• The plan employ the modified rollback model, at a minimum,
in urbanized areas with measured quarterly lead means in
excess of 4.0 yg/m3 since January 1, 1974 [51.83]; or that
• In other areas the plan employ the modified rollback model,
at a minimum, in the vicinity of any air quality monitor
that has recorded lead concentrations in excess of the lead
NAAQS [51.85]; and that
• In all areas dispersion models be used to calculate the
magnitude and location of the maximum lead air quality con-
centration in the vicinity of the significant lead point
sources listed in Sec. 2.1.1 whether or not the lead NAAQS
are violated in the vicinity of these significant point
sources [51.84]. ......... .„_„...'.
The requirements are also discussed in Ref. 17. States are encouraged to use
dispersion models for the entire analysis whenever possible, particularly in
cases where such models are already available and being used to analyze other
pollutants. The additional accuracy generally yielded by dispersion models
is deemed sufficient to make their use highly desirable.
Regardless of the type of model used, it should be noted that credit
may be taken for expected reductions in background lead levels resulting from
federal programs for the reduction of lead in gasoline, the prohibition of the
use of leaded gasoline in catalyst-equipped vehicles, and improved fuel econ-
omy.
5.2 APPLICABLE MODELS
Dispersion Models
EPA's supplementary guidelines for lead17 indicate that the point
source models for sulfur dioxide and particulates listed in Ref. 31 are accept-
able for dealing with lead concentrations in the vicinity of significant lead
point sources. States should be aware, however, that hand calculation or pro-
gram modification would be required before quarterly averages could be obtain-
ed from the referenced Single Source (CRSTER) Model. This model has the ad-
ditonal problem of not treating fugitive emissions directly.41 The model can,
-------
66
however, treat up to 19 emission points. It is suggested here, based on a
recommendation made by EPA for use in multisource dispersion models, that
the situation could be handled by a pseudostack for the fugitive emissions.
Such a stack would be 5-10 m'in height, depending on the average release
height of the,fugitive emissions being modeled. The emissions would be as-
sumed to be released at ambient temperature with an initial upward velocity
of about 0.1 m/sec. Since this guidance does not represent official policy,
states desiring to handle fugitive emissions from significant lead sources in
this fashion should consult with the Regional Office's air pollution meteor-
ologist for the latest available information.
However, as noted in Ref. 31, the CRSTER Single Source model should
be applicable provided that the lead emission can be assumed to behave as a
gas. This assumption may sometimes be inaccurate for lead, because deposi-
tion and fallout are also expected to be important for lead in certain cir-
cumstances. The Single Source (CRSTER) Model does not account for these pro-
cesses. Reference 17 contains reprints of several articles which states de-
siring to account for the deposition of lead may find useful.
Recognizing that models accounting for deposition and fallout and cap-
able of treating fugitive emissions would be useful for dealing with lead, EPA
is currently developing an Industrial Source Complex Model with the capability
of treating these factors. The model is expected to be available by mid-1979.
EPA is also preparing a quarterly model (PBLSQ) which should be avail-
able by summer 1979. This model will be applicable to lead and could be used
in cases where dispersion modeling is required in an entire urbanized area
or where states choose to use a dispersion model to treat the entire region
of interest. Meanwhile, the models described in the section on multi-source
models for sulfur dioxide and particulates in Ref. 31 are acceptable although
these models do not account for deposition and fallout of larger particles.
These models are:
• Climatological Dispersion Model (CDM or CDMQC),
• Air Quality Display Model (AQDM), and
• Texas Climatological Model (TCM).,
CDMQC is a version of CDM which gives individual point and area source contri-
butions at selected receptors. These are useful aids in strategy development.
-------
67
GDM gives only the total concentration at each receptor. If a model is being
used that does not have any capability to provide source contribution lists,
it is recommended that the effort be expended to develop and implement one.'
These three models are presented as models for predicting annual average air
quality. However, use of quarterly rather than annual wind roses as the meter-
ological input will result in the simulation of quarterly air quality esti-
mates. The remarks presented above concerning fugitive emissions, fallout,
and deposition also apply to these models. It should also be noted that these
multisource models do not have the capability of treating line sources. Thus,
if line sources have been inventoried, they must be allocated to the area
source grid by a procedure such as that described in Sec. 4.1.
EPA's modeling guideline31 notes that in cases where (1) the recom-
mended air quality is not appropriate, (2) the required data is unavailable,
or (3) a better applicable model or analytical procedure is available, models
other than those mentioned above may be used when deemed appropriate by the
Regional Administrator. It also encourages early discussions between state
and Regional Office staffs in such circumstances and suggests that all devi-
ations from the guideline be "fully supported and documented." It is further
suggested here that Refs. 43 and 44 be consulted for guidance in documenting
alternative models and determining how alternative models compare with the
models recommended in the guideline.
Validation and Calibration
As noted in Refs. 16 and 31, some validation is almost always attempt-
ed with dispersion models in order to improve the agreement between measured
and predicted concentrations. References 16 and 31 discuss some of the spe-
cific steps involved in the normal validation procedure.
Long-term models are frequently calibrated against observed air qual-
ity. The most frequently used procedure involving linear regression of ob-
served against predicted concentrations is discussed in Ref. 16. When using
linear regression, a correlation coefficient, r, (significant at the 5% level)
has frequently been taken as an indication that the regression is acceptable.
Another test is suggested in Ref. 31 which notes that if less than 50% of the
variation in the measured concentrations is.accounted for by the model, it is
doubtful that there is justification for its use. This test would require
45
-------
68
that r2, not r, exceed 0.50... If the number of pairs of values is less than
eight the significance test on r is more stringent than the test requiring
that r2 exceed 0.50. Since there is currently no widely accepted standard,
states should consult with the Regional Office if there are any questions
concerning the acceptability of the calibration after some effort has been
expended in validating a model.
Modified Rollback
Acceptable forms of modified rollback are described in Ref. 30;*
equivalent forms also may be used.17 Reference 30 presents four forms of
modified rollback which account progressively for:
1. Multiple categories of sources which may have
different growth rates and control efficiencies,
2. Average stack heights in different categories.
3. The source-receptor distance, and
4. The wind direction frequency.**
The first form is the form actually employed in many prior rollback analyses,
simple rollback being a form which applies the same rate of growth and control
efficiency to all the sources being modeled. Reference 30 itself concludes
the third or fourth form would be more appropriate than the first two forms
in most circumstances. States deciding to use modified rollback may want to
consult with the Regional Office to ensure that the form chosen will be ac-
ceptable.
Two additional points should be made about the use of modified roll-
back. First, both the third and fourth forms require that emissions be al-
located to emission grid cells composed of annular rings or sectors, respec-
tively, centered about each receptor where an analysis is required. Each
*This article is reproduced as Appendix G in Ref. 17.
**States desiring to use the fourth form of modified rollback should note
that the illustration of wind direction frequency in Ref. 30 (or Ref. 17)
is potentially misleading. If north is taken as being vertical on the
printed page, then standard NCC wind roses of eight sectors would give a
wind direction sector centered about north; the illustration in the ref-
erences appears to show a wind direction sector centered about the north-
northeast. It would, of course, be possible to develop such a set of
wind direction sectors, but not from the meteorological data normally used
-------
69
receptor analyzed will require a separate emission grid and a separate emis-
sion allocation. Second, the equations in Ref. 30 assume the background, b,
is constant over the time frame of the analysis. States wishing to account
for expected reductions in background lead concentrations would need to mod-
ify the equations presented in the reference.
5.3 EXAMPLE AREA MODELING RESULTS
The ANLCDM dispersion model was used to perform the lead air quality
analysis for the example region. ANLCDM is a version of CDMQC using an im-
proved algorithm for calculating area source contributions.
Based on the 1975 inventory, an initial run of the model was used to
develop a calibration. A scatterplot of the data is shown in Fig. 5.1. Lin-
ear regression with two parameters yielded a regression line
Xobs * °-46XCalc + °'58 (5.1)
and a correlation"coefficient r = 0.315 (r2 = 0.0992). The correlation coef-
ficient r was significant at the 5% level but the regression explained only
about 10% of the variation in the data. It should be noted, however, that
ANLCDM, like CDMQC, does not treat deposition, fallout, or resuspension and
there are a number of uncertainties in the analysis. In such situations at-
tempts would have to be made to validate the model by the procedures listed
in Ref. 31. Resources for extensive validation were not available, so the
model was used in an uncalibrated form for most of the analysis. The tabu-
lated values that follow, however, can easily be converted to calibrated
values by applying Eq. 5.1 after subtracting the 0.15 yg/m3 background from
the listed noncalibrated values.
Figures 5.2 and 5.3 illustrate the uncalibrated modeled lead air
quality in 1975 and 1982, respectively. In these two figures, calibrated
air quality concentrations have been indicated in parentheses following the
uncalibrated values. These' isopleths were plotted using a computerized plot-
ting routine to interpolate between lead concentrations calculated at the 400
receptor points shown in Fig. 5.4. This receptor network contains a large
number of points; states would not normally be expected to calculate air
quality at such a large number of points. Tables 5.1 and 5.2 present the
computed air quality values at each of the receptor points in Fig. 5.4 for
-------
70
1.6
^ 1.4
fO
E
^ 1-2;
CD
O
X
o
h-
o:
CO
0.8
o
o
0.6
3 0.4
tr.
CO
m
o
0.2
0.0 0.2 0.4 0.6 0.8 1.0
CALCULATED LEAD CONCENTRATION, XCALC (/ig/m3)
T?i°. 5.1. Scatterplot of Baseyear Lead Concentrations
at Calibration Receptors
-------
71
0.65(0.81)
.10(1.02)
00(0.97)
50(1.20)
74(1
.77)
43)
00(1.43)
.32(1.12)
Fig. 5.2. Modeled Example Area Lead Air quality in Base Year 1975
-------
72
£45(0172}
.65(0.81)
187(0.91)
0.70(0.83)
A 65 (0.81)
0.52(0.75)
Fig. 5.3. Projected Example Area Lead Air Quality in 1982
-------
73
40
38
36
34
32
30
28
Off
26
24
22
20
• «"i
18
16
14
12
in
IU
+ + 4- 4- H
1 — 4- 4- 4- 4- H
4- 4- 4- 4- H
— 4- 4- 4- 4- H
4- 4- 4- 4- H
— 4- 4- -V 4- H
+ 4- 4- 4- H
— 4- 4- 4- 4- H
4- 4- 4- 4- -
— 4- 4- 4- 4- -
+ [~+ 4- ~=F =F~=
— 4- 4- 4- 4- 4- -
4- 4- 4- + 4- -
— I- 4- 4- 4- 4- -
f 4- 4- + 4- -
4- + 4- 4- 4- -
— 4- 4- 4- 4- + -
4- 4- 4- 4- 4- -
— 4- 4- 4- 4- 4- -
4- 4- 4- 4- 4- -
— 4- 4- 4- 4- 4- -
4- 4- 4- + 4- i
4- + 4- + 4- H
— H
H
1
12 14 H
T . _
>- 4- 4-
h 4-4-
K + +
h 4- 4-
h 4- 4-
h 4- 4-
1- 4-4-
h 4- 4-
h 4- 4-
h 4- 4-
F =F T~~
h 4- 4-
h 4- 4-
h 4- 4-
1- 4- 4-
h 4- 4-
1- 4- 4-
h 4- 4-
h 4- 4-
h 4- 4-
h 4- 4-
h 4- 4-
• 4- 4- J
•—±-A
• 4- 4-
h 4- 4-
1
5 18
4-4-4-
4- 4- 4-
4- 4- 4-
4- 4- 4y
+ + f
+ +r
4- 4- H-
4- 4- X
4-4- 4-N
+ + 4-
4-4-4-
4-4-4-
4- 4- 4-
4-4-4-
+ i i
4- 4-
4-4-4-
+ 4-4-
4-4-4-
4- 4- 4-
4-4-4-
4-4-4-
4- 4: 4-
4-4-4-
4-4-4-
4-4-4-
4-4-4-
4-
4-
4-
4-.
4-
20
4-
4-
s.
4- \
4- 4-\r
4- + 4\
4- 4- +\
+ t i 1 1
V
4- 4- 4- +
4- 4- 4- +\ 4-
4- 4- 4- 4- W
4- 4- 4- 4- >
4- 4- 4- 4- V
4-4-4-4- 4-\ 4-
4- 4- 4- 4- 4- \
i
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- 4- 4- 4- 4- 4-
4- f 4- 4- 4- 4-
1 1 1
22 24 26 i
4-
\
4-
4-
4-
+
4-
4-
4-
?8
Fig. 5.4. Air Quality llodel Receptor Network
-------
74
-------
75
. oooooooooooooooooooooooooo
i • i i t i i i i i i i i i i i i i r i i i i i i i
UIMJUlUJtUHIUIUllUUIUllUUilUUIUllUUIUIHtlUUllUlUlUtU
cd
. CO
4J
cd
o
8,
cd
CU
OOOOOOOOOOOOOOOOOOOOOOOOOO
I t I I I i t I I I t I I I I t I I I I t t I I t I
tii m m m m 111 m mm in tii m m m u? MJ m Hf 111 HI tu uf «'tn-tf-fnm(M(M(MCM
ooooooooooooooo
I I t I I 1 I I I I I I I I I
bO
•H
0)
0)
00
ooooooooo ooooooooo o o o o o
JitLiLuiuiiLiLtLiLiL JiiLtLiL ' ' * * * ' ' ' *
>^
ooooo ooo ooooooooo
> o o ooc
000 0000
ooooooooooooooooooooooo
i i i i i i i i i r i i i i i t t i t i i i t
m m in in m in in in in ifi ^n n< ni in nf m in in »n ni III m 111
>-
OOOOOO^OOt> OOOOOOOOOOO
i t i i i i i i i i i t i i i i i i i i
--^
13
---^
M
O
4J
ft
0)
o
cu
•
o
•H
4J
cti
o o'o o o'o o
-------
76
1975 and 1980, respectively. In making these computations, it was assumed
that no new control programs were in force, only existing state programs,
federal NSPS, and the federal programs for the reduction of the lead content
of gasoline and improved fuel economy were assumed to be in effect.
Figure 5.2 shows several areas exceeding the lead NAAQS in 1975. De-
spite growth and development and increased vehicular travel, Fig. 5.3 shows
compliance with the lead NAAQS in 1982. This conclusion remains the same
regardless of whether the calibrated or uncalibrated air quality values are
used. The improvement in air quality between 1975 and 1990 is due almost
entirely to the reduction in emissions from highway vehicles as was noted
above in Sec. 3.3. This reduction in emissions results from the federal pro-
grams for the reduction of the lead content of gasoline and improved fuel
economy.
-------
77
6 ANALYSIS OF MODELING RESULTS
6.1 GENERAL
The analysis of modeling results is outlined fairly thoroughly in Ref.
16 which discusses four basic considerations involved in such an analysis:
• The magnitude of the problem,
• The geographical extent of the problem,
• The temporal extent of the problem, and
• The identification of sources contributing to the problem.
The guidance presented in Ref. 16 emphasizes point sources and the use of dis-
persion models. Little need be added to that guidance to account for lead
except to add several comments concerning area sources and the use of modi-
fied rollback.
Area Sources
As noted in the development of the emission inventory, highway vehi-
cles may be the predominant source of lead emissions in some areas. It is
thus important to be able to assess the contribution of this source category
in analyzing modeling results and developing control strategies. Highway
vehicle emissions will almost always be treated as part of the area source
emission rates in dispersion models. However, the dispersion models recom-
mended for use in Sec. 5 '31 give at most the contribution of each individ-
ual area source, of which the allocated highway vehicle emissions are one
component, to the concentration at a given point. (This problem will be
eliminated with the publication of the Industrial Source Complex Model and
the PBLSQ model.) A more useful piece of information might be the total con-
tribution of the individual area source categories to the concentration, since
regulations are generally developed to cover entire area source categories.
The category-specific contributions can be found by making multiple model runs
with a model which gives an area source culpability list. Each run would be
made with an emission inventory consisting of the allocated emissions from the
area source category whose contributions were desired. The sum over all area
sources of the individual area source contributions at each receptor of in-
terest would then give the total contribution from the category at these
receptors. If the use of this technique is contemplated, care should be taken
-------
78
during emission allocation to tabulate the emissions separately for those
categories whose individual contributions are likely to be desired. Also,
since multiple model runs are required, the categories should be chosen judi-
ciously to avoid the use of a large amount of computer resources. States
might also want to consider developing computerized routines to produce cul-
pability lists for specific area source categories for use with their dis-
persion models. If neither of these suggestions appears viable, the area
source emission inventory can be examined to determine the largest emitting
categories. These are likely to be'the largest contributors to problems
caused by area sources and hence should be the initial categories 'selected
for control during strategy development. This last method relies only on
total emissions and neglects the -effects of allocation and dispersion. It
is thus likely to be less precise than developing category-specific 'contri-
butions using a dispersion model.
Modified Rollback
Modified rollback need .only be used in the vicinity .of monitors ex-
ceeding the lead NAAQS unless a monitor exceeds 4.0 yg/m3, quarterly average,
in which case attainment must be demonstrated in the entire urbanized area
using modified rollback as a minimum [51.83 and 51.85]. If one of the forms
of modified rollback which allocates the emissions to grid cells is used,
the projected .changes in air quality apply only at the central .point of the
emission grids. Thus, in areas with only a few lead monitors, .an assessment
of the overall magnitude, geographical extent, and temporal extent of the
lead problem is likely to be inaccurate. This situation is not unexpected
given the approximations made in the modified rollback models and it is
still within the spirit of these approximations to make these assessments to
the degree possible in analyzing the results of modified rollback.
It is possible to estimate the concentration contributions of indi-
vidual source categories using modified rollback. An example illustrating
the technique is given in Refs. 17 and 30. In general, all forms of modi-
fied rollback can be derived from Eq. 34 in Ref. 30:
(c, - b)
max-base
(6.1)
-------
79
where c is the projected concentration and wf. is proportional to the emis-
th J
sions in the j source category. Thus, by keeping track of the various terms
in the sum, t"he contributions of the source categories j to the concentration
c. can be determined.
6.2 ILLUSTRATION FROM EXAMPLE AREA
The modeling results presented at the end of Sec. 5 indicate that the
lead NAAQS will be attained in 1982 and hence that no extension will be re-
quired. Furthermore, no additional strategies beyond those already in effect
will be required. It should be noted that the modeled point source contri-
butions used in developing the air quality information in Table 5.2 and Fig.
5.3 are conservative in that they do not include the impact of reasonably
available control technology (RACT) for particulates which will be required
as part of the example area's nonattainment SIP revision. The uncalibrated
hot spot in Table 5.2 is only 0.87 yg/m3 which corresponds to a calibrated
value of 0.91 Ug/m3. Both values are well below the 1.5 yg/m3 NAAQS. Since
the meteorological data used appeared to be the worst available from an air
quality perspective, there is little doubt that the lead NAAQS will be at-
tained.
Had there been NAAQS violations in 1982, it would have .been necessary
to identify the sources contributing to the air quality problem. ANLCDM pro-
vides the source contributions for both point and area sources. The model
was used in a way which allowed the determination of the contributions of
individual area source categories rather than the contributions of individual
area sources to air quality at a given receptor. Table 6.1 presents a sum-
mary of such results for receptor 297 (I = 22, J = 27) in 1982. In the table,
area sources have been placed in groups; the same allocation parameter has
been used for each group. Except for allocation by UTM coordinates, comparison
of the total emissions from each source category within a group with the total
emissions from the group will accurately reflect the individual category's
contribution to the air quality. For example, Table 3.5 shows that the 1985
emissions for the first group of area sources in Table 6.1 are projected to
be:
-------
so
Table 6.1 Sample Source Contribution
Analysis for Example Area3
Source Lead
Category
Major Point
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
Area
Commercial/Institutional
fuel combustion, industrial
fuel combustion, incineration,
and arterial foadsb
Freeways
Residential fuel combustion
"Residual" points and
industrial fugitives
Other
Background
TOTAL
Quarterly
Concentration
(yg/m3)
0.000
0.081
0.000
0.000
0.000
0.000
0.000
.0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.193
0,039
0.082
0.079
0.011
0.150
0.64
Percentage of
Total Concentration
0.05
12.69
0.02
0.02
0.01
0.00
0.00
0.01
0.01
0.00
0.00
0.05
0.01
0.01
0.00
30.31
6.07
12.85
12.50
1.77
23.60
100
At receptor 295 (I = 22, J = 27).
The^quarterly lead concentration due to arterial roads is 0.127 ug/m3.
Similarly, the air quality contribution of each.individual area source
category can be determined by proportion as discussed in the text on
pp. 79 and 81. However, the contributions of individual "residual"
points cannot be determined by this method.
-------
81
Industrial Fuel Combustion
Commercial/Institutional Fuel Combustion
Arterial Roads
Incineration (Solid Waste Disposal)
0.7368 tons/yr
6.7273
620.6600
314.8500
942.9741 tons/yr
Total 1985 Lead Emissions
Thus, arterial roads account for (620.66/942.9741) x 100 = 65.82% of the 1985
emissions from the group. Since all source categories in the group were al-
located to the model's emission grid using population as an allocation param-
eter (see Table 4.1), arterial roads would also account for 65.82% of the
group's air quality contribution at any receptor. For receptor 297:
Air quality contribution from arterial roads
= .193 x .6582 = 0.127 yg/m3 lead, quarterly average.
Although this receptor is not the area-wide hot spot, it is close to the hot
spot located at I = 24, J = 24. Based on this receptor and the relative mag-
nitudes of the emissions given in Table" 3.5, vehicles, residential fuel com-
bustion, and industrial process fugitives would be the most likely candidates
for additional control if there were a lead problem. The potential for ad-
ditional control at Plant B would also need to be investigated. In actual
practice, these initial indications would need to be evaluated further and
modified, if necessary, by examing the regional hot spot itself and the other
local hot spots in the example area.
-------
-------
83
7 SELECTION, TESTING, AND EVALUATION OF STRATEGIES
A fairly complete discussion of the methods for selecting, testing,
and evaluating control strategies is available in Ref. 16. Since additional
controls beyond those already contemplated do not appear to be required for
attainment of the lead NAAQS in the example area, this discussion is limited
to a lead-specific amplification of the material in Ref. 16 and no illustra-
tive material from the example region is included.
7.1 AVAILABLE STRATEGIES .
Reference 16 discusses both technological and land use and planning
controls. The technological strategies include:
• New Source Performance Standards (NSPS),
• Existing source retrofit,
• Phase-out of emission sources,
• Fuel conversion,
• Energy conservation,
Combination of emission sources, and
Fugitive dust controls.
It should be noted that special operating conditions or supplementary control
systems (SCS) are suggested as a pocential strategy for short-term standards
in Ref. 16. Given that there is currently no short-term lead standard and
the current Clean Air Act requirements that emission control systems operate
continuously, SCS need not be considered as a lead control strategy. Also,
since the recognition of the energy crisis, fuel conversions can probably
not be used as a control technique in most situations. In fact, fuel conver-
sions are most likely to be from scarce clean fuels to more abundant dirty
fuels giving an emission increase as discussed in the example in Sec. 3.1.2.
Land use and planning measures are summarized in Tables 8-1 and 8-2
of Ref. 16 and discussed in the accompanying text. States desiring to inves-
tigate these types of strategies should consult the reference. The measures
listed there could be applied to lead.
Given the large amount of process fugitive emissions likely to be
associated with lead point sources, recent legislative and regulatory re-
-------
84
quirements, and the large amount of emissions associated with vehicular
sources, states may want to consider several other options in developing
lead strategies. These include:
Industrial fugitive emission controls,
• Reasonably Available Control Technology (RACT),
Best Available Control Technology (BACT), or
Lowest Achieveable Emission Rate (LAER) controls,
• Limitations on the lead content of gasoline more
stringent than the federal limitations,
• Transportation control measures, and
• Retrofit of vehicles with lead traps (particulate
collection devices).
Fugitive emissions are expected to be significant around several sig-
nificant lead point sources such as primary and secondary lead smelters and
primary copper smelters. In general, the whole subject of industrial fugi-
tive emissions is currently under intensive investigation.47'48'49 The in-
formation in these references pertains to particulates, but could potentially
be revised to account for lead in cases where the source types, covered are
contributing to the lead problem.
The generic control technologies listed above are required by other
federal programs. RACT applies to retrofit controls on existing sources;
BACT ^LAER are new source controls and must be at least as stringent as
NSPS. ' EPA has established a clearinghouse for information on -determi-
nations for these technologies. As an aid in setting control requirements,
states considering either stringent retrofit controls or very stringent con-
trols for new sources may want to check with the EPA Regional Office for the
latest particulate information available from the clearinghouse.
In Sec. 211(c)(4)(C), the Clean Air Act provides that states may im-
pose limitations on the lead content of gasoline more stringent than the
federal requirements. This strategy can be considered as an alternative to
vehicle use constraints in areas where the existing federal program and sta-
tionary source controls fail to provide for a demonstration of attainment of
the lead NAAQS within the mandated time. This strategy may present substan-
tial administrative and enforcement problems and requires approval by the
EPA Administrator.
-------
85
Since vehicles are frequently a major source of lead emissions, ve-
hicle use constraints provide another set of strategies that may be consider-
ed. References 51 and 52 provide a list of such strategies. In particular,
Ref. 52 provides a list of reasonable transportation control measures (RTCMs).
Not all the strategies listed in these references would be effective for lead
so some care should be exercised in selecting candidate strategies for anal-
ysis. In addition9 vehicular lead emissions can be reduced by the use of land
traps as described in Ref. 24.
7.2 SELECTING AND TESTING STRATEGIES
There are too many strategies available to permit an agency to eval-.
uate them all. Reference 16 discusses a screening procedure that can be used
to select a set of strategies for detailed analysis. States need not apply
this procedure but consultation of Ref. 16 is recommended, because it pre-
sents the considerations which are important in selecting the strategies for
analysis.
Reference 16 also describes how to model the various strategies it
lists. The same procedures can be used for lead. With respect to NSPS,
however, it should be noted that additional material is available in Refs.
53, 54, and 55. Updated information on controls applicable to fugitive dust
sources is also available in Ref. 56. Control of reentrained road dust
is discussed in Ref. 57 and should be applicable to lead. Reference 24 pro-
vides information on the control of lead emissions from several source cate-
gories including lead traps on highway vehicles. These technological controls
would be simulated by modifying either the control efficiency, W, or the emis-
sion factor, EF (in the event of process change or modification), in Eqs. 2.1
and 2.3 as discussed in Ref. 16.
Controls on fugitive lead emissions from industrial processes would
be simulated by using the appropriate value for W in Eq. 2.1. References 24
and 25 contain information developed specifically for lead fugitives while
Refs. 47, 48, and 49 deal with particulates in general. The control infor-
mation in these three references could be used if lead-specific information
is unavailable.
RACT, BACT, and LAER determinations from the clearinghouse could pro-
vide useful estimates on available control technologies and efficiencies for
-------
86
both existing and new sources. Some LAER determinations might preclude cer-
tain emissive processes so that if LAER-level controls were being contemplated
for new/modified sources, the simulation might require changes in both the
emission factor, EF, and the control efficiency, W, in Eq. 2.1.
Reduced limitations on the lead content of gasoline can be simulated
by using appropriate vales of Pbn in Eqs. 2.3, 2.4, or 2.5. The value of Pb
used will depend upon the fraction of vehicles affected and the lead limita-*
tions being considered. Appendix C in Ref. 17 discusses the method used to
develop the pooled average lead content values used in estimating lead emis-
sions under the current federal program. The method presented there is
straightforward and can easily be extended to the case where limitations more
stringent than the federal are in effect.
Vehicle use limitations will generally be simulated by changing the
average daily traffic (ADT) or vehicle miles travelled (VMT) values in Eqs.
2.3, 2.4, or 2.5 to reflect the effects of the candidate strategy. The
Regional Office should be consulted for the latest information on estimating
the impacts of these measures; new guidelines are currently being issued and
states should attempt to utilize the latest available information. When use
of these measures is being considered, the local transportation planning
agencies should be involved as early as possible in the lead planning process.
These agencies should have valuable insights concerning the effectiveness and
ease of implementation of these strategies.
7.3 STRATEGY SELECTION
After the strategies have been selected and modeled, it is quite like-
ly that several strategies or combinations of strategies will be technically
acceptable, that is, provide a basis for demonstrating attainment of the lead
NAAQS. Additional considerations which may aid in selecting the most appro-
priate strategies for actual implementation include:16
• Economic analysis,
• Social impact analysis, and
• Institutional review.
Regulations do not require either an economic or a social impact anal-
ysis. Guidelines for states desiring to do an economic analysis are provided
-------
87
in Appendix D. Additional material is available in .Ref. 16 which also pro-
vides some assistance in evaluating social impacts.
.Certain institutional reviews are required by regulation. Aside from
the technical reasons for involving transportation planning agencies when
developing vehicle use constraints, the involvement of these agencies is a
regulatory requirement^ ' ' it is important in an area as complex and
sensitive as transportation that all agencies.cooperate in the development of
consistent and effective plans. Public hearings are required prior to the
adoption of any air pollution control regulations.50 This process and other
institutional questions are addressed very briefly in Ref. 16 but neither
this work nor the reference is intended to provide specific guidance for this
process.
These additional considerations should enable the agency to choose a
final set of strategies for implementation. If the analysis has been carried
out carefully, the appropriate agencies consulted, and the required comments
considered, the plan should be approved with a minimum of difficulty and re-
vision.
-------
-------
89-
APPENDIX A. BASE-YEAR ACTIVITY LEVELS FOR AREA,
FREEWAY, AND ARTERIAL SOURCES
The tables, in this appendix do not present the data in its original
form. The data in Table A.I is based on Ref. 58 and includes only the data
for those area source categories with lead emissions. Tables A.2 and A.3
present data based on freeway and arterial VMT and speed data for 1,714
traffic zones covering the example area and five nearby counties. These
traffic zones are subcounty areas defined, for planning and data gathering
purposes by the regional transportation agency from which the original VMT
data was obtained.
-------
90
Table A.I Base-Year (1975) Area Source Activity Levels3'b
Activity Level
(SGCUnits/yr)
Source Category
Residential Fuel Combustion
Anthracite Coal
Bituminous Coal
Distillate Oil
Commercial/institutional Fuel Combustion
Bituminous Coal
Distillate Oil
Residual Oil
Industrial Fuel Combustion
Bituminous Coal
Distillate Oil
Residual Oil
Residential On-Site Incineration
Of f -Highway Gasoline Vehicles
Fugitive Dust
Unpaved Roads
Land Tilling
C
4610
639070
206630
16490
140160
184540
503210
140480
388610
751890
22966
4668
184
County
D
10
1110
91410
1780
14550
19970
'45370
13510
37350
378550
21202
2481
163
L
Q
330
25120
770
6550
5840
56590
17230
47670
n.
1841
7612
193
Only categories with nonzero activity levels are listed. Highway
vehicle activity levels are given in Table A.2.
Compiled from Ref. 58.
-------
91
Table A.2 Example Area Freeway Link Lead Emissions
for Base Year 1975a
Link
No.b
1A
3C
4A
4B
6
7A
7B
7C
7D
8
9A
9B
10AC
11
12AC
18AC
18B
19
20
21
22A
22B
23
A f
24
25
26
27A
27B
27C
28
29C
30A
30B
31A
31B
32
33
Average
Vehicle Speed
(miles/hr)
57.5
52.7
57.5
55.5
50.4
48.2
40.9
38.9
35.9
45.3
45.6
55.1
57.4
54.7
54.2
57.1
43.5
53.1
50.5
48.6
55.0
56,3
56.5
40,5
48.1
51.2
37.2
41.7
47.2
5A.1
59.1
51.9
54.5
54.5
56.0
51.3
50A
VMT/day
100
3200
8651
12,164
8576
5532
14,039
8345
11,563
7578
5860
8613
6478
3693
2914
7256
7536
25,733
26,223
11,144
8197
5977
2202
416
26,138
21,772
15,110
8374
13,116
15,504
5596
1663
3812
3362
16,237
20,898
3124
28,421
Lead
Emissions
(ton/yr)c»d
5.70
3.94
21.3
14.1
7.88
18.8
9.20
12.1
7.26
7.28
7.36
10.5
2.10
4.69
5.75
5.85
30.5
40.3
15.9
11.1
9.70
3.72
0.707
28.2
28.9
22.0
8.33
8.75
20.3
8.85
1.88
5.66
10.3
25.7
35.0
4.56
40.5
3Lead emissions are total from passenger, light duty and heavy duty
gasoline vehicles. y y
For location of freeway link in example area see Tig. 2.2.
c
Lead emissions computed only for that portion of link in example area.
Based on VMT from passenger light-duty and heavy-duty gasoline
powered vehicles.
-------
92
Table A.3 Base-Year (1975) Tailpipe Lead Emissions
from Vehicle Activity on Arterial Roads
County
L
D
C
VMT/day
100
65,543
94,299
527,847
Lead Emissions
(tons/yr)a»b
271
390
2183
a_
Lead emissions calculated from passenger, light-duty,
a-.id heavy-duty gasoline powered vehicles.
Calculated using county-level VMT with average speed
of 19.6 mph for all vehicles26 and 70% lead exhausted.
-------
93
APPENDIX B. EMISSION PROJECTION MODEL
The model incorporates parameters affected by air quality control strat-
egies and exogenous parameters representing growth and replacement. Alter^-
ation of parameter values can be used to simulate various control strategies
and permits emissions to be projected in a manner consistent with a chosen
strategy. The primary symbols used are listed in Table B.I. Fig. B.I de-
picts the model schematically. The figure shows how emissions from a source
can be estimated. It is important to realize that the model is applied at n>-
the source (or source category) level: each individual source (emitting unit)
is assumed to grow and to be replaced at appropriate rates. Such an assump-
tion is clearly incorrect with respect to any individual source, for as time
passes a particular source is either still as it was originally or has been
totally replaced or new sources have been added; no source is partially re-
placed and partially new. What the model estimates is the total amount of
growth and replacement and the associated emissions over the ensemble of
similar sources in the inventory. These estimates are made by considering
each source individually. In the aggregate they may be expected to be reason-
ably representative of the emissions picture but for any specific source
they may be inaccurate.
Following Fig. B.I, consider a source category whose level of activity
at time tQ is PQ. As time passes, the activity level (i.e., the number of
production units produced per year) in this category increases linearly at r.a
the rate of GR per year. The total activity level at time t is
n
Pn - Po
Po X GR(tn - to) - V1
(B.I)
In addition, replacement has taken place linearly at the rate of RR per year.
At t the amount of existing (original) activity left is
P - P x RR(t
o o n
=Po[l-RR(tn- to)].
(B.2)
It is also assumed that at some time T such that t < T < t the policy atrali-
n — — n * J rr
n
cable to new growth and replacement construction changes. The growth in
*The development presented here closely parallels that found
in Ref. 59.
-------
94
Table B.I Primary Symbols Used in Emission Projection Model
Symbol
Meaning
Units
n
n
gf
"gi
ri
rf
EF
en
Ef
gi
EF
RR
GR
gf
Base year .
Projection year.
Year in which regulation applicable to
new/modified sources changes (t: £T£t ).
Total source category emissions in t .
n
Source category activity level or capacity
in t .
n '
Source category activity level or capacity
in t . (Capacity is assumed to be fully
utilized in t .)
o
Source category activity level or capacity
in fcn associated with growth between T and
t .
n
Source category activity level or capacity
in t associated with growth between t and
T. n °
Source category activity level or capacity
in t associated with replacement .between
tQ and T.
Source category activity level or capacity
in t associated with replacement between
Source category activity level or capacity
associated with existing (original sources
in t .
n
Emission factor for existing sources in t .
n
Emission factor in t for sources which are
new/modified in initial period between t
and T. °
Emission factor in t for sources which are
new/modified in f inaS period between t and
T.
Replacement rate per year.
Growth rate per .year .
n
tons/yr
production units/yr
production units/yr
production units/yr
production units/yr
production units/yr
production units/yr
production units/yr
tons/production unit
tons/production unit
tons/production unit
yr
-1
yr
-1
-------
95
g
O
H
cu
c
o
•H
j-i
CJ
(U
fl
o
•H
CO
CO
ta
•H
pt.
-------
96
activity level that took place prior to the time T is P . and the growth i
activity that took place after T is P ...
in
- T) and Pg± -
x GR(T - tQ).
(B.3)
Similarly, replacement activity can be divided into two parts before
and after T:
Prf = Po
~ T) and Pri = PO
- tQ). (B.4)
The total activity in the category at time t has now been separated into five
classes: P^f, Pg±, P^, prft and Pen, representing activity from growth since
T, growth before T, replacement before T, replacement after T, and the re-
maining existing original sources, respectively (see Fig. B.I).
Emissions can now be calculated by multiplying each of these activity
levels by an appropriate controlled emissions factor EF. In general, emis-
sions Q(tons/yr) can be found by multiplying the activity level (production
units/yr) by an emissions factor EF (tons/production unit):
Q - P x EF. (B.5)
It should be noted that the emission factor as used here includes both
the uncontrolled emission factor and the emission reduction due to any con-
trol devices. Figure B.I shows the emissions factors appropriate to the five
classes of activity in the year tn. EFgn applies to existing (original)
sources and can be changed to represent changing retrofit control require-
ments. EF applies to all growth and replacement construction since T.
^g± aPPlies to all growth and replacement construction before T. Total
emissions in t^ can be calculated by repetitive application of Eq. B.5 to
each of the activity levels in Eqs. B.2 - B.4 and summing the results:
Qn
[Prf XEFgf]
Using Eqs. B.2 - B.4 to substitute for the production activity levels:
po[i - m(tn - to)] x EF
+ P x GR(t - T) x EF
o n gf
+ P x GR(T - t ) x EF .
o o' gi
+ P x RR(t - T) x EF „
on gf
-------
97
•+ P x RR(T - t ) x EF .
o o gi
or
+ [PQ x EF ±][(T - tQ)(GR + RR)]
+ [PQ x EF f][(tn - T)(GR + RR)]
where similar terms have been gathered together.
(B.6)
For allocation purposes, new and modified (growth and replacement)
sources are frequently treated differently than existing sources and it is
desirable to allocate Q between Qe from existing sources and Q1™1 from new
n ' ii _ iR.
and modified sources. As noted in Eq. B.2, existing source activity at t
is PQ[1 - ^(^ ~ t )] corresponding to the first term in Eq. B.6. Thus,
= ' {P x
o
en
- RR(t - t )] and
n o
t )(GR + RR)]
EF f}[(tn - T)(GR + RR)].
> (B.7)
The factors in eutjly brackets (PQ x EF^, PQ x EF ±, and PQ x EF f) represent
the emissions that would occur in the baseyear t if the control requirements
corresponding to these three emission factors were fully met in that year.
It is frequently convenient to rewrite Eq. B.7 in terms of the ratios of the
controlled emissions in the ye-r n to the base-year emissions. If EF is
eo
the base-year emission factor for the source category being considered, the
base-year emissions for the category, Q , are given by
0 = P x EF
o o eo
(B.8)
and three ratios may be defined
R = EF
en
R. = EF
i e
/EFeQ, the ratio, on a per unit of production
basis, of emissions from sources emitting
at year n retrofit levels to emissions from
sources emitting at base-year levels;
the ratio, on a per unit of production basis,
of emissions from new/modified sources built
before time T to emissions from similar
sources emitting at base-year levels; and
the ratio, on a per unit of production basis,
of emissions from new/modified sources built
between T and t to emissions from similar
sources emitting at base-year levels.
(B.9)
-------
98
Combining Eqs. B.7, B.8, and B.9 gives
t)] and
x
t)(GR+ RR)]
R [(t - T)(GR + RR)]
It
of
should
both
(B.10)
be noted that the emission factors, EF., could include the effects
±.
process changes and control device efficiencies. In general
EF
id
_ n±j/100)
(B.ll)
where EF^.. is the uncontrolled emission factor for the source category or
process being considered and n is the appropriate control efficiency. If
EF
is the
no process changes are contemplated as is frequently the case
same for all categories ij of source activity and Eq. B.9 can be rewritten
?nc
EF
r,unc
R
en
EF (i - n /ioo) (i - n /ioo)
EF
'en
eo
UnC
EFUnC(l -
- /100)
0
Thus, if no process changes occur,
- neo/ioo>
R.
(i -
and
(B.12)
(i - neo/ioo)
(i - ngf/ioo)"
Rf = (i - neo/ioo) '
where neo, Hen, ngi> and ngf , respectively, are the control efficiencies re-
quired on original sources in the base year, existing (original) sources in
the projection year n, growth and replacement sources built between t >.and T,
and growth and replacement sources built between T and t .
n
-------
99
APPENDIX C. HATREMS EMISSION FACTORS .;-,.,-.
This appendix contains HATREMS emission factors for lead. It should
be noted that the list does not include each SCC for boilers. Only one SCC
per fuel type is listed since all boilers using the same fuel type would use
the same emission factor. For example, the lead emission factor for pulver-
ized bituminous coal-fired industrial boilers (SCC 1-02-002-01) would be "
.00160, the same as the lead emission factor for SCC 1-01-002-01 which applies
to all bituminous coal-fired boilers. The same default multiplier, 8,300,
and control efficiency mulitplier, 100.00, would also apply to both SCGs if
HATREMS values were to be used in the emission calculations.
-------
100
I- O f-
ffi -I
«« Z UJ
-I U) _l -.
CO
Q)
a
IM MM
OM O <
HO* r-CL r»
a. i-« a.
k r- «a r-
Or- or- w r-
•H
CO
CO
O O UJ O
UJ O UJ O
Z O iu O
MO h* O
r t- M«-
l O _JO _l
05
« r- < O
a: o a: o
ui o uj r-
O
0)
i o o o
o o u <
u o u o
UJ O uJ O w O 'i-f O
w o oo urvj wO u<
JO uj a
Md. MiD M
X t-l Oj K4 OJ **
o oo «- a
OIM orxj
ru orvj
T- Ir-
a oo o co oaa
o < o <
-------
101
UJ
ac
o
ac
o
9 =
ui ac
=» -I
o a
a
UJ
u
a
o
ac
a.
v>
z
o
a
o
o
rr
o.
in
z
c-
a
o
oc
a.
v>
z
o
o
ct
o.
o
I-
o
ac
a.
a
o
ac
Q.
Z
o
a
o
oc
a.
Ui
z
o
z
o
o
I Z
V9
Z
O
ui:
u
z
a
o
z
o
=
a
o
z
o
o
ac
a.
v>
z
o
o
u
Z
O
z
UJ
Z
in
z
o
o
z
z
o
4-1
G
O
i-J
•
U
(U
iH
•3
H
a: z
O 3
Z O
o v>
V) UJ
1-1 a.
3 C
O UJ
o o
a. a.
o
a.
c
u. o
U.
u, O
a
o
a
^1
vt
UJ
u.
X
ir»
u
Z
u.
Z
UJ
a.
3 O O
9 O
5 O O
3 -10
•c
or
UJ
41k
_J
<
h-
Uj
JL
Z --
O
u
_J
V)
w
z
z
ul
a.
o o
0
0 O
JO
<
CK
UJ
u.
UJ
z
<
u?
z
•c
-"- X -
o
<_)
-J
to
u
z
u.
2
UJ
a.
o o
o
c o
-J O
•c •
oc
tu
u.
N6ANESE
«;
£
CK
OVFNCF
Jt
UJ O
v> 0
>• o
o'o
«e
oe
UJ
u.
UJ
QC
0
1
u
z
u.-
H
<
CD
Z
0
u
3 0
a o
a o
o. o
z
o
OC
M
^
(D
«t
U
z
„ 1^
h-
«
O
Z
o
u
30
0 O
CC O
a. o
z
o
ac
1-1
z
ui
<0
a
X
M ,
CC
UI
Z
v»
z
0
u
3 a
0 O
cc 0
a. o
z
o
CC
IH
_l
X
O
X
K
ac
«t
UJ
z
o
z
o
l-»
3
0 O
00
a. o
0
UJ
UJ
h»
*/>
X
o
z
X
K
<
UJ
z
o
--z
o
t>4
3
o a
00
a. 0
o
u
UJ
h-
v>
_l
•e
ac
UJ
Z
UJ
UJ
u.
CD
Z
o
M4
=1
O O
o o
a. o
o
UJ
UJ
>/-
v>
_l
X,
•M
u
oc
^«e .
t-
UJ
UJ
z
o
M
=)
00
oo
0.0
o
UJ
UJ
1-
v>
.1
c>
z
u
cr
1-
Ul
UI
z
o
IH
a
00
o o
a. o
o
UJ
UJ
h-
V)
U)
z
a:
UI
z
V)
UJ
t- o
_l 0
It 0
v> o
A
•t
UJ
_J
UJ
u
<
z
ac
a
i-
«c
m
UI
t- o
-ID
at o
o
a
«
UJ
u
FURNACl
IB
at
>
ac
at
UI
t-
•j
C
M
O
f
UJ
J
o
ui o
-JO
O =>
UJ <
.< u.
u a:
wo
O
o
o
CC O
e. o
VI O
U •
OL
<
K O
a. »-
u o
o •
a.
_i
o
o
V) O
-J O
o o
VJ O VJ O
-JO -JO
000
V) O V> O VJ O
_IO U O -IO
CC
«t
E
o
VIO
VJ O
_i o
O
VI O
o
V) O
a. c
VI O
u ir\
o •
a: o
& o
vi o
0. O
0.0
v> *-
UJ O
u o
o •
CC
a.
CC O
0.0
VI f
UJ Kl
o •
a. cc
a. a.
-I _i
t-
V)
K O
e. o
vi m
uo
o •
CC O Cc O
0.0 0.0
V) O VJ O
UJ •* UJ N.
UP- u a
o • o •
cc ac
a. a.
cc
h»
VI
3
K O
a. o
vio
ul O
O (M
O •
ac o ac o cc o
a. CD a. o a. o
V) C3 V) O V> O
Ul 09 UJ Q Ul O
U «" U O WO
o • o « o •
cc ac u\ ac in
a. o. o a. KI
cc
«t
c
(•»
o
in
o
I
•o
00)
O rvi
I r-
K,«\J
O «
I
O co
O (VJ
K! (M
O «
I
I
90
O 00
O (VI
I V
(M
O
I
QO
O CO
o rsi
I r-
o
I
CO
o
-------
102
UJ UJ
£ JL
to (o
K h-
*-< r-4
Z Z
U UJ
ec M
3 -I
c a.
in
z
o
0
LU
to
LU
C
W
z
o
1-
0
UJ
to
CO
UJ
O
W
z
o
H*
a
UJ
w
j^3
UU
O
Z
O
UJ
ae
CO
z
o
h-
UJ UJ
VS 10
ae CK
X X
,
yj ,
06 < <
z z w
o o o
a
UJ UJ
ae ae
x x
w o
-J «J
z z
0 0
t- h-
o
UJ
ce
X
cv
-J
z
0
u.
UJ
10
ae
x
cv
ac
w
z
0
k_
Ct
UJ
CO
to
UJ
0.
w
z
o
S PRODUCED
UJ
cc
UJ
h-
H
CO
O
to
O
S PRODUCED
LU
cr
LU
r-
u ui
>- a.
x
*u a
O
CL
O
u. o
I- O h-
a: o
CO O
z
o
wo
ul •
E O
v> o
"I «-
ac
O
X
h- O
JO
UJ •
01
ae
UJ **
o
a.
3
t/t t-
W
m
19
1
u>
z
h- O
-JO
r o
CO O
S/BRON
100.00
30
O O
OJ O
-* o
0
z o
o »
IE O
11 O
O
z o
o •
CC O
l-l O
o
z o
o •
a: O
It O
«J O
u) •
IE O
CO 0
I- O
-I O
K O
to O
It O
wo
C
O
UJ O
(- O
« 0
03 O
o
0
o
0
u>
U K
U O
a
o
o
UJ
O
WO
-j a
UJ
ac
v
cc
If
iL PROCESPR1H,
.13000
«*
£C
C/>
»-4
O
to o
-JO
Lu
>.
ce
*•
>L PROCESPRIHi
19.00000
«.
Cc
to
M
N
O
WO
-1 O
UJ
>_
cc
L PROCESPRIMi
1.ZOOOO
5
cc
w
t-t
^
o
w o
-JO
« •
r- t"
UJ
JC
or
L PROCESPRIM/
2.20000
t-H
CC
w
QC
m
u o
* o
t- o
ku •
SL •-
ii.
L PROCESSECOK
5.00000
5
a
to
ac
m
to
-JO
« o
h- o
UJ •
*• *"
O»
L PROCESSECOH
4.20000
<
ac
v:
&
u>
VJ
_! 0
« O
1- O
uu •
V «-
>•
&c
o
L PROCESSECON
.60000
«s
a
a.
. PROCESSECON
.07000
«
a
tsi
ac
03
_l O
« o
h- O
E. «-
fit:
- PROCESSFCON
.05000
— <«
•-*
CC
h*
«
UJ
-I
_J O
< O
H- O
4. r-
CC
Jtt
. PROCESSECONI
.18000
_l
CC
h-
<
J
W
-J O
< o
h- O
ae ^
CK.
,e
. PROCESSECONI
34.60000
•j
Ce
H>
«
lj
(^
-JO
«o
>- o
C I-
IX
PROCESSECONI
45.00000
;j
a:
h-
•x
UJ
_l
(O
-J 0
« o
1-0
u. •
CC
PROCESSECOND
.44000
«
ac
K
«
UJ
«J
V)
-1 0
« o
K 0
UJ •
X. t-
tc
PROCESSECOND
.88600
j^
M
t-
f
UJ
-1
«
H
a.
cc
PROCESSECOND
-J
£
1-
79
03/
o
LJ (_)
O
Q.
O
eo «• co
(M o rg
T- I »-
fNJ K) (M
« O <
I
O
*O OO
O (M
I «-
O <
I
a
z
0
Kl CQ
O f\J
1 f-
K) rw
0 -C
I
o
I
o
r*\ oo
o ru
u^
o
I
1M
O 03
O fVj
I •-
•» 1\1
o <
I
r
O rg
I «-
^ rg
a «
i
o no
o rg
I T-
o <
I
a
t
o
f
-------
103
f. m.
3
0
oc
a,
v>
UJ
M
Or
3
0
ac
a
in
Ui
M
OC
a
UJ
u
3
a
o
ac
a.
i-
a
UJ
u
3
a
o
ac
a.
a
UJ
w
o
oc
a a
UJ UJ
oc ec
a
UJ
o
a
O
BC
a
iu
3
a
o
oc
t-
a
UJ
u
3
O
ac
a
u
3
O
ac
a
u
0
UI
Z
3 _)
O O.
in t-i
ec z
o 3
»*-
U UJ
« u
u. oc
3
z o
O in
UJ
z
c
z
u.
o
in
z
o
M
in
Z
o
z
QC
UI
X
IH
X
tfe
o
CO
z
o
o
ac z
IU
a.
o
in
u
w
o
cc
a.
in in
-J _l
UJ UJ
or ac
ac ac
0 0
z u z
UJ
m
a
Z
oc
U)
^«
E
U
1/9
Z
0
O
z
a
UJ
M
Ik
_J
M
O
E
UJ
in
Z
c
c
z
a
UJ
M
U.
...
U)
E
UJ
in
z
o
o
z
o
UJ
a:
u.
^
w
• o- • • •
w
V) to
ac ac
oc oc
z z
i «i
UJ
oc
UJ
z
M
OC
IB
E
UJ
V3
Z
o
z
0
z
UJ
oc
hk
.J
M
O
M«
C
UI
in
z
o
z
o
z
UJ
ac
u.
10
UJ
E
Z
O
Z
0
Z
a
UJ
OC
M
M^
o
u
u
o
en
z
o
z
o
z z
M
H
<«
O
UJ
-J
or
^
t~
_j
V)
z •
o
z
o
z
u>
z
M
1-
.J
UJ
- E -
z
oc
3
m
in
*
O
a:
UJ
CD
E
«
X
u
- Uj
•w
4-J
-J
*J
a.
Oc
z
V)
z ^>
in
z «-
-j
i
U)
a «-
<
Iu
a «-
«
UJ
O
O
u- o
K •
tO O l£> O
U-O U. O
* • X •
o o
1-0 (- o
(A O
U. O
E •
O
u o
3 O
<->
V)
I-O
UO
30
a •
Of*
I- O
u O
3 O
UJ
c
UJ
o «-
OC
a.
c
I-O
i-a
u a
30
O *- Mr-
O
<-> O
O
O
MO
(-0
Z O
MO
f- •
HO
ui o
MO
MO
3
E
u ac
u o
Z
o
u
L PROCESSE
.05500
«
M
ec
h-
3
a
z
•— dj
Z
O
u
L PROCESSE
.73000
«
M
OC
t-
3
a
X
oc
UJ
z
L PROCESHI
.02060
^
M
OC
t-
in
3
a
-z
ac
UJ
z
M O
E m
in 1*.
UJ O
u O
o •
ac
a.
^
M
flj
H
3
a
z
ac
UJ
z
. PROCESHI
.11000
••rf
«c
K.
A
Z
£
z
PROCESHll
.11000
«e
£•
W5
3
Z
OC
or
PROCESHll
.11000
— ?
OC
3
A
Z
m
fie
»
PROCESHll
.01875
«J
oc
u
3
2
•c
OC
w
PROCESHI*
.00375
•J
«C
OC
V)
3
o
z
OC
iu
PROCESHIl!
.10000
_i
«e
a
VI
3
Z
a
UJ
PROCESHIA
.10000
«J
<
O:
v>
3
z
OS
UJ
PROCESHIN
.10000
j
ac
fr»
3
3?
Ml
PROCESHET
.50000
_j
cc
p-
VI
3
A
•a*
£ 2
Z UJ
PROCESPRI
.25000
: GOV
Ud
-I K
oc 2
1-
in a
a _i
O
o.
00
n
o
fw
O 00
O AJ
I f
in ru
o «
I
I
o
-» CO
O (M
J »-
O- IM
O <
I
O co
O IM
I f
O IM
•O <
I
O
I
t—
O
o
I
^
o
i
m
-------
104
t- Z.
*tJ u. t-
•. u I-
4J a -i
u- O _io
"- • < .
•J *~ w »-
O „
m oj
-------
105
•• S s | « g
•j Hj , ^
uj ' < 3! j -j
z 5 o u? 33
*?* OC *» ^
- *- " §S oo
*> " ui ,» • z Z
S - - » ^ I 1 55
« 3 §- g. g L f g 3 § g
a rit w; " * Sri ^ g. g, *j x-
0 "• z z « <=-J
W *v UJ O *""*
* " W) fcl UJ O
O UJ W IM
iflt-oE 3 -1 «_i
Gjuocz '* ^t S
S<3= ri= -J3
•^""S a i i •- « '• j _, K S
H o « . „ 2 2 S S S ° " 2
O •- Ul ---j . ° « 0 (j 0 ~
h">- = £ £ S f * «: a: c
S -5 ' "• * t fc t t l|
« §" ' ".''"• —
HW O V>
in d
fi z
0 * =
T! « a- -J
CO « 5
co £
21 g'lliiiiss
i 2 s
z z o o • » «« =
o o ~ x -i j£ 5 ' * z °
*• '° Sri S g £ £ i
O ,J *"
ui o
h-
« «l
-1 <
•J a
o S 2 2
u w a or.
«. t «• ,- S " >u
t 2- * £ ' 5 o S
»• o- o. a. z z S
i I I I I i §
"• °> - - ^ ^
S J § § s§ z§ g
o o o S rS °° z o
«» § § § '55 ?§ 55 ' 2
. * ' •*•«•!-."
«C _i *" •" f U
0 M Z UJ OC «
U 0 , ^J Z UJ X
"•* w *^ Z uj
•^ z »_j ^ ir
S S ° - * 3
O ^ ^^ SC UJ
z 5 " « z
S j i- uj 5
it ri i- i- ui j
5 £ •j1 « H. o
E ~ ES £ §£ |s fs i
-J _l < ° *•
s Hs Hs s js •?* z°° »
r- *•
"*"*«« 5 2 o
^
V" IM _ "^ IM Ki f>
*^ "<
-------
106
4J
O
O
0)
-------
107
APPENDIX D
GUIDELINES FOR ECONOMIC IMPACT
ANALYSIS FOR LEAD SIP
-BY
Allen C. Basal a
n«- Economic Analysis Branch
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
and
Susan Karash
Energy and Environmental Analysis, Inc.
March 16, 1979
-------
-------
109
GUIDELINES FOR ECONOMIC IMPACT-ANALYSIS OF LEAD SIP
INTRODUCTION
of a
o Design alternative strategies and develop costs;
o Evaluate cost effectiveness of alternatives;
Each of these steps will be discussed below.
states may elect to use mode' "ants as a basil far ^,t" aljf?ls *re unavailable,
situations, it should be estimated tn fo
differ from the mode? "ant estimates.
DESIGNING ALTERNATIVE
a basi far ,t
nl for,coft estimates
9 " Plant Spec1flc
In such
in orJer l?l?M™*
'ffe-"* stationary sources
of
a detailed source emissions invent2rv fc f Aspersion models, preparing
step which is also nece sary In SSlteoJ dL?™6S?ar^fir?I Step* A
methods which are technically feas^p
a basic source of such information
*
, is identnfying the control
ntro1 Techn1^es Document24 i
°ntro1
not address control of mobile source lead emissions.
-------
110
PROGRAM COSTS
For every alternative control strategy (achieving the standard) the
control costs for each source are estimated using primarily the Control
Techniques Document.24 Other cost references such as the GARD Manual63
may also prove useful. The cost estimate for a facility includes an incre-
mental investment cost and an annualized cost. In addition, an estimate of
incremental administrative costs should be made.
Incremental investment costs reflect incremental costs (for emission
reductions beyond the level of any other emission regulation) which occur
only once, such as the design, development, purchase, and installation of
pollution control systems.
The annualized cost is comprised of three components: 1) direct operating
cost, 2) indirect costs, and 3) recovery credit. The first component, direct
operating cost, includes operating and maintenance costs such as labor, utilities,
and materials needed to operate and maintain the equipment. The indirect costs-
include additional administrative overhead, property taxes, insurance, and
the annualized capital charges for depreciation and interest. The annualized
capital charges for depreciation and interest are computed using a capital
recovery factor which depends on the life of the equipment and the cost of
capital. The annual allocation for administrative overhead, taxes, and
insurance is usually estimated as a fixed percentage of the installed capital cost.
In some cases, valuable material is reclaimed from the control device. In
these instances the recovery credit accounts for the value of the reclaimed
material. The annualized cost is then obtained by adding the direct operating
costs to the indirect costs and subtracting any recovery credit.
It is also necessary to estimate the administrative costs associated with
the alternative control strategy. The incremental administrative costs reflect
increased needs for data gathering, monitoring, enforcement, and management
activities resulting from the control program. This is best estimated by the
state based on experience in administering existing plans and previous revisions.
COST EFFECTIVENESS
With control strategies designed and program costs estimated, a comparison
of costs among strategies, cost-effectivenss, can be made. An evaluation of
cost effectiveness is needed to identify the initial least cost solutions. The
optimal (minimum cost) program is one where the marginal costs of control for all
sources are equal. However, what is initially identified as a least cost strategy
may not be least cost when plant specific economic effects are assessed. Further-
more, the least cost strategy may be unacceptable given the associated distribution
effects. Consequently, the strategy selection process should give preference to the
least costly control strategy (considering plant specific economic effects) which
attains and maintains the ambient standard with acceptable distribution effects.
PLANT SPECIFIC ECONOMIC EFFECTS
After the initial least cost solutions are identified, the plant specific
economic effects are assessed. The main question asked in a plant specific economic
effects assessment is: Will control requirements preclude the existence of this
plant? To wit: Is control economically feasible; affordable? In answering this
question two related issues should be addressed. The first is the availability of
-------
Ill
financial capital. The second is the price/sales/profitability of the
plant after control.
Capital Availability
An important issue to consider is whether firms have adequate financial
resources to make the capital outlays required for emission control In
nf rLTf; can-they raise t!?e "ecessary capital? If so, does the allocation
of capital to emission control projects restrict the firm's ability to fund
other productive investments, thereby imposing significant opportunity losses?
To address these questions one must consider the firm's capital structure
and financial resources. Internal sources of capital are equal to after- tax
profits less dividends, plus depreciaton, less additions to working capital
External sources of capital include debt (borrowing and bond sales), and '
issues of new stock.
The analysis should compare capital requirements with total resources in
nSS^^'^K JV?ternal/?sources* the annua1 caPital requirement would
be compared with total annual internal resource projections. The impact may
be measured in terms .of .the change in the firm's debt ratio measurers the -
Iytlnf°tn 11°"?-term debt to long-term debt plus equity. This would indicate the
extent to which a firm's debt position would worsen if the firm wished to
finance emissions controls while continuing all other planned investment
projects.
*mn -r S?a11 *hange in a firm's debt P°sl'tion could represent a significant
impact. The investment community evaluates firms on the basis of their
expected earnings, the stability of their earnings, the growth in sales, and
the riskiness of the firm in light of its debt repayment burdens in relation to
its resources. Firms may try to keep their debt ratios low and reserve debt for
applications such as boosting sales or entering new markets. Hence, the
opportunity costs of even a small change in a firm's debt position may be
considerable. J
Pri ce/Sal es/Prof i tabi 1 i ty Effects
' The second issue addressed in determining economic feasibility is-
Will prices, sales, and profits be such that after control the plant's'minimum
iSrfSX IS r?-J ?VeHUrVS achieved? In addressing this issue considerations
include the likelihood and consequences of full control cost absorption versus
f-fl n? y complete control cost pass back or pass forward. Characterization of
the plant s market and industry in terms of demand and supply elasticities will
aid in identifying the likelihood and consequences of the various cost shifting
o
For example, full cost pass through with increased prices and no output
nnnS-5 woul <* "?6 ^^ where demand was highly inelastic and there were
no competing suppliers Where the potential for substitution is great (demand
rnn^ni ioctIC)> ??? ^T ^l™* **% te a more 11kel* "sumption. Absorbed
control costs could lead to plant closure and concomitant unemployment.
However, this need not always be the case. ipiuymeni.
-------
112
Resource i-.-.g '.nrenients to Do the Analysis
"' "'
ne nateant, tho assessment can be as thorough as
.
have already been developed. For example, the Economic Impact
IS
DISTRIBUTION EFFECTS
^^
om i Effe$ts information need not be limited to control cost orice and
employment. Geographic, income, and productivity effects can aTIo HP
SSS^SAfSSA-
OTHER STRATEGIES
addition, the distribution effects of these other strategies should a U
determined. As mentioned before, the strategy selection process shoJld
preference to the least costly control strategy (
^ -1"^^i
Cn'terion that 1s generally accepted is the "polluter pays
-------
•<•:
i* " * '
113
REFERENCES
1
'
2.
3.
4
5.
6
..."
7
Guidelines for Air Quality Maintenance Planning and Analyses, Vol. 1.
Designation of Air Quality Maintenance Areas U.S. Environmental
Protection Agency Report No. EPA-450/4-74-001, Research Triangle
Park, N.C. (September 1974).
, Vol. 2: Plan Preparation, U.S. EPA Report No. EPA-450/4-74-002,
RTP, N.C. (July 1974).
_ f vol. 3: Control Strategies, U.S. EPA Report No. EPA-450/4-74-003,
RTP, N.C. (July 1974).
Vol 4: Land Use and Transportation Considerations, U.S. EPA
No/EPA-450/4-74-004, RTP, N.C. (August 1974).
, Vol. 5: Case Studies in Plan Development, U.S. EPA Report No.
EPA^450/4-74-006, RTP, N.C. (December 1974).
Vol 6- Overview of Air Quality Maintenance Area Analysis,
u7sT~EPA Report No. EPA-450/4-74-007 , RTP, -Nv&.~ (September 1974)— ..... "»
Vol. ?• Projecting County Emissions, Second Edition, U.S. EPA
No. EPA-450/4-74-008, RTP,' N.C. (January 1975).
Vol 8: Computer-Assisted Area Source Emissions Gridding Pro-
* U.S. EPA Report No. EPA-450/4-74-009, RTP, N.C. (September 1974)
,,
'
15
} Voi. 9; Evaluating Indirect Sources, U.S. EPA Report No. EPA-
450/4-75-001, RTP, N.C. (January 1975).
Vol 10 (Revised): Procedures for Evaluating Air Quality Impact
-5TS& Station^ Sources, U.S. EPA Report No. EPA-450/4-77-001, RTP, N.C.
(October 1977).
vol 11- Air Quality Monitoring and Data Analysis, U.S. EPA
Re^o7t No/EPA-450/4-74-012, RTP, N.C. (September 1974).
vol 12- Applying Atmospheric Simulation Models to Air Quality
Ma^ten^e Areas,™ .< i .EPA Report No. EPA-450/4-74-013 , RTP, N.C.
(September 1974) .
Vol 1S. Allocating Projected Emissions to Suloeounty Areas,
Report No. EPA-450/4-74-014, RTP, N.C. (November 1974).
v t 14; Allocating Projected Emissions to Subcounty Areas -
lcesA and B, U.S. EPA Report No. EPA-450/4-74-Ol4a, RTP, N.C.
(March 1975).
Accounting for New Source Performance Standards in Proje cting
AlUcatin^ Emissions - Hypothetical Example, U.S. EPA Report No. EPA-
450/4-7 4-Ol4b, RTP, N.C. (October 1975).
-------
114
16.
drill-, p.P.,eJ.F. Tschanz, A. E. Smith, R.F. Freeman, J.E. Camiaiomi
nd V. R- i Air Quality Analysis Workshop Volume I - Manual, U S EPA
Report Nr DA-450/3-75-080a, RTF, N.C. (November 1975).
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
. ^ G tidelines for Lead Implementation Plans, u.S EPA Report
No. EPA-450/2-78-038, RTF, N.C. (August 1978).
40 CFR 51, Implementation Plans for Lead, 'National Ambient Air Quality
Standard, 43 FR 46264 (October 5, 1978).
Guidelines for Compiling a Comprehensive Emission Inventory, u S EPA
Report No. AFTD 1135, RTF, N.C. (March 1973).
40 CFR 50, National Primary and Secondary Ambient Air Quality Standards
for Lead, 43 FR 46245 (October 5, 1978).
Technical Manual for the Measurement of Fugitive Emissions: Upwind -
Downwind Sampling Method for Industrial Fugitive Emissions, U.S. EPA
Report No. EPA-600/2-76-089a, RTF, N.C. (April 1976).
Technical Manual for the Measurement of Fugitive Emissions: Roof Monitor
Sampling Method for Industrial Fugitive Emissions, U.S. EPA Report No
EPA-600/2-76-089b, RTF, N.C. (May 1976) .
Technical Manual for the Measurement of Fugitive Emissions: Quasi-Stack
Sampling Method for Industrial Fugitive Emissions, U.S. EPA Report No
EPA-600/2-76-089c, RTF, N.C. (May 1976).
Control Techniques for Lead Air Emissions, 2 vols., U.S. EPA Report No
EPA-450/2-77-012, RTF, N.C. (December 1977).
Zoller, J.M., G.A. Jutze, and L.A. Elfers, A Method for Characterization
and Quantification of Fugitive Lead Emissions from Secondary Lead
Smelters, Ferroalloy Plants, and Gray Iron Foundries, U.S. EPA Report
No. EPA-450/3-78-003, RTF, N.C. (January 1978).
Compilation of Air Pollutant Emission -Factors, 3rd ed. with Supp 1-8
U.S. EPA Report No. AP-42, RTF, N.C. (May 1978).
AEROS Manual Series Volume II: AEROS User's Manual, U.S. EPA Report No
EPA-450/2-76-029, RTF, N.C.
40 CFR 51, 53, Ambient Air Monitoring Reference and Equivalent Methods
for Lead, 43 FR 46272 (October 5, 1978).
Maxwell, C.M. and D.W. Nelson, A Lead Emission Factor for Reentrained
Dust from a Paved Roadway, U.S. EPA Report No. EPA-450/3-78-021 RTF
N.C. (April 1978).
de Nevers, N. and R. Morris, Rollback Modeling - Basic and Modified,
paper 73-139 presented at Air Pollution Control ASsoc. Meeting Chicago
(June 24-28, 1973). S S
Note: This reference has been reproduced as an Appendix G to Reference
•*- ' •
-------
115
31. Guideline on Air Quality Models, U.S. EPA Report No. EPA-450/2-78-027,
Office of Air Quality Planning and Standards Guideline No. 1.2-080,
RTP, N.C. (April 1978).
32. Standard Industrial Classification Manual, U.S. Government Printing
Office Stock No. 4101-0066, Washington, B.C. (1972).
33. Langford, T.W., Regional Air Quality Maintenance Planning Employment
Projections 19?'5-2000', Chicago, IL3 S.M.S.A., unnumbered Illinois
Environmental Protection Agency Report (March 1978).
34. Commonwealth Edison Co., personal communication (1978).
35. Illinois Population Projections (Revised 1976): Summary and by County
1970-2025, unnumbered State of Illinois Bureau of the Budget Report,
Springfield, 111. (July 1976).
36. Hooper, T.G. and W.A. Marrone, Input Variable Calculation Sheets for EPA-
450/3-76-017, Volumes a-f, U.S. EPA Report No. EPA-450/3-76-01, RTP, N.G.
37. Hooper, T.G. and W.A. Marrone,' Impact ofTew'Source Performance Standards
on 2985 Emissions from Stationary Sources, U.S. EPA Report No. EPA-450/3-
76-017, RTP; N.C. (April 1977). ~
38. Memorandum of Understanding between the Department of Transportation
and the Environmental Protection Agency Regarding the Integration of
Transportation and Air Quality Planning, Washington, D.C. (June 14, 1978).
39. Transportation - Air Quality Planning Guidelines, issued jointly by the
U.S. Environmental Protection Agency and the U.S. Department of Trans-
portation, Washington, D.C. (June 1978).
40. Environmental Protection Agency Appendices to Transportation - Air
Quality Planning Guidelines, (June 1978).
41. User's Manual for Single-Source (CRSTER) Model, U.S. EPA Report No.
EPA-450/2-77-013, RTP, N.C. (July 1977).
42. Tyler, D.D., Supplementary Guidelines for Lead Implementation Plans —
Corrections, memo to U.S. EPA Air Branch Chiefs, Regions I-X (December 21,
1978).
43. Workbook for Comparison of Air Quality Models, U.S. EPA Report No.
EPA-450/2-78-028a, RTP, N.C. (May 1978).
44. Workbook for Comparison of Air Quality Models - Appendices, U.S. EPA
Report No. EPA-450/2-78—28b, RTP, N.C. (May 1978).
45. TRW Systems Group',""~Air Quality Display Model, National Technical Infor-
mation Service No. NTIS PB 189194, Springfield, Va. (November 1969).
46. Clean Air Act, 42 U.S.C. 1857 et seq.
-------
116
47
48
*"
50
51 "
52.
*>' ^-600/7-77-148,
ion, Adoption, and Submitted, of
, 36 FR 22398 (November 25, 1971 et seq) .
ion Controls, preamble,
on Requirements for flonattalrment Area Plans, revised ed
U.S. EPA Report No. OAQPS 1.2-103, RTF, N.C. (April 1978). '
SovL?er26,'
54
ir'pon* °f P^f°*man°e ^ New Stationary Sources
ir Pollution - A Swmary of Regulations, "Journal of the Air Pollu-
tion Control Assn.," 20(11):1055 (November 1976).
Monarch, M.R., R.R. Cirillo, B.H. Cho, G.A. Concaildi, A.E. Smith
1116^ J.7K'L%Brubaker> Priorittes for New Source Performance
55 '
?).
P°" H0- °AQPS 1-2-°71' E", B.C. (October
57 •
' °-s-
j,
^
l Re^°n> draft reporTprSred for
the U.S. EPA, Washington, B.C. (November 17, 1978).
60. Jurbaker K.L. P.., Brown, and R.R. Cirillo, Addendum to the User's Guide
7^015! S^4*"8*11 "^^ u*s- EPA Report No-
5> T °{ ^ Specialists' Conference on tne EPA Model-
EP Modli r unnumberedjeport of the Specialists' Conference on the
EPA Modeling Guideline, Chicago (Feb. 22-24, 1977) prepared under Inter-
Agency Agreement No. EPA- IAG-D7 -0013. ' inter-
-------
/JW TECHNICALREPOR-;
(Please read instructions.™ the reverse before c,oj.
Development o* an Example Control Strategy for Lead
4. SPONSORING AGENCY CODE
• ORGANIZATION NAME AND ADDRESS
Energy and Environmental Systems Division
Argonne National Laboratory
Argonne, .Illinois 60439
~2. SPONSOR.NG AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
plllf £frA.r Q"al1ty Planning and Standards
Research Triangle Park, North Carolina 27711
J15. SUPPLEMENTARY NOTES~
EPA Project Officers: John Silvasi and Daniel deRoeck
"16. ABSTRACT
cnmQ -P 4-u allw-^i-iuuco5 lur all CT1 I Cl nilC T~ny*Q^_^<-\i i^4-«> «,,- i _
3. RECIPIENT'S ACCESSIOf*NO'.
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE~
8. PERFORMfNG ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
'1. CONTRACT/GRANT NO. '
EPA-79-D-F05Q2 _
3. TYPE OF REPORT AND PERIOD COVERED"
of
KEY WORDS AND DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
13-B
Air Pollution
State implementation
plan
Atmospheric contamination control
Lead
Control stfaiegy
National ambiifvt \
18. DISTRIBUTION STATEMENT
Release unlimited
25. SECURITY C1.ASS (Ttiispaff)
EPA Form 2220-1 (9-73)
-------
117
62. Brier, G.W., Statistical Questions Relating to -the Validation of Air
Quality Simulation Models3 U.S. EPA Report No. EPA-650/4-75-010,
RTF, N.C. (March 1975).
63. CARD, Inc., Capital and Operating Costs of Selected Air Pollution Control
Systems, U.S. EPA Report No. EPA-450/3-76-014, RTP, N.C.
64. Economic Impact Assessnient, unnumbered U.S. EPA Report, Office of Air
Quality Planning and Standards, RTP, N.C. (September 1978).
65. Environmental Impact Statement, unnumbered U.S. EPA Report, Office of Air
Quality Planning and Standards, RTP, N.C. (September 1978).
------- |