EPA-600/7-89-0l0a
October 1989
THE 1985 NAPAP EMISSIONS INVENTORY:
OVERVIEW OF ALLOCATION FACTORS
Prepared by
Robert A. Walters
Lysa G. Modica
David B. Fratt
ALLIANCE TECHNOLOGIES CORPORATION
213 Burlington Road
Bedford, Massachusetts 01730
EPA Contract No. 68-02-4274
Work Assignment Nos. 33 and 35
Robert C. Lagemann, EPA Project Officer
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
This study was conducted in cooperation with the
National Acid Precipitation Assessment Program
AIR AND ENERGY ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC 27711
-------
FOREWORD
This report documents the development of allocation factors for the 1985 National Acid Precipitation
Assessment Program (NAPAP) Emissions Inventories. As part of the NAPAP inventory development effort,
EPA currently plans to publish additional reports which will address various elements of allocation factor
development and application in greater detail:
The 1985 NAPAP Emissions Inventory (Version 2): Development of the Annual Data and Modeler's Tapes
An integrated report of the inventory development process, this document will contain sections on the
development of the allocation factors as well as a description of the software used to apply the factors to
annual emissions data.
The 1985 NAPAP Emissions Inventory: Development of Temporal Allocation Factors
This report will provide a detailed description of the data sources and methodologies used to develop and
enhance the NAPAP temporal factors. It will also contain descriptions of data processing software and
complete temporal factor listings.
The 1985 NAPAP Emissions Inventory: Development of Spatial Allocation Factors
This report summarizes spatial allocation factor development for U.S. and Canadian area sources and
documents the pertinent software and peripheral files used to create the factors. It also contains detailed
descriptions of the quality control checks undertaken and subsequent enhancements made to the factors.
The 1985 NAPAP Emissions Inventory: Development of Species Allocation Factors
This report will contain information on speciation methodologies for hydrocarbons, NO, and particulate matter,
and summarize computer programs and software used to generate speciation factors. Additional topics
addressed in the report will include development of class-average molecular weights and hydrocarbon
preprocessing.
Interested readers may wish to refer to these documents as they become available.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
A89-380
ii
-------
TABLE OF CONTENTS
Page
Foreword ii
Figures i v
Tables iv
Abstract vi
Executive Summary vii
1. Introduction 1
Background 1
Summary 2
Report Organization 3
2. Temporal Allocation Factors 4
Introduction 4
Area Source Temporal Allocation Factors 5
Point Source Temporal Allocation Factors 17
Quality Assurance and Factor Enhancements 19
Temporal Allocation Factor File Formats 20
3. Spatial Allocation Factors 25
Introduction 25
Spatial Factor Development 31
Quality Control and Enhancements of Spatial Factors for the
1985 NAPAP Inventory 37
Spatial Allocation Factor File Format 41
4. Species Allocation Factors 44
Introduction 44
Hydrocarbon/NO, Speciation Factor Development 45
Particulate Speciation Factor Development 49
Hydrocarbon Preprocessing 56
5. References 69
General References 69
Temporal Factor References 69
Spatial Factor References 70
Speciation Factor References 71
Appendices
A Adjustments to Spatial Fractions for 1985 NAPAP Processing A-l
B Development of Allocation Factors for Canadian B-l
Anthropogenic Sources
C Development of Allocation Factors for Natural Sources C-l
D Description of Seasonal Electric Utility Factors D-l
A89-380
iii
-------
FIGURES
Number Page
3-1 The extent of the NAPAP grid 26
3-2 Simplified flow diagram of spatial factor processing 32
3-3 Example of county-to-grid cell areal relationship 36
4-1 Example of VOC and THC calculation, FREDS
Hydrocarbon Preprocessor 60
TABLES
Number Page
2-1 Source Categories for U.S. Area Sources, 1985 NAPAP Inventory 6
2-2 Format of SAS Data Set containing Temporal Allocation Factors
for 1985 Area Source Emissions Data 21
2-3 Format of SAS Data Set Containing Temporal Allocation Factors
for 1985 Point Source Emissions Data 22
2-4 Temporal Allocation Scenario Types 24
3-1 Spatial Allocation Surrogates Available in the 1980 and 1985
NAPAP Spatial Allocation Factor Files 27
3-2 Spatial Allocation Factor Surrogates for the 1985 NAPAP Area
Source Emissions Categories 28
3-3 Summary of 1980 Spatial Factor Problems and Impacts 38
3-4 Summary of Spatial Factor Adjustments for the 1985 NAPAP Inventory 42
3-5 File Format for the EBCDIC 1985 Spatial Allocation Factor File 43
4-1 Hydrocarbon Species Classes, 1985 NAPAP Resolved Inventory 46
4-2 PSPLIT File Format 50
4-3 Particulate Species Classes, 1985 NAPAP Resolved Inventory 54
4-4 Particulate Spcciation Factor File Format 57
4-5 Hydrocarbon Augmentation Flag File 62
A89-380
iv
-------
TABLES (Continued)
Number Page
B-l Format of SAS Data Set Containing Temporal Allocation Factors for
1985 Canadian Area Source Emissions Data B-4
B-2 Format of SAS Data Set Containing Temporal Allocation Factors for
1985 Canadian Point Source Emissions Data B-5
B-3 Spatial Allocation Factor Surrogates for 1985 NAPAP Canadian Area
Source Emissions Categories B-6
B-4 File Format for the EBCDIC 1985 Canadian Spatial Allocation File B-9
B-5 Assignment of Hydrocarbon and Particulate Speciation Profiles for
Extra Canadian Point Source SCCs B-ll
B-6 Assignment of Hydrocarbon and Particulate Speciation Profiles for
Canadian Area Source Categories B-12
C-l Natural Alkaline Particulate Categories Included in the 1985 NAPAP
Emissions Inventory (Version 2.0) C-3
C-2 Speciation Factors for U.S. Natural Alkaline Particulates C-6
D-l Methods Used to Allocate Fuel Use to Seasons D-5
A89-380
V
-------
ABSTRACT
The 1985 NAPAP Emissions Inventory has been developed by the National Acid Precipitation Assessment
Program's (NAPAP) Task Group on Emissions and Controls. Temporal, spatial, and pollutant species
allocation factors were developed to convert annual emissions data into resolved modeling formats to support
the Regional Acid Deposition Model (RADM) and the Regional Oxident Model (ROM). The level of
resolution required for the allocation factors was determined by the modelers based on analyses of model
performance and emissions data bases. Allocation factors were originally developed for processing the 1980
NAPAP anthropogenic emissions inventory. These 1980 NAPAP allocation factors, and the methodologies
developed, served as the basis for the 1985 allocation factor development
Temporal allocation factors are comprised of three separate multipliers which permit resolution of annual
emissions to the hourly level for a typical weekday, Saturday, or Sunday in each season. Spatial allocation
factors apportion county-level area source emissions to modeling grid cells; point sources are assigned to grid
cells based on location data. Speciation factors allow hydrocarbon emissions to be resolved into 32
representative chemical classes. NO, emissions are resolved into NO and N02 and TSP emissions into 15
classes based on alkalinity and size distribution.
This report describes the allocation methodologies applied to convert the 1985 NAPAP Emissions Inventory
annual data base into a resolved format suitable as input to the Regional Acid Deposition Model (RADM).
The completion of the allocation process produced the 1985 NAPAP Modelers' Emissions Inventory Version
2. The Modelers' Inventory is being used as input to RADM for NAPAP assessment studies. This report is
intended to provide an overview of the allocation factor data bases and allocation methodologies.
A89-380
vi
-------
EXECUTIVE SUMMARY
This report documents the development of temporal, spatial, and species allocation factors for the 1985
National Acid Precipitation Assessment Program (NAPAP) anthropogenic point and area source emissions
inventories. These allocation factors are used to apportion annual emissions totals into gridded, hourly,
speciated emissions estimates suitable for use as input to atmospheric transport models such as the Regional
Acid Deposition Model (RADM).
Resource and logistical constraints prevented the collection of daily emissions data for the entire base year of
record (1985). Therefore, allocation factors have been developed to represent a typical weekday, Saturday and
Sunday for each of the four seasons. The modelers inventory is divided into 12 temporal scenarios, one
representing each of the day types. Each of the 12 scenarios is further resolved to the hourly level, for a
total of 288 hourly values.
The RADM includes a chemical reaction mechanism to simulate the transformation of primary emissions into
stable product species during transport. Emissions are reported in the annual inventory system as estimates of
SO„ NO,, VOC, CO and TSP. Speciation factors are applied to the aggregated emissions such that the
modeling inventory includes emissions estimates of N02 and NO, 32 chemical classes of hydrocarbons and
three size fractions of four reactive components of alkaline particulate species in addition to S02, CO, SO,,
HF, HCL, AND NH,. The grid pattern used to distribute the emissions data is defined by grids that are 1/4
degree longitude by 1/6 degree latitude (approximately 20 km by 20 km). There are 63,OCX) grid cells in the
NAPAP grid system, including Canada.
Allocation factors are statistical representations of the spatial and temporal distribution of annual emissions, or
representative speciation profiles for particular source types. Factors arc generally applied to NAPAP annual
missions records on the basis of point source SCC (Source Classification Code) or NAPAP area source
category.
The temporal, spatial, and species allocation factors are discussed in detail in separate report sections. Each
section contains a description of the methodology for application of the factors, a discussion of data sources,
and documentation of the activities undertaken to create the allocation factor data sets used in the 1985
NAPAP resolved modeling inventories.
A89-380
vii
-------
TEMPORAL ALLOCATION FACTORS
In support of modeling applications, the annual emissions totals must be resolved temporally into 24 hourly
totals for a typical weekday, Saturday or Sunday in each of the four seasons of the year. The NAPAP
temporal allocation factors were developed to accomplish this resolution.
Temporal factors were created for U.S. point and area source emissions categories in the NAPAP inventories.
Factors for the 1985 NAPAP inventory were derived primarily from temporal factors developed for the 1980
NAPAP effort. They reflect data from a variety of sources, which include previous modeling studies, as well
as enhancements deemed necessary as part of the ongoing development of the NAPAP inventories.
Factors were developed for each of the 97 U.S. area source categories included in the 1985 inventories. In
most cases, temporal allocation of point source data is accomplished using operating schedule information
provided with the emissions records. However, given the magnitude of emissions from electric utilities,
process-level (fuel and state specific) factors were developed from data supplied by the Department of Energy
and Tennessee Valley Authority to mane accurately characterize these sources.
Efforts to enhance the accuracy of temporal allocation arc ongoing. Factor normalization has eliminated
summation errors which occurred in previous inventories when temporally apportioned emissions were
reaggregated. Other modifications included the incorporation of data sources which enhance the accuracy of
temporal allocation.
SPATIAL ALLOCATION FACTORS
Spatial allocation factors were developed to apportion area source emissions from counties to individual grid
cells as required for modeling applications. The actual spatial distribution of emissions are estimated
according to the distribution of surrogate indicators. Fourteen such indicators were developed for use with
the NAPAP inventory based on housing, population, and land-use data. For the 1985 NAPAP application, 6
of the 14 surrogates are used for spatial allocation.
To assure the quality and representativeness of the spatially resolved 1985 area source inventory, extensive
quality control checks were performed on the existing spatial factors. Quality control procedures were both
data aid software intensive. Data analyses focused on evaluating spatial factors at the county level and
ensuring the quality of national-, state-, and SCC-level gridded emissions totals. Software-intensive
evaluations included reviewing computer code and implementing modifications to the spatial factor software
for other applications. Based on the results of the quality control procedures, adjustments were made to the
spatial factors and the computer programs. Once the adjustments were made, additional quality control checks
were performed to assure the quality of the modified spatial factors.
A89-380
viii
-------
SPECIATION FACTORS
Several of the pollutants in the 1985 NAPAP annual inventory represent composites of various individual
species. To accommodate RADM requirements, annual hydrocarbon emissions estimates are split into 32
chemical species classes, annual NO, estimates are divided into N02 and NO, and TSP emissions are resolved
into 15 classes based on alkalinity and size fraction.
The hydrocarbon and particulate matter speciation factors were based on profiles developed by the U.S. EPA
(Shareef et al., 1988). Percentage splits for NO and N02 were based on factors taken from AP-42 (U.S.
EPA, 1985). A default split of 95 percent NO and 5 percent N02 was applied to processes that were not
expected to be major NO, sources. Specific species-class assignments for VOC were developed by the
National Center for Atmospheric Research (NCAR). Alliance obtained these data, performed quality assurance
checks, and developed software to convert the files into a format suitable ft* use with the 1985 NAPAP
inventory processing software. Hydrocarbon species data were also used to create files for the preprocessing
of VOC and THC to adjust for the presence of formaldehyde and methane in some NEDS emissions
estimates.
A89-380
ix
-------
SECTION 1
INTRODUCTION
BACKGROUND
The "Acid Precipitation Act of 1980" (Title VII of PI. 96-294) established a long-term interagency program
to coordinate and expand research on the problems posed by acid deposition in and around the United States.
Among the priority research objectives that Congress identified for the National Acid Precipitation Assessment
Program (NAPAP) was the development of a comprehensive nationwide inventory of emissions sources
thought to contribute to the formation of acid rain. Complete and accurate inventories of acid deposition
precursor emissions are necessary to support assessment activities and studies of source-receptor relationships
using atmospheric process models. The NAPAP Task Group on Emissions and Controls has undertaken this
objective by developing inventories of acid deposition precursor emissions for the base year 1985. The
Environmental Protection Agency Office of Research and Development is responsible for developing the
NAPAP emissions inventories.
The 1985 NAPAP anthropogenic inventories were developed in two phases. The first phase involved data
collection and quality assurance of source characteristics and emissions totals on an annual basis (Zimmerman,
el al., 1988). The data collection effort used as a basis the existing Environmental Protection Agency (EPA)
National Emissions Data System (NEDS). NEDS data were supplemented by the application of noncriteria
pollutant emission factors to yield a final data base of 10 pollutants (SOa, S04, NO,, TSP, CO, NH„ HCI,
HF, VOC, and total hydrocarbons).
The second phase of the inventory development involved the generation of a resolved emissions inventory
containing gridded, hourly, and speciated emissions data. The spatial, temporal, and species allocation of the
annual emissions totals are required to render the data suitable as input to regional atmospheric process
models. For example, the Regional Acid Deposition Model (RADM), the principal model developed by
NAPAP, is an event-specific analysis tool that requires specific daily inputs of meteorological and emissions
data to simulate deposition episodes.
Resource and logistics constraints precluded the collection of sufficiently resolved emissions data for the entire
base year of record (1985). For example, to collect hourly emissions data for each 1/4 degree longitude by
1/6 degree latitude grid cell for each of the 59 species would be a monumental task at best Therefore, a set
of allocation factors was developed to resolve annual emissions data, using statistical representations of the
A89-38G
1
-------
spatial and temporal distribution of annual emissions and representative speeiation profiles for particular source
types. The level of resolution of the allocation factors have been determined by the modelers based on
analyses of model performance and emissions data bases (Middleton, 1987).
Temporal allocation factors are comprised of three separate multipliers: seasonal, daily and hourly. Spatial
allocation factors apportion county-level area source emissions to modeling grid cells; point sources are
assigned to grids cells based on location data (e.g., latitude and longitude or Universal Transverse Mercator
(UTM) coordinates). Speeiation factors allow hydrocarbon emissions to be resolved into 32 representative
chemical classes. NO, emissions are resolved into NO and N02, and TSP emissions are resolved into 15
classes on the basis of alkalinity and size distribution.
The temporal, spatial, and species allocation data were used as input to the Flexible Regional Emissions Data
System (FREDS) (Modica, et al., 1989), which matches the annual process-level emissions with appropriate
allocation factors to derive the gridded, hourly, speciated emissions estimates required for RADM input.
Version 1 of the point and area source modeling inventories generated by FREDS was delivered to EPA in
September 1988.
SUMMARY
Allocation factors were originally developed for processing the 1980 NAPAP anthropogenic emissions
inventory (Sellars, et al., 1985). These factors, and the methodologies developed for assigning them to
emissions records, formed the starting point for the 1985 allocation facte* development effort. Following the
completion of the 1980 inventory, the methods, data, and documentation used to develop allocation factors
were examined.
limitations were identified, the availability of new data was investigated, and requirements as to die necessity
and extent of changes to existing allocation factors were determined. In particular, updated hydrocarbon and
particulate speeiation data became available and were used to create entirely new speeiation factors. Temporal
and spatial factors remained largely based on data gathered for the 1980 inventory.
The 1985 NAPAP inventory contains county-level emission estimates for several area source categories not
included in previous inventories (Demmy, et al., 1988). In sane cases, these tow source types could be
accurately characterized by application of existing allocation data. Where this was not possible, new profiles
were developed. Additional issues were noted and addressed on a case-by-case basis. For example, a county
in New Mexico which was formed after 1980, had no corresponding spatial data in the original files, and
required a reapportionment of land area to account for its presence.
A89-380
2
-------
In addition, to ensure the quality of the resolved emissions estimates, extensive quality control checks were
performed both on the new factors and those adopted from the 1980 data base, and corrections were made
accordingly. The computer programs used to process the allocation data into a format suitable for use with
FREDS were also tested and checked for errors in logic or data format
REPORT ORGANIZATION
The remainder of this report documents the development of the various allocation factors ft* the 1985
NAPAP anthropogenic point and area source inventories. Section 2 provides a discussion of the temporal
factor development from a current and historical perspective. In Section 3, spatial factor development and
processing requirements are detailed. Section 4 provides information on the development of hydrocarbon, NO,
and particulate speciation factors. Each of these three sections includes background information on the
application of the factors, a discussion of data sources, and a description of the activities undertaken to create
allocation factor data sets suitable for use with the NAPAP inventory. Each section also contains a brief
description of the allocation factor files and computer formats as used by FREDS to create a resolved
modeling inventory. References, contained in Section 5, are organized by allocation factor type to facilitate
the identification of relevant documentation.
A89-380
3
-------
SECTION 2
TEMPORAL ALLOCATION FACTORS
INTRODUCTION
In support of modeling applications, annual emissions totals must be resolved temporally into 24 hourly totals
for a typical weekday, Saturday and Sunday in each of the four seasons of the year. To accomplish this task,
seasonal, daily, and hourly allocation factors were developed and applied to NAPAP point and area source
data.
The temporal allocation factors take the form of three sets of fractional multipliers, applied to the NAPAP
annual emissions records in sequence as follows:
1. Four seasonal factors divide the annual total into four subtotals representing emissions in each season.
2. Three daily factors per season divide each seasonal total into three subtotals representing emissions far a
typical weekday, Saturday and Sunday in each season.
3. Twenty-four hourly factors per day divide each daily total into 24 subtotals representing emissions for
each hour of that day. After this final step, the annual emissions have been divided into 288 subtotals
(four seasons x 3 day "typesVseason x 24 hrs/day "type").
Hie seasonal multipliers for each record sum to unity, as do the hourly multipliers for each season/day type
combination. Since daily emissions totals represent emissions for one typical weekday, Saturday or Sunday in
each season, the governing equation for daily allocation factors is:
(65 x weekday factor) + (13 x Saturday factor) + (13 x Sunday factor) = 1 (2.1)
where a season is defined as 91 days (13 weeks). For the purposes of the NAPAP inventory, the four
seasons are defined as:
Season
Months
Winter
December, January, February
Spring
March, April, May
Summer
June, July, August
Fall
September, October, November
A89-380
4
-------
All temporal factors are defined according to local time (e.g., the factor for "Hour 1" corresponds to the
period between midnight and 1:00 a.m„ local time). However, modeling applications require that these factors
be offset to Greenwich Mean Time (GMT) to produce standardized allocation for all U.S. sources. Offset to
GMT is accomplished during processing of the annual emissions inventories by integrating local factors with
time zone information. Where appropriate, additional data adjust the factors for Daylight Savings Time.
Temporal allocation factors were developed for U.S. point and area source categories in the NAPAP
inventory. The factors reflect data from a variety of sources, including previous modeling studies. Many of
the factors are based on data from the Northeast Corridor Regional Modeling Project (NECRMP). In the
NECRMP effort, estimates of temporal pattens were developed for point and area source emissions in 15
States in the eastern United States. Factors from the NECRMP States were often retained for the NAPAP
inventories; other factors were developed for states outside the NECRMP study area.
Point-specific factors were developed for only a subset of all point sources. Specifically, temporal profiles
were derived fa electric utility processes, as they are among the largest emitters erf acid rain precursors. In
most cases, however, temporal allocation of point-source emissions was estimated using operating schedule
information provided with annual emissions records.
Where specific temporal factors were not developed, emissions are allocated uniformly during processing of
the NAPAP point and area source data. The following default profiles are used:
Factor Type Uniform Default Value
Seasonal 1/4
Daily 1/91
Hourly 1/24
AREA SOURCE TEMPORAL ALLOCATION FACTORS
Temporal factors were developed for the 102 source categories in the NAPAP area source file. Depending cm
the magnitude of emissions within the category and the availability of data, temporal factors were frequently
resolved to the State or regional level (i.e., different sets of factors for each State and county for a given
source category). A complete list of area source categories used in the 1985 NAPAP inventories, including
the level of resolution for each SCC-specific temporal pattern is provided in Table 2-1.
A description of the development of area source factors for each category or set of categories follows.
Categories which were recently added or factors which have been modified for the 1985 NAPAP effort are
noted in the text Categories added for the 1985 inventories include:
A89-380
5
-------
1
2
3
4
5
6
7
8
9
to
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
40
41
42
43
44
45
46
47
48
49
50
51
52
53'
54
55
56
57
58
59
60
TABLE 2-1. SOURCE CATEGORIES FOR U.S. AREA SOURCES, 1985 NAPAP INVENTORY
Category description
Level of
temporal factor
resolution
Residential Fuel - Anthracite Coal
Residential Fuel - Bituminous Coal
Residential Fuel - Distillate Oil
Residential Fuel - Residual Oil
Residential Fuel - Natural Gas
Residential Fuel - Wood
Commercial/Institutional Fuel - Anthracite Coal
Commercial/Institutional Fuel - Bituminous Coal
Commerdai/lnstitutional Fuel - Distillate Oil
Commercial/Institutional Fuel - Residual Oil
Comrneraal/lnstitutional Fuel - Natural Gas
Commercial/Institutional Fuel - Wood
Industrial Fuel - Anthracite Coal
Industrial Fuel - Bituminous Coal
Industrial Fuel - Coke
Industrial Fuel - Distillate Oil
Industrial Fuel - Residual Oil
Industrial Fuel - Natural Gas
Industrial Fuel - Wood
Industrial Fuel - Industrial Process Gas
Inane ration - Residential
Inane ration - Industrial
Incineration - Commercial/Institutional
Open Burning - Residential
Open Burning - Industrial
Open Burning - Commercial/Institutional
Light Duty Gas Vehicles - Limited Access Roads
Light Duty Gas Vehicles - Rural Roads
Light Duty Gas Vehicles - Suburban Roads
Light Duty Gas Vehicles - Urban Roads
Medium Duty Gas Vehicles - Limited Access Roads
Medium Duty Gas Vehicles - Rural Roads
Medium Duty Gas Vehicles - Suburban Roads
Medium Duty Gas Vehicles - Urban Roads
Heavy Duty Gas Vehicles - Limited Access Roads
Heavy Duty Gas Vehicles - Rural Roads
Heavy Duty Gas Vehicles - Suburban Roads
Heavy Duty Gas Vehicles - Urban Roads
Off-Highway Gas Vehicles
Heavy Duty Diesel Vehicles - Limited Access Roads
Heavy Duty Diesel Vehicles - Rural Roads
Heavy Duty Diesel Vehicles - Suburban Roads
Heavy Duty Diesel Vehicles - Urban Roads
Off-Highway Diesel Vehicles
Railroad Locomotives
Aircraft - Military
Aircraft - Civil
Aircraft - Commercial
Vessels - Coal
Vessels - Diesel
Vessels - Residual Oil
Vessels - Gasoline
Solvents Purchased (not used)
Gasoline Marketed
Unpaved Road Travel
Unpaved Airport LTOs
(Not used)
(Not used)
(Not used)
Forest Fires
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
STATE
(NECRMP STATES),
(NECRMP STATES),
(NECRMP STATES),
(NECRMP STATES),
(NECRMP STATES),
(NECRMP STATES),
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
STATE
NATIONAL
NATIONAL
NATIONAL
NATIONAL
STATE
REGION
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL
STATE
NATIONAL
NATIONAL
NATIONAL
NATIONAL
NATIONAL (OTHER)
NATIONAL (OTHER)
NATIONAL (OTHER)
NATIONAL (OTHER)
NATIONAL (OTHER)
NATIONAL (OTHER)
(continued)
6
-------
TABLE 2-1 (continued)
Level of
Category
temporal factor
No.
Category description
resolution
61
Managed Burning - Prescribed
NATIONAL
62
Agricultural Field Burning
NATIONAL
63
(Not used)
64
Structural Fires
NATIONAL
65
(Not used)
66
Ammonia Emissions - Light Duty Gasoline Vehicles
NATIONAL
67
Ammonia Emissions - Heavy Duty Gasoline Vehicles
NATIONAL
68
Ammonia Emissions - Heavy Duty Diesel Vehicles
NATIONAL
69"
Livestock Waste Management - Turkeys
REGION
70"*
Livestock Waste Management - Sheep
REGION
71**
Livestock Waste Management - Beef Cattle
REGION
72**
Livestock Waste Management - Dairy Cattle
REGION
73"
Livestock Waste Management - Swine
REGION
74"
Livestock Waste Management - Broilers
NATIONAL
75"
Livestock Waste Management - Other Chickens
NATIONAL
76
Anhydrous Ammonia Fertilizer Application
REGION
77
Beef Cattle Feed Lots
NATIONAL
78
Degreasing
NATIONAL
79
Dry cleaning
NATIONAL
80
Graphic Arts/Printing
NATIONAL
81
Rubber and Plastics Manufacturing
NATIONAL
82
Architectural Coating
NATIONAL
83
Auto Body Repair
NATIONAL
84
Motor Vehicle Manufacture
NATIONAL
85
Paper Coating
NATIONAL
86
Fabricated Metals
NATIONAL
87
Machinery Manufacture
NATIONAL
88
Furniture Manufacture
NATIONAL
89
Flatwood Products
NATIONAL
90
Other Transportation Equipment Manufacture
NATIONAL
91
Electrical Equipment Manufacture
NATIONAL
92
Ship Building and Repairing
NATIONAL
93
Miscellaneous Industrial Manufacture
NATIONAL
94"*
(Not used)
NATIONAL
95*"
Miscellaneous Solvent Use
NATIONAL
96
Minor Point Sources-Coal Combusiton
NATIONAL
97
Minor Point Sources-Oil Combustion
NATIONAL
98
Minor Point Sources-Natural Gas Combustion
NATIONAL
99
Minor Point Sources-Process Sources
NATIONAL
100
Publicly-Owned Treatment Works (POTWs)
NATIONAL
101
Cutback Asphalt Paving Operation
NATIONAL
102
Fugitives from Synthetic Organic Chemical
Manufacture
NATIONAL
103
Bulk Terminal and Bulk Plants
NATIONAL
104
Fugitives from Petroleum Refinery Operations
NATIONAL
105
Process Emissions from Bakeries
NATIONAL
106
Process Emissions horn Pharmaceutical Manufacture
NATIONAL
107
Process Emissions from Synthetic Fibers
Manufacture
NATIONAL
108
Crude Oil and Natural Gas Production Fields
NATIONAL
109
Hazardous Waste Treatment, Storage, and
Disposal Facilities (TSDFs)
NATIONAL
NECRMP STATES: Connecticut, Delaware, District of Columbia, Maine, Maryland,
Massachusetts, New Hampshire, New Jersey, New York, Ohio,
Pennsylvania, Rhode Island, Vermont, Virginia, West Virginia.
* - Category 53 is disaggregated into process categories 78 to 95.
** - These categories formerly referred to as 'manure field application.*
*** - Formerly "miscellaneous industrial solvent use* (94) and "miscellaneous
non-industrial solvent use* (95); now combined into one category (95).
A89-380
-------
• Ammonia Emissions - Vehicular Sources (Categories 66-68);
• Livestock Waste Management (Categories 69, 70); and
• Miscellaneous VOC Categories (Categories 100-109).
Residential Fuel Combustion (Categories 1-6)
The residential fuel categories account for emissions from residential water heating, space heating, and
cooking. The surrogate for the temporal variability in emissions is heating degree days, which reflects
variability in average temperature.
To develop state-specific seasonal factors, monthly average heating degree days for each state for 1980 were
obtained from State, Regional, and National Monthly and Seasonal Heating Degree Days Weighted by
Population (U.S. Department of Commerce, 1983b). For each month, temperature averages were calculated
for state climatic divisions, which are regions within a state that are considered to be climatically
homogeneous. Monthly heating degree day totals for each division were calculated from the monthly mean
temperature. State-level average degree day totals for each month were derived from divisional values by
weighting each division by its percentage of total state population, as obtained from the 1980 Census; degree
day totals were therefore biased toward areas of the state having the greatest population. Seasonal temporal
factors were developed by determining the fractional distribution of annual heating degree days by month and
summing over each season.
Since the variability of residential heating emissions is assumed to be dependent only on temperature, a
uniform daily temporal pattern was assigned in the NAPAP effort. The hourly variations in residential fuel
use were developed with data from NOAA (U.S. Department of Commerce, 1980). Monthly averages of
3-hour meteorological records were obtained few a representative meteorological station in each state. For
each month, the averaged 3-hour temperatures were subtracted from 18.7*C (65*F), which is the value used to
calculate degree days. Negative values (indicating temperatures above 18.7'C) were set to zero. The
resulting eight values were proportional to the variation in diurnal heating for a selected station for a month.
Months were then averaged to obtain seasonally adjusted diurnal factors for each state.
Commercial/Institutional Fuel Combustion (Categories 7-12)
Area source emissions from fuel use by commercial and institutional sources consist of emissions from all
fuel burned in stationary sources that are not included under residential sources, industrial sources, power
plants or commercial point sources.
Seasonal allocation factors were developed from Procedures for the Preparation of Emission Inventories for
Volatile Organic Compounds, Volume II (EPA, 1979b), hereafter referred to as the EPA Guidelines. These
A89-380
8
-------
Guidelines recommend that 25 percent of emissions be spread uniformly across the seasons, and that the
remaining 75 percent be uniformly allocated to those months in which the average temperature is 18.7°C
(65°F) or less. The seasonal splits in the temporal allocation factors presented in NAPAP are 35 percent
winter, 25 percent spring, 15 percent summer, and 25 percent fall.
EPA Guidelines were also used to generate daily temporal factors. The Guidelines recommend that
95 percent of emissions be uniformly allocated from Monday through Saturday, and 5 percent be allocated to
Sunday. For NAPAP, it was thought that too little of the emissions were allocated to Sunday, since a certain
amount of energy is needed to keep buildings at an acceptable temperature. The revised profile allocates
90 percent of emissions uniformly from Monday through Saturday, and the remaining 10 percent to Sunday.
Hourly patterns for commercial/institutional fuel use for NECRMP states were taken from the NECRMP study
data. This pattern was developed in the Philadelphia AQCR inventory effort (Engineering-Science, 1982).
The profile assumes thai 50 percent of daily emissions are released uniformly from 7:00 am. to 4:00 p.m.,
with the other 50 percent released during the remaining hours in the day. Outside the NECRMP area, hourly
allocation factors were taken from the EPA Guidelines, which indicate 90 percent of emissions are released
between 5:00 a,m. and 10:00 p.m„ and 10 percent during the remaining hours in the day.
Industrial Fuel Combustion (Categories 13-20)
These categories include emissions from the industrial sector which are not accounted for by pant source
categories. National seasonal patterns were developed from the EPA Guidelines (EPA, 1979b), in which a
uniform distribution is recommended.
The daily pattern was based on U.S. Bureau of Labor statistics on average overtime at manufacturing facilities
(U.S. Department of Labor, 1981). All overtime hours were assumed to be worked on weekends, and three
times as much overtime was worked on Saturdays as on Sundays. The daily pattern developed from these
data indicate that approximately 93 percent of emissions are produced on weekdays, 5 percent on Saturdays,
and 2 percent on Sundays.
The hourly pattern was taken from NECRMP (Sellars, et al., 1982), and was developed during the
Philadelphia AQCR inventory effort (Engineering-Science, 1982). Fifty percent of industrial coal and oil fuel
use emissions were allocated uniformly from 7:00 a.m. to 4:00 p.m., to reflect greater production during
business hours. The remaining 50 percent of emissions were uniformly allocated to the remaining hours in a
day.
A89-3S0
9
-------
Onsite Incineration and Open Burning (Categories 21-26)
For the purposes of determining solid waste generated, onsite incineration is defined as disposal in a small
incinerator. Using this definition, incineration encompasses the following types of disposal units: backyard
burners, industrial incinerators, and incinerators used by food and department stores, hospitals, and schools.
Since large municipal incinerators are usually classified as point sources, emissions resulting tan disposal in
this type of incinerator have not been included in this category. For the purposes of estimated open burning
practices, open burning refers to uncombined burning of wastes such as leaves, landscape refuse, and other
rubbish. Large open burning dumps are usually included under point sources. Both onside incineration and
open burning categories are considered separately in NAPAP for the residential, commerciaWnstitutional, and
industrial sectors.
A single temporal profile set, developed in the NECRMP effort, was assigned to these categories on a
national level. The profile assumes that emissions occur uniformly throughout the year. The daily pattern
assumes that 91 percent of emissions are produced mi weekdays, and 9 percent are generated on Saturdays.
The hourly factors allot missions uniformly from 6:00 a.m. to 5:00 p.m. NECRMP documentation indicates
that this profile was based on temporal data in the St. Louis RAPS effort (EPA, 1979a), tie Tulsa, OK
inventory (EPA, 1980), and data from the states of New Jersey and New York (Sellars et al, 1982).
Highway Vehicles, Light and Medium Duty (Categories 27-34)
These eight categories account for exhaust and tire wear emissions from automobiles and light trucks
(0-8,500 lb gross vehicle weight [GVW]) on limited access, rural, suburban and urban roads. National-level
factors were developed for each of these eight categories, using data obtained from die U.S. Department of
Transportation (Welty, K. personal communication. Mr. Welty supplied traffic statistics compiled by the
Highway Statistics Division, Federal Highway Administration, U.S. Department of Transportation, on a
diskette formatted for use on an IBM Personal Computer.) These data were analyzed with FORTRAN
software designed specifically to calculate temporal allocation factors in the format required by the FREDS
data processing system. This agency collects continuous traffic count data from all 50 states, and uses a
subset of these data covering 12 states as a basis for estimating national traffic patterns. Data from these
12 states allow hourly temporal allocation factors to be derived for each day of die week in each month of
the year for six roadway types: rural and urban Interstate, rural and urban other arterial, and rural and urban
collector/local. This data set was obtained and analyzed to yield seasonal, daily, aid hourly temporal
allocation factors for light and medium duty vehicles. The pattern for limited access toads was derived from
an analysis of the rural and urban Interstate groups; the pattern for rural roads was taken from an analysis of
the rural other arterial group; and the pattern for suburban and urban roads resulted from an analysis of the
urban other arterial group.
AW-380
10
-------
Heavy Duty Highway Vehicles (Categories 35-38, 40-43)
These categories account for emissions from gasoline and diesel trucks weighing more than 8,500 lb GVW.
Emissions from these vehicles are inventoried separately for limited access roads, rural roads, suburban roads
and urban roads. A uniform seasonal pattern has been assigned to all these categories, after its applicability
was confirmed by the U.S. Tracking Association (Sellars et al, 1985).
A lack of daily factor data necessitated a uniform daily split for the Heavy Duty Gasoline Vehicle (HDGV)
and Heavy Duty Diesel Vehicle (HDDV) truck categories. The hourly profile for HDGV and HDDV was
derived from NECRMP data and was based on a profile developed for the Philadelphia AQCR inventory
effort (Engineering-Science, 1982). Although not explicitly staled in the Philadelphia report, it is likely that
these data were compiled by the Delaware Valley Regional Planning Commission (DVRPC) and documented
in a 1982 report (Delaware Valley Highway Vehicle Emissions Inventory, September 1982).
Off-Highway Vehicles (Categories 39,44)
The off-highway vehicle area source categories account for gasoline and diesel emissions generated by farm
equipment, construction equipment, industrial equipment, motorcycles, lawn and garden equipment, and
snowmobiles. Seasonal patterns for the off-highway vehicle categories were derived from data contained in
Highway Statistics (U.S. Department of Transportation, 1980). For this effort, monthly distributions of
off-highway motor vehicle fuel use were calculated on a state-specific basis by subtracting monthly
on-highway fuel use from total monthly fuel use. state-specific seasonal patterns were then derived from the
monthly data.
Daily patterns were derived from the EPA Guidelines (EPA, 1979b), and are a composite (weighted by
emission strength) of the daily patterns for five subcategories of off-highway vehicles (agricultural equipment,
construction equipment, industrial equipment, lawn and garden equipment, and motorcycles) for which factors
are separately described in the Guidelines. The suggested daily patterns for each subcategory are as follows:
• Farm equipment - uniform through the week
• Construction equipment - uniform Monday through Saturday
• Industrial equipment • uniform Monday through Saturday
• Lawn and garden equipment - 50 percent Monday through Friday, SO percent Saturday through
Sunday
OfT-highway motorcycles - 30 percent Monday through Friday, 70 percent Saturday through
Sunday
• Snowmobiles - 30 percent Monday through Friday, 70 percent Saturday through Sunday
A89-380
11
-------
The resulting weighted daily pattern assigns emissions uniformly over the week.
The NECRMP documentation describes the development of separate hourly patterns for gasoline and diesel
vehicles, as follows. First, a single hourly pattern was developed for the composite of five of the component
subcategories identified previously. Next, percentages of gasoline and diesel usage were derived for the five
subcategories of off-highway use with information obtained from Highway Statistics, 1980 (U.S. Department
of Transportation, 1980) (Table MF-24), "Private and Commercial Non-Highway Use of Gasoline" and a
combination of material from various NECRMP area source emission inventories (Ohio, Virginia, New
Hampshire, and Delaware) (Sellars, et al., 1982). These weighted percentages were then applied to the
composite hourly curve to produce composite curves for gasoline and diesel off-highway fuel use.
Railroads (Category 45)
This category accounts for emissions from locomotives and fuel used by railroad stations and workshops for
space heating. The latter fuel consumption was included primarily because it is difficult to separate from
total railroad fuel use; railroad space heating emissions are considered insignificant compared to those
generated by locomotive fuel consumption (Sellars, et al., 1982).
Within the NECRMP region, temporal patterns for locomotives were based on information provided by
Conrail for the Philadelphia AQCR inventory effort (Engineering-Science, 1982). Most emissions in this
category are due to freight service rather than passenger service. A sales manager at Conrail indicated that
36 percent of the traffic occurred uniformly between 5:00 p.m. and 11:00 p.m. EST, and the remaining
64 percent was equally distributed around the clock. Midday local traffic involved the movement of boxcars
for loading and unloading. No seasonal variations were assumed. The Conrail sales manager indicated there
was only a slight weekday-weekend day preference, so an operating schedule based on a 6.3-day work-week
was assumed.
Outside the NECRMP region, temporal patterns were developed according to EPA Guidelines (EPA, 1979b).
The Guidelines recommend uniform emissions throughout the year and week, with 70 percent of hourly
emissions being generated from 7:00 a.m. to 6:00 p.m. and 30 percent from 6:00 p.m. to 7:00 a.m.
Aircraft (Categories 46-48, 56)
Aircraft are divided into civil, military and commercial subgroups in the NAPAP inventory. For civil aircraft,
seasonal daily and hourly patterns were derived from the EPA Guidelines (EPA, 1979b). A uniform seasonal
distribution is recommended, with 40 percent of operations occurring on weekends and 60 percent on
weekdays. Emissions are allocated uniformly from 7:00 a.m. to 9:00 p.m.; levels outside this period are set
to zero. The same temporal profile was used to allocate emissions from unpaved airport LTOs (Category 56).
A89-380
12
-------
For military aircraft, patterns were derived from the Philadelphia AQCR Inventory (Engineering-Science,
1982). Interviews from this study led to the adoption of a uniform seasonal and daily pattern for military
aircraft, with 55 percent allocated between 9:00 a.m. and 9:00 p.m., and 45 percent spread uniformly over the
remaining hours. For commercial aircraft, seasonal patterns were derived from unpublished data presented in
the U.S. Civil Aeronautics Board (CAB)'s Seasonally Adjusted Traffic and Capacity. The CAB report
provided national statistics on monthly domestic traffic handled by the major air carriers. Commercial aircraft
emissions were assumed to occur 7 days/week, with 90 percent of all commercial emissions occurring
between 6:00 a,m. and midnight
Vessels (Categories 49-52)
Emissions from vessels are split into four area source categories on the basis of fuel type. Coal-, diesel-, and
residual fuel-powered vessels include large river- and ocean-going barges. Gasoline-powered vessels are
mostly far recreational use. Temporal patterns for all vessels were developed from the EPA Guidelines (EPA,
1979b). The following patterns were used:
• Ocean and river cargo vessels
Seasonal—uniform
Daily—uniform
Hourly—75 percent from 7:00 a.m. to 7:00 p.m. and
25 percent from 7:00 p.m. to 7:00 am.
• Pleasure craft
Seasonal—uniform during months with an average temperature of 45°F or higher.
Daily—30 percent Monday through Friday, 70 percent Saturday through Sunday
Hourly—uniform from 7:00 a.m. to 6:00 p,m„ otherwise zero
The seasonal patterns for pleasure craft were developed using State-level average temperature data compiled
by the NOAA (U.S. Department of Commerce, 1983a).
Gasoline Marketing (Category 54)
This area source category accounts for missions from tanker truck loading and transit, gasoline station
loading, storage tank breathing and vehicle fueling. Sources Investigated in the development of seasonal
patterns were refiner sales to end users, and consumption data from the U.S. Department of Energy's (DOE's)
Petroleum Marketing Monthly (U.S. DOE, 1984) and the U.S, Department of Transportation's Highway
Statistics 1980 (U.S. Department of Transportation, 1980). From analyses of these it was concluded that
A89-380
13
-------
the evidence was insufficient to justify using other than a uniform seasonal distribution. Daily and hourly
patterns were based on the EPA Guidelines (EPA, 1979b) which indicate uniform emissions Monday through
Saturday, with uniform emissions from 6:00 a.m. to 8:00 p.m., and no emissions at other times.
Unpaved Roads (Category 55)
The seasonal, daily and hourly patterns for this category were taken from those for light duty vehicles cm
rural roads.
Forest Fires (Category 60)
Forest fires were assumed to occur randomly, 7 days per week, 24 hours per day. It was estimated that
90 percent of forest fires occur during summer or fall, and that the remaining 10 percent are split evenly
between winter and spring.
Agricultural Burning (Categories 61 and 62)
Both of these categories were assigned the patterns developed for field/slash burning in the NECRMP study
(Sellars, et al., 1982). The NECRMP pattern assumed burning to occur 7 days per week, and uniformly
between 5:00 am. and 8:00 p.m. It was also assumed that no burning would occur during die summer
season. The actual NAPAP seasonal splits are 10 percent winter, 70 percent spring, 0 percent summer, and
20 percent fall.
Structural Fires (Category 64)
Structural fires are assumed to occur randomly throughout die year. Thus a uniform seasonal, daily and
hourly pattern was assumed for this category.
Ammonia Emissions - Vehicular Sources (Categories 66-68)
(Added for 1985 NAPAP Inventory)
These new categories contain ammonia emissions for three vehicle types: light-duty gasoline, heavy-duty
gasoline, and heavy-duty diesel. Temporal profiles were taken directly from corresponding urban vehicular
categories for which emissions of other pollutants were reported (Categories 30, 38, and 43, respectively).
Livestock Waste Management (Categories 69~75)
(SCCs 69 and 70 added for 1985 NAPAP inventory)
A89-380
14
-------
Ammonia emissions from livestock waste management include those resulting from production and field
application of different types of livestock manure. Emissions for these categories were estimated cm the basis
of number of each type of livestock, weighted to account for the percentage confined and unconfined animals.
National-level temporal patterns were generally uniform for seasonal, daily, and hourly allocation. To
establish state-level profiles Agricultural Extension Agents in 12 states across the country (Indiana, Wisconsin,
Nebraska, Kansas, North Dakota, Idaho, Iowa, North Carolina, Alabama, New York, Texas and Arkansas)
were contacted; they provided information on the patterns of manure application in their states. These
patterns were extended to nearby states not contacted directly.
For the 1985 effort, emissions for turkey and sheep waste management (Categories 69 and 70) were provided
in addition to the livestock categories reported previously (Categories 71-75). Temporal profiles for these new
sources were assumed to be similar to those used for other livestock waste management categories, and were
copied directly from Category 71 (beef cattle waste management).
NH, Fertilizer Application (Category 76)
The 12 Agricultural Extension Agents contacted to provide information on manure-spreading practices (see
above) were also asked to share their knowledge of NH fertilization practices. Daily factors were derived
from the assumption that no fertilizer application occurs on Sundays; the national daily pattern assigns
emissions uniformly Monday through Saturday. The national hourly pattern assumes approximately 60 percent
of emissions occur between 8:00 a.m. and 7:00 p.m., 17 percent occur between 9:00 p.m. and 4:00 a.m., and
the remainder are distributed over the intervening period (Wagner et al, 1986).
Beef Cattle Feed Lots (Category 77)
Emissions were allocated according to uniform seasonal, daily, and hourly patterns.
Solvent Use Categories (Categories 78-95)
Early versions of the NAPAP area source inventories reported a large composite category (53) for organic
solvent evaporation based cm data from NEDS. Certain NEDS-generated reports, however, split this aggregate
category into 18 individual categories of solvent use. Beginning with Version 5 of the 1980 NAPAP
inventory, emissions previously reported under Category 53 were disaggregated into these individual
categories (78-95).
U.S. Department of Labor statistics on 1980 working hours (U.S. Department of Labor, 1981) were consulted
to derive national-level seasonal and daily temporal variation ft* each of these emission categories. Monthly
A89-380
15
-------
data cm total hours worked served as the basis for seasonal allocation factors. For most categories, seasonal
allocation is approximately uniform. Data on overtime hours worked served as the basis for daily factors; the
assumption was made that all overtime hours were worked on weekends, and that three times as much
overtime was worked on Saturdays as on Sundays. The length of a working day (e.g., 7am-6pm) was used
to generate hourly factors for each category. In general, most emissions were allocated uniformly during this
period, with a smaller percentage (20-25 percent) divided throughout the remainder of the day.
Minor Point Sources (Categories 96-99)
These area source emissions categories were originally created for the 1980 NAPAP inventory to account for
small point sources of SO,, NO,, and TSP. For Version 2 of the 1985 inventory, these categories were based
on the following emissions criteria:
• All individual points with emissions of less than 5 TPY of each pollutant (SOJf NO,, VOC, TSP,
and CO);
• All points at any plant found to emit less than 100 TPY of each pollutant (100 TPY plants are
determined by summing emissions far all points excluding the less than 5 TPY points noted in
the criterion above).
Categories 96-98 account for emissions from minor coal, oil, and gas combustion sources, respectively. The
seasonal factors for these categories were derived from electric generation data contained in the EPRI
Regional Systems (EPRI, 1981). The EPRI report presents monthly peak load data for each of the si* EPRI
regions. It is likely that these regional data were averaged nationwide to produce the national seasonal factor.
The seasonal splits used in NAPAP are 27 percent winter, 23 percent spring, 27 percent summer, and 23
percent Call. The uniform daily pattern is based upon the assumption that utility activity is constant
throughout the week. The hourly profile indicates that most activity takes place during normal working hours.
Category 99 includes combustion sources for fuels such as coke and process gas, evaporative emissions, and
all process emissions. Owing to the diversity of source types within this category, the temporal pattern
assigned reflected a general operating schedule of 52 weeks per year, 5 days per week, and 8 hours per day.
Miscellaneous VOC Categories (Categories 100-109)
(Added for the 1985 NAPAP inventory)
The 1985 NAPAP inventory provides county-level VOC emissions estimates for several categories which have
not previously been included as area source categories. Some of these categories, such as Bakeries and
Synthetic Fiber Manufacturing, were included to reconcile the difference between the total emissions reported
in the National Air Pollutant Emissions Estimates 1940-1984 (EPA, 1986), aid the emissions already
accounted for by the NEDS point source data files. The remaining categories such as Publicly-Owned
A89-380
16
-------
Treatment Works (POTWs) and hazardous waste Treatment, Storage, and Disposal Facilities (TSDFs) have
been included due to the difficulty of measuring emissions from specific points within these categories (e.g„
aeration basins).
To develop temporal profiles for these ten categories, operating schedule data for similar point source
categories in the NAPAP inventory were analyzed. Operating data included seasonal throughput percentages,
as well as hours per day and days per week of process operation. As a result, it was found that all but two
of the new categories were best represented by uniform profiles. For the remaining two categories (101,
105), factors were derived from mean operating schedule data for similar categories in the point source fife.
The derived factors are as follows:
Category 101
Seasonal percentages of 10, 25, 40, and 25 (for winter, spring, summer, fall, respectively);
Daily uniform over weekdays;
Hourly uniform from 7 a.m, to 5 p.m.
Category 105
Seasonal
Daily
Hourly
percentages of 24, 24, 26, and 26 (for winter, spring, summer, fall, respectively);
uniform over weekdays and Saturdays;
uniform from midnight to 5 p.m.
POINT SOURCE TEMPORAL ALLOCATION FACTORS
Temporal allocation of point source data is accomplished by one of two methods. In most cases, operating
schedule data included in point-source data records saved as the basis for point source factors. However,
given the magnitude of emissions from electric utility sources, process-specific factors were developed to more
accurately characterize these sources.
Operating Data-Derived Factors
Most NAPAP point source emission records contain operating schedule data which make possible the
point-specific temporal allocation of emissions. These data consist of information on:
• Seasonal throughput percentages;
• Days/week of process operation; and
Hours/day of process operation.
A89-380
17
-------
These operating data woe used to temporally apportion point source data for which specific temporal factors
were not developed
Seasonal factors were taken directly from seasonal throughout percentages. Daily factors were derived from
the number of days per week which a process operates. Factors were calculated according to the following
schedule:
If the process operates
x days/week
Emissions are allocated as follows
1
Saturdays only
2
Equally on Saturdays and Sundays
3-5
Equally on weekdays only
6
Equally on weekdays and Saturdays
7
Equally cm all days of the week
Hourly factors are based on the number of hours per day which a process operates. Factors were developed
using the following method:
If the process operates
x hours/day Emissions are allocated as follows
1-17 Zero for midnight to 7:00 a.m.; equally far x hours
beginning with 7:00 a.m.; and zero for hours remaining
before midnight
> 17 Equally among 24 hours of the day.
Point Source Temporal Factors for Electric Utilities
Because of the importance of electric power plants to total U.S. emissions of S02 and NO,, and because
detailed data are available for these sources, special process-level (fuel and state specific) temporal factors
were developed for utilities during the NAPAP effort Factors were originally developed during the NECRMP
study for power plants within the Northeast Corridor. In development of the NAPAP inventories, most of the
NECRMP factors were retained, while some were replaced by new factors derived from additional data. For
sources outside the NECRMP region, some factors were based on NECRMP figures; however, most were
developed using other data sources.
A89-380
18
-------
Seasonal factors for the temporal distribution of point source emissions were developed on a fuel- and
state-specific basis for facilities in the NECRMP study area, using information from the U.S. DOE's 1979
Energy Data Reports (U.S. Department of Energy, 1979). Outside the NECRMP region, uniform seasonal
allocation factors were assigned.
Daily factors were developed at the national level from weekly load cycle listings in the EPRI Regional
Systems (Electric Power Research Institute, 1981). These daily factors were calculated by normalizing the
averages of the daily load statistics contained in the EPRI report The factors, which were compiled on a
season-specific basis, assume 75-76 percent of weekly emissions are produced on weekdays, 12-13 percent are
produced on Saturdays, and the remainder are generated on Sundays.
Fuel- and state-specific weekday hourly patterns for utility operation were developed during the NECRMP
effort (Sellars, et al., 1982). Profiles were derived from hourly power plant fuel-use data collected during the
development of the SURE inventory (Klemm and Brennan, 1981). The SURE data base included hourly
fuel-use data for approximately 300 power plants within the SURE region (the eastern portion of the United
States). These data were collected for several study periods in 1977-1978 and included data for each season.
Fuel- and season-specific fuel-use patterns were averaged by state, then normalized to generate the NECRMP
hourly patterns. Since there were no hourly fuel-use data available for Rhode Island or Vermont, weekday
factors were assigned from Connecticut and New Hampshire, respectively. Outside the Northeast Corridor,
hourly factors for weekdays were developed by taking averages of the NECRMP values. Since no weekend
hourly factors were developed during the NECRMP study, national Saturday and Sunday hourly profiles were
developed from weekly load cycle listings in EPRI Regional Systems (Electric Power Research Institute,
1981). These were calculated by normalizing the averages of hourly load statistics.
QUALITY ASSURANCE AND FACTOR ENHANCEMENTS
Quality assurance and enhancement of the NAPAP temporal allocation factors is an ongoing process. As
described above, the addition of new source categories for the 1985 area source inventory required the
creation of corresponding factor profiles. Continuing work includes enhancing the accuracy of these and
existing allocation factors by examining alternative data sources which may provide more representative
temporal data. Efforts focus on sources which are significant contributors to acid precipitation precursors
(e.g., electric utilities), and for which data are readily available.
Factor Normalization
During development of the 1985 NAPAP inventory, quality control checks of temporally allocated emissions
revealed that inaccuracies sometimes resulted when reaggregation of point and area source emissions was
attempted. These inaccuracies resulted from the fact that temporal factors themselves did not always sum to
A89-380
19
-------
unity according to the prescribed algorithms in the Introduction to this section. This in turn, was caused by
rounding errors which were produced when factors were stored as 3-4 place values in computerized data files.
To preserve annual emissions totals following temporal allocation, it was necessary to: (1) normalize the
temporal factors; and (2) store the normalized factors more accurately in data files (as eight-byte values of
computerized variables).
Other Studies
As described above, emissions from utility point sources outside the NECRMP study area were assigned a
uniform seasonal profile in early versions of the NAPAP temporal allocation factor file. As a result, large
amounts of SO, and NO, emissions were allocated according to these default factors. Efforts to improve
point source temporal allocation in the 1985 NAPAP inventory included:
using seasonal throughput percentages on plant-level records as the primary source of seasonal allocation
data; uniform profiles were used secondarily, and
incorporation of more accurate utility operating data.
Updates to seasonal factors for electric utility sources were made primarily on the basis of monthly plant-level
fuel use data from DOE/EIA Form 759. These data were aggregated to form seasonal activity factors at the
plant-fuel or state-fuel level (See Appendix D). Fuel-specific, state-level factors for utility boilers were also
developed on the basis of unpublished computerized monthly reports from DOE/EIA titled "R080-Report on
Consumption" from December 1984 to December 1985.
Hourly generation data tan 58 TV A coal-fired boilers were evaluated to produce point-level seasonal, as well
as daily and hourly, temporal factors. These profiles were added to the temporal allocation factor file and
necessitated data processing modifications to accommodate point-level factor assignment
All seasonal factors derived from these analyses woe incorporated as seasonal throughput percentages on
point source annual emission records.
TEMPORAL ALLOCATION FACTOR FILE FORMATS
Formats for area and point source temporal factor files are given in Tables 2-2 and 2-3. Both factor files are
*
organized as data sets in the programming language of SAS. All data are represented as 8-byte variables.
The factors themselves (SEA-, DAY-, and HOUR-type variables) are decimal fractions. All other values are
integers.
* SAS is a registered trademark of the SAS Institute, Cary, NC.
A89-380
20
-------
TABLE 2-2. FORMAT OF SAS DATA SET CONTAINING TEMPORAL ALLOCATION
FACTORS FOR 1985 AREA SOURCE EMISSIONS DATA
SAS variable length
Variable name (bytes) Description
NUM_DAY
8
Temporal Scenario Code
STATE
8
AEROS Stats Code
COUNTY
8
AEROS County Code
see
8
Area Source Category Code
SEA
8
Seasonal Temporal Allocation Factor
DAY
8
Daily Temporal Allocation Factor
HOUR1-HOUR24
8
Hourly Temporal Allocation Factors
A89-380
21
-------
TABLE 2-3. FORMAT OF SAS DATA SET CONTAINING TEMPORAL ALLOCATION
FACTORS FOR 1985 POINT SOURCE EMISSIONS DATA
Variable name
SAS variable length
(bytes)
Description
NUM_DAY
8
Temporal Scenario Code
STATE
8
AEROS State Code
COUNTY
8
AEROS County Code
PLANTJD
8
NEDS Plant Identifier
POINTJD
8
NEDS Point Identifier
see
8
Source Classification Code
SEA1-SEA12
8
Seasonal Temporal Allocation Factors
DAY1-DAY12
8
Daily Temporal Allocation Factors
HOUR1-HOUR2B88
8
Hourly Temporal Allocation Factors
A89-380
22
-------
The value of the variable NUM_DAY defines the day type of a given temporal scenario record. In all, there
are 12 day types, which are defined in Table 2-4. F
-------
TABLE 2-4. TEMPORAL ALLOCATION SCENARIO TYPES
Reference number
(NUM_DAY)
Scenario
1
Winter weekday
2
Winter Saturday
3
Winter Sunday
4
Spring weekday
5
Spring Saturday
6
Spring Sunday
7
Summer weekday
8
Summer Saturday
9
Summer Sunday
10
Fall weekday
11
Fall Saturday
12
Fall Sunday
-380
24
-------
SECTION 3
SPATIAL ALLOCATION FACTORS
INTRODUCTION
Spatial allocation factors were developed for the NAPAP inventories to apportion area source emissions from
counties to a series of grid cells as required for modeling applications. The NAPAP grid system covers the
United States and Canada and extends from 50° to 125° west longitude and 25° to 60° north latitude.
However, for the purpose of this section of the document, only spatial surrogates developed for U.S. area
source categories are considered. Canadian spatial factors are discussed in Appendix B. Grid cells are 1/6°
latitude by 1/4° longitude (or approximately 20 x 20 km). The actual size of a grid cell is a function of
latitude and decreases to the north. Figure 3-1 illustrates the extent of the NAPAP grid. It should be noted
that this figure depicts 2/3° latitude by 1° longitude grids for visual clarity.
A spatial allocation facte* assigns a portion of an individual county's area source emissions to a specific grid
cell. Since the actual subcounty distribution of area source emissions is usually not known, emissions are
assumed to be distributed according to the known distribution of another quantity, i.e., a surrogate indicator.
For example, in County A, the total housing count is 1,000 units and the housing count of County A within
grid cell (1,1) is 200. The spatial fraction for housing in grid cell (1,1) for County A is 0.20. Therefore,
20 patent of County A's area source emissions spatially allocated by housing (e.g., residential fuel) would be
allocated to grid cell (1,1). This methodology assumes area source emissions occur uniformally across the
entire area of the grid cell.
The goal of the NAPAP spatial allocation factor development effort was to create as many surrogate values as
possible for each county, allowing the user maximum flexibility in assigning county-level emissions to specific
grid cells. These surrogates are used to estimate the subcounty distribution of area source emissions.
Fourteen surrogate indicators were developed for use with the NAPAP inventory based on housing,
population, and land-use data. The categories and sources of data are summarized in Table 3-1. For 1985
NAPAP application, 6 of the 14 surrogates are used for spatial allocation. Table 3-2 contains a listing of the
spatial surrogates and their assignment to each area source category.
A89-380
25
-------
Figure 3-1. The extent of NAPAP grid. Note: 2/3° latitude by f longitude
(approximate 80 x 80 km) grid cells are plotted in this figure.
Source: The Flexible Regional Emissions Data System for the 1980 NAPAP
Emissions Inventory (Lebowitz and Ackerman. 1987).
-------
TABLE 3-1. SPATIAL ALLOCATION SURROGATES AVAILABLE IN THE 1980
AND 1985 NAPAP SPATIAL ALLOCATION FACTOR FILES
Surrogate
Indicator No. Surrogate Indicator Source
1
Population
1980 Census
2
Housing
1980 Census
3
Urban Land
Landsat
4
Agricultural Land
Landsat
5
Rangeland
Landsat
6
Deciduous Forest
Landsat
7
Coniferous Forest
Landsat
8
Mixed Forest and Forested Wetland
Landsat
9
Water
Landsat
10
Barren Land
Landsat
11
Nonforested Wetland
Landsat
12
Mixed Agricultural Land and Rangeland
Landsat
13
Composite Forest
Landsat
14
Land Area
EPA/Alliance
Note: The Landsat data are for 1972-1973.
A89-380
27
-------
TABLE 3-2. SPATIAL ALLOCATION FACTOR SURROGATES FOR THE
1985 NAPAP AREA SOURCE EMISSIONS CATEGORIES
Category Surrogate Surrogate
ID ID indicator Emissions category
1
2
Housing
Residential Fuel-Anthracite Coal
2
2
Housing
Residential Fuel-Bituminous Coal
3
2
Housing
Residential Fuel—Distillate Oil
4
2
Housing
Residential Fuel—Residual Oil
5
2
Housing
Residential Fuel—Natural Gas
6
2
Housing
Residential Fuel-Wood
7
3
Urban Land
Commercial/Institutional Fuel-Anthracite Coal
8
3
Urban Land
Commercial/Institutional Fuel—Bituminous Coal
9
3
Urban Land
Commercial/Institutional Fuel-Distillate Oil
10
3
Urban Land
Commercial/Institutional Fuel-Residual Oil
11
3
Urban Land
Commercial/Institutional Fuel-Natural Gas
12
3
Urban Land
Commercial/Institutional Fuel-Wood
13
3
Urban Land
Industrial Fuel-Anthracite Coal
14
3
Urban Land
Industrial Fuel-Bituminous Coal
15
3
Urban Land
Industrial Fuel-Coke
16
3
Urban Land
Industrial Fuel-Distillate Oil
17
3
Urban Land
Industrial Fuel-Residual Oil
18
3
Urban Land
Industrial Fuel-Natural Gas
19
3
Urban Land
Industrial Fuel-Wood
20
3
Urban Land
Industrial Fuel-Process Gas
21
2
Housing
Incineration—Residential
22
3
Urban Land
Incineration-Industrial
23
3
Urban Land
Incineration—Commercial/Institutional
24
2
Housing
Open Burning-Residential
25
3
Urban Land
Open Burning-Industrial
26
3
Urban Land
Open Burning-Commercial/Institutional
27
14
Land Area
Light Duty Gas Vehicles-Limited Access
28
14
Land Area
Light Duty Gas Vehicles-Rural
29
2
Housing
Light Duty Gas Vehicles—Suburban
30
3
Urban Land
Light Duty Gas Vehicles-Urban
31
14
Land Area
Medium Duty Gas Vehicles-Limited Access
32
14
Land Area
Medium Duty Gas Vehicles-Rural
33
2
Housing
Medium Duty Gas Vehicles-Suburban
34
3
Urban Land
Medium Duty Gas Vehicles-Urban
35
14
Land Area
Heavy Duty Gas Vehicles-Limited Access
36
14
Land Area
Heavy Duty Gas Vehicles—Rural
37
2
Housing
Heavy Duty Gas Vehicles-Suburban
38
3
Urban Land
Heavy Duty Gas Vehicles-Urban
39
14
Land Area
Off-Highway Gas Vehicles
40
14
Land Area
Heavy Duty Diesel Vehicles-Limited Access
41
14
Land Area
Heavy Duty Diesel Vehicles-Rural
42
2
Housing
Heavy Duty Diesel Vehicles-Suburban
43
3
Urban Land
Heavy Duty Diesel Vehicles—Urban
44
14
Land Area
Off-Highway Diesel Vehicles
45
3
Urban Land
Railroad Locomotives
46
1
Population
Aircraft-Military
(continued)
A89-380
28
-------
50
51
52
54
55
56
60
61
62
64
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
95
96
97
98
99
TABLE 3-2 (continued)
Surrogate Surrogate
ID indicator Emissions category
1
Population
Aircraft-Civil
1
Population
Aircraft-Commercial
1
Population
Vessels-Coal-Powered
1
Population
Vessels-Diesel
1
Population
Vessels-Residual Oil
1
Population
Vessels-Gasoline
1
Population
Gasoline Marketed
14
Land Area
Unpaved Roads
14
Land Area
Unpaved Airstrips
13
Composite Forest
Forest Wild Fires
13
Composite Forest
Managed Burning - Prescribed
4
Agricultural
Land
Agricultural Field Burning
2
Housing
Structural Fires
14
Land Area
Ammonia Emissions-Light Duty Gasoline Vehicles
14
Land Area
Ammonia Emissions-Heavy Duty Gasoline Vehicles
14
Land Area
Ammonia Emissions-Heavy Duty Diesel Vehicles
4
Agricultural
Land
Livestock Waste Management - Turkeys
4
Agricultural
Land
Livestock Waste Management - Sheep
4
Agricultural
Land
Livestock Waste Management - Beef Cattle
4
Agricultural
Land
Livestock Waste Management - Dairy Cattle
4
Agricultural
Land
Livestock Waste Management - Swine
4
Agricultural
Land
Livestock Waste Management - Broilers
4
Agricultural
Land
Livestock Waste Management - Other Chickens
4
Agricultural
Land
Anhydrous NH3 Fertilizer Application
4
Agricultural
Land
Beef Cattle Feed Lots
1
Population
Degreasing
1
Population
Drycleaning
1
Population
Graphic Arts/Printing
1
Population
Rubber and Plastic Manufacturing
1
Population
Architectural Coating
1
Population
Auto Body Repair
1
Population
Motor Vehicle Manufacturing
1
Population
Paper Coating
1
Population
Fabricated Metals
1
Population
Machinery Manufacturing
1
Population
Furniture Manufacturing
1
Population
Flat Wood Products
1
Population
Other Transportation Equipment Manufacturing
1
Population
Electrical Equipment Manufacturing
1
Population
Ship Building and Repair
1
Population
Miscellaneous Industrial Manufacturing
1
Population
Miscellaneous Solvent Use
1
Population
Mine* Point Sources - Coal Combustion
1
Population
Minor Point Sources - Oil Combustion
1
Population
Minor Point Sources - Natural Gas Combustion
1
Population
Minor Point Sources - Process Sources
1
Population
Publicly Owned Treatment Works (POTWs)
(continued)
29
-------
TABLE 3-2 (continued)
Category Surrogate Surrogate
ID ID indicator Emissions category
101
14
Land Area
Cutback Asphalt Paving Operation
102
1
Population
Fugitives from Synthetic Organic Chemical Mfg.
103
1
Population
Bulk Terminal and Bulk Plants
104
1
Population
Fugitives from Petroleum Refinery Operations
105
1
Population
Process Emissions ton Bakeries
106
1
Population
Process Emissions from Pharmaceutical Mfg.
107
1
Population
Process Emissions from Synthetic Fibers Mfg.
108
1
Population
Crude Oil and Natural Gas Production Fields
109
1
Population
Hazardous Waste Treatment, Storage, and
Disposal Facilities (TSDFs)
A89-380
30
-------
Once the subcounty distribution of each surrogate indicator was determined, area source emissions categories
were matched to the spatial fractions corresponding to the most appropriate surrogate indicators. The resultant
file, the Spatial Allocation Factor File (SAFF), is input to the Spatial Allocation Module (SAM) of the
Flexible Regional Emissions Data System (FREDS) (Modica, et al, 1989) to spatially distribute area source
emissions among 1/6° latitude by 1/4° longitude grid cells.
SPATIAL FACTOR DEVELOPMENT
The development of spatial allocation factors for the NAPAP effort (Beaulieu, et al., 1988) was based on two
main sources of data: U.S. Department of Commerce, Bureau of the Census, Census of Population (POP)
and Housing (HSG), 1980, and land use/classification data derived from 1972-1973 Landsat satellite imagery
and land use/cover maps. Separate software was created to process each type of data into a format suitable
for NAPAP applications. A simplified flow diagram of spatial factor processing is presented in Figure 3-2.
Population and housing data are processed through a series of three sequential computer programs. These
programs, written in Fortran, are executed on the NCC S perry UNI VAC. CREATE7A reads EPA's STF3A
tape (census summary tape File 3A) for each state requested and writes a condensed census file. CREATE5A
reads the condensed census file and a user-defined grid to generate subcounty gridded census data (e.g.,
population). For the State of Virginia, gridded census fractions are modified by the program VIRGINIA to
avoid double counting or omission of the state's independent cities.
The land use spatial factor processing is accomplished by a Fortran program NEWLAND, executed on the
NCC Sperry UNTVAC. Prior to NEWLAND, software developed by the Environmental Sciences Research
Laboratory (ESRL) uses the NAPAP grid and county boundary definitions from the County Dime File to
generate the county-grid relationships. These fractions are used by NEWLAND to reprocess the land use
spatial fractions supplied as the percent land use of each grid cell to the fractional county land use for each
grid cell. The resultant output file contains land use spatial fractions far 12 surrogates by state, county,
column, and row.
Once spatial fractions are generated from the Census and Landsat data, the files are merged in
SPACEMERGE to form a data set containing state, county, column, row, and 14 spatial surrogates (2 census
and 12 land use surrogates) for each record. In addition, Massachusetts land use fractions, reported relative to
county totals are modified relative to Air Pollution Control Districts using the Massachusetts adjustment data.
Once the Census and Landsat data are merged, the data are converted to EBCDIC characters and transferred
to the NCC's IBM 3090. For final processing, spatial fractions for land use and census surrogates are
matched to area source categories in the Spatial Allocation Factor Preprocessor (SAFP) using a surrogate
A89-380
31
-------
Land Use / Cover
Data
State Specific Census
Population and Housing Data
County -
Grid
Relationships
F1PS -to-
NEDS-
Codes
Grid -
Definition
Land Use Spatial
Fractions
G ridded
Census
Fractions
Massachusetts
Adjustment
Data
IBM
Conversion
Spatial Fraction
File
Surrogate
Selection
File
Used for these specific states only.
VIRGINIA'
Spatial Allocation
Factor Preprocessor
NEWLAND
SPACEMERGE
CREATE5A
CREATE7A
Spatial Allocation
Factor File
Figure 3-2. Simplified flow diagram of U.S. spatial factor processing.
32
-------
selection file containing category number and the spatial surrogate assigned to each category, SAFP writes an
EBCDIC file which is compatible with the Spatial Allocation Module of FREDS.
Spatial allocation factors for the 1985 NAPAP inventory are based on those developed for the 1980 effort.
Quality control checks performed on the spatial factors and adjustments made to the factors for use with the
1985 NAPAP inventory are discussed in this section.
Grid Description
The grid system defined for U.S. spatial factor development is comprised of 37,440 grid cells (156 rows, 240
columns), each 1/6° latitude by 1/4° longitude (approximately 400 km2). Grid cell size is a function of
latitude and varies in the United States from about 335 km2 in Lake of the Wood County, Minnesota to
466 km1 in Monroe County, Florida. The boundaries extend from 65° to 125° west longitude and from 25°
to 51° north latitude. The cell at the southwestern owner of the grid is denoted (1,1), with column numbers
increasing west to east, and rows from south to north. Column and row numbers are calculated from latitude
and longitude as follows:
The resultant integer parts are the column and row numbers; fractional parts are truncated.
Population and Housing Factors
The starting point of the census-based spatial allocation surrogates is the U.S. Department of Commerce,
Bureau of the Census, Census of Population and Housing, 1980. The file consists principally of sample data
expanded to represent the total population. However, data pertinent to persons and housing units are based
upon 100 percent counts and unweighted sample counts (U.S. Bureau of the Census, 1982),
Census data are summarized at the state (or state equivalent) level, and are broken down in hierarchical
sequence, to the following levels: counties or county equivalents, minor civil divisions (MCDs) or census
county divisions (CCDs), places or place segments within MCDs/CCDs and remainders of MCDs/CCDs,
census tracts or block numbering areas (BNAs), and block groups (BGs) or, for areas that are not
column = ((WLNG-Longitude) x DLON) + 1
row = ((Latitude - SLAT) x DLAT) + 1
(3.1)
(3.2)
where: WLNG = Western boundary of the grid (125° for NAPAP)
DLON = The number of grids/degree longitude (4 for NAPAP)
SLAT = Southern boundary of the grid (25° for NAPAP)
DLAT = The number of grids/degree latitude (6 for NAPAP)
A89-380
33
-------
block-numbered, enumeration districts (EDs). In addition, place and congressional districts (districts delineated
for the 96th Congress) are presented separately from the hierarchical summary organization.
The longitude, latitude, and land area data fields in the state-specific census population and housing Summary
Tape File 3A (STF3A) (U.S. Bureau of the Census, 1982) are blank. EPA inserted values for these fields
using the Master Area Referenced File (MARF)2 (U.S. Bureau of the Census, 1983). The modified census
tapes were used to generate the NAPAP spatial factors.
Generation of the housing and population-based spatial allocation factors involves executing three separate
Fortran programs: CREATE7A, CREATE5A, and VIRGINIA. For Massachusetts, census-based spatial
allocation factors are calculated relative to Air Pollution Control Districts (APCDs). (Massachusetts repots
emissions Anta by APCD rather than at the county level). In the census data, independent cities are reported
with the county level records for Virginia. Ten of the 41 independent cities are treated as county equivalents
for NAPAP. These include Alexandria, Chesapeake, Fairfax, Hampton, Newport News, Norfolk, Portsmouth,
Richmond, Suffolk, and Virginia Beach. The remaining 31 independent cities are merged into their respective
counties (thus the need for the program VIRGINIA).
Land Use Spatial Factors
Development of the land use derived spatial allocation factors required two data files: land use/cover
percentages for 1/6° latitude by 1/4° longitude grid cells and country-grid area relationships. The land
use/cover data base was developed by Lockheed Engineering and Management Services Company's Remote
Sensing Laboratory under contract for the Meteorology and Assessment Division of EPA's Environmental
Sciences Research Laboratory (ESRL), currently Atmospheric Sciences Research Laboratory (ASRL). The
data base was developed using Landsat mosaic images covering the periods July 23 to October 31, 1972 and
January 1 to March 31, 1973, and from Land Use and Land Cover Maps developed in the middle-to-late
1970's. The development of the land use percentages has been previously summarized (Sellars, et al, 1985).
The total land use/cover in each grid cell is divided into the following classifications:
Urban Land • Water
Agricultural Land • Barren Land
Rangeland • Nonforested Wetland
Deciduous Forest Land • Mixed Agricultural Land
and Rangeland
Coniferous Forest Land
• Rocky and open areas occupied by low
Mixed Forest Land (includes growing shrubs and lichens
forested wetlands)
A89-380
34
-------
Each land classification percentage reported by Lockheed represents the portion of a grid cell comprised of a
specific land use category. For NAPAP spatial factor processing, the category "Rocky and open area
occupied by low growing shrubs and lichen" was not used. In addition to the eleven land use/cover
categories supplied by Lockheed, two additional categories were
calculated for NAPAP:
• Composite Forest • Land Area
Composite Forest is comprised of coniferous, deciduous, and mixed forest lands for each grid cell (this
summation is performed prior to the allocation of grid cells to counties). Land area is the fraction of a
county's total land area contained within a grid cell and is calculated from data in the county-grid file
discussed below.
Use of the land classification data for allocation of county-wide emissions totals to grid cells required that the
gridded land use/cover data be summed to the county level. To allocate grid cells to counties, a data file
containing county-grid relationships was created. The county-grid file was generated using a file containing
county boundaries (the County Dime File) in conjunction with the NAPAP grid system and software
developed by ESRL. County-grid relationships were calculated as the fraction of each county in each grid
cell. These relationships were used to aggregate the grid-level land use data to the county-level such that
county-level allocation factors could be determined. An example of this process is illustrated in Figure 3-3,
Applying the county-grid file to the land use/cover percentages allows gridded land use fractions to be
expressed as the fraction of a county's land use/cover in a grid cell. Spatial allocation factors for each land
use classification are calculated as follows:
(AJ (AJ
SAFai =
jp (AJ (AJ
i = 1
where: SAF^ = The spatial allocation factor for County C, land
use type X, and grid i.
Aq = The portion of County C that falls within grid i.
Ae = The portion of grid i with land use type X.
n = The number of grids covering County C.
Hie fractional totals for each county sum to 1; fractional totals by grid cell do not aim to a particular
number. The land use and county-grid fractions are processed by the program NEWLAND.
A89-380
35
-------
COUNTY A
AREA = 500
COUNTY B
AREA = 1000
OVERLAP IN COUNTY A
/ GRID CELL X
/ AREA = 150
OVERLAP IN COUNTY
GRID CELL X
AREA = 200
COUNTY C
AREA = 800
OVERLAP IN COUNTY B
COUNTY B GRID CELL X
AREA = 50
Figure 3-3. Example of county-to-grid cell areal relationship.
Source: National Acid Precipitation Assessment Program Emission
Inventory Allocation Factors (Sellars et al., 1985).
36
-------
QUALITY CONTROL AND ENHANCEMENTS OF SPATIAL FACTORS FOR THE 1985 NAPAP
INVENTORY
Spatial allocation factors were originally developed for processing the 1980 NAPAP area source inventory.
To assure the quality of the spatially resolved 1985 area-source emissions, extensive quality control checks
were performed cm spatial factors adopted from the previous NAPAP data base. The goal of the spatial
factor quality control procedures was to ensure complete and accurate apportionment of NAPAP area source
emissions. Based on the results of these checks, adjustments were made to the spatial fractions for use with
die 1985 NAPAP inventory. Additional modifications to the file were required for spatial allocation of
emissions from the new 1985 NAPAP area source categories and Cibola County, New Mexico. Quality
Control procedures and modifications made to the Spatial Allocation factors are discussed below and in
Appendix A.
Quality Control Procedures
Quality control procedures for the spatial allocation factors fell into three general categories: evaluation of
emissions losses during inventory processing in the Spatial Allocation Module (SAM) (based on the Quality
Control Module (QCM) processing messages); analyses of counties requiring normalization for spatial
surrogates deviating by 20 percent or more as noted in the Spatial Allocation Factor Preprocessor (SAFP)
processing messages; and "other analyses" including a review of the spatial factor computer programs and the
Massachusetts county-to-APCD conversion.
The causes of the emissions losses during inventory processing were evaluated by comparing county- and
SCC-level emissions totals before aid after spatial allocation, matching stale/county codes in the emissions file
with those in the Spatial Allocation Factor File (SAFF), and tracking specific counties through various stages
of the spatial factor processing. Counties and surrogates identified by the SAFP processing messages as
requiring excessive normalization were analyzed using grid maps for each state, atlases, and data from various
segments of the spatial factor development process. In addition, segmented maps were plotted as necessary
and included the locations of various towns for reference. Location data for these maps were obtained from
the National Cartographic Information Center. Other quality control procedures included close examination of
the spatial software and applications of these programs for other projects.
Quality control checks applied to the 1980 NAPAP spatial allocation factors revealed several problems with
the data. These are summarized in Table 3-3 and are discussed in greater detail in Appendix A.
Adjustments to Spatial Allocation Factors
For 1985 NAPAP applications, adjustments were made to the 1980 spatial fractions to account for missing
and new counties, missing grid cells, incorrect county code assignments and location data, inaccuracies in the
A89 380
37
-------
TABLE 3-3. SUMMARY OF 1980 SPATIAL FACTOR PROBLEMS AND IMPACTS®
Problem
Areas Affected
Reason
Impacts
Missing Counties
1. From Land Use File
2. From FIPAEROS File0
3. New County
Missing Grid Cells
1. From Land Use File
County Code Assignments
1, FIPAEROS File0
Sabine, LA; Barbour, WV
Alexandria, VA; Chesapeake,
VA; Fairfax, VA; Portsmouth,
VA; Richmond, VA
Unknown
File
Present in CNTY2GRID
Menominee, WI
Nansemond, VA
Cibola, NM
Some Virginia independent cities
were incorrectly coded in
CNTY2GRID File
Menominee is an Indian reserva-
tion within Shawno County. For
Census processing, Menominee was
assigned the county code for
Shawno
Instead of assigning Suffolk
independent city to Nansemond
County, the code for Suffolk
was used
Separated from Valencia County
in 1982; not a county for 1980
Alexander, IL; Queen Annes, Grid cells not in CNTY2GRID File
Monroe, FL
Saginaw, MI
Some segments below NAPAP
southern boundary; the
remainder is unknown
Input as two county codes -
4600 and 4780 in land use;
4600 only in Census data.
County 4600 does not exist so
emissions are matched to 4780
which shows no Census fractions
Loss of emissions
Loss of emissions
Loss of emissions
Loss of emissions
Loss of emissions
Incorrect Census fractions
(off by 20% or more before
normalization)
Incorrect Census fractions
(off by 20% or more before
normalization)
Incorrect Census fractions
(off by 20% or more before
normalization)
(continued)
-------
TABLE 3-3
(continued)
Problem
Areas Affected
Reason
Impacts
Jackson, SD; Washabaugh, SD
Treated as separate counties in
land use and emissions files,
but as single county in Census
data (for certain applications,
these are treated as a single
county)
Incorrect Census fractions
(off by 20% or more before
normalization)
2. Land Use File
Lincoln, OR; Linn, OR
These were interchanged in the
land use file - grids belonging
to Lincoln were assigned to Linn
and vice versa
Incorrect Census fractions
(off by 20% or more before
normalization)
Location Data
1, Census File
Rappahannock, VA
Grid cell located outside county
Incorrect Census fractions
(off by 20% or more before
normalization)
Adams, WI
Latitude/longitude of county
and subcounty centroids do not
correspond to Adams County grid
cells
Incorrect Census fractions
(off by 20% or more before
normalization)
2. Independent city assigned
to county which it is not
within geographically
Pulaski, VA
Unknown
Incorrect Census fractions
(off by 20% or more before
normalization)
Massachusetts Countv to
APCD Conversion
1. County to APCD Area
Adjustment
Berkshire, Central Massa-
chusetts, Merrimack Valley,
Metropolitan Boston, Pioneer
Valley, and Southeastern
Massachusetts APCDs
County to APCD area conversion
was not done in SPACEMERGE;
Border grid cells assigned to
only one grid cell based on a
value of population rather than
area
Spatial distribution of
emissions not accurately
represented
Aleorithas
1. CREATE5A column and row
equations
Border grid cells when a
subset of the NAPAP grid
is specified (incorrect
coordinates)
Equation does not account for a
subset of the NAPAP grid
Does not impact NAPAP
(continued)
-------
TABLE 3-3 (continued)
Problem Areas Affected Reason Impacts
2.
SPACIMERGE grid to column
and row equations
Eastern border grids
Equation does not correctly deal
with the eastern boundary of the
grid
Land use data in the eastern
boundary is shifted one cell
to the north; the COL
equation yields a column
number of 0 instead of 2U0
3.
SAFP Normalization
County totals off by less
than 5%
Algorithm only checks for
county/surrogate totals off
by 5% or more; normalization
algorithm doesn't allow county
to sum exactly to 100%
(approaches it, but never
reaches it)
Minor Iofs of emissions
(<0.1%)
¦fc.
o
4.
SAFP Combination
Massachusetts
Equation did not combine each
succeeding state, county, col,
row combination which matched
the previous record. Instead
it added the spatial fractions
from the first similar record
to itself
Spatial distribution of
emissions not accurately
represented
aMore complete discussions of spatial factor problems and adjustments are presented in Appendix A.
^Contains county-to-grid fractions
cContains FTPS to NEDS code conversions
-------
Massachusetts county-to-APCD conversion and the new area source categories. All corrections were applied
to the SAFP input file (the spatial fraction file) and are summarized in Table 3-4. Once adjustments were
made to the spatial fraction file, SAFP was executed and the processing messages were evaluated to assure
the factors were correctly applied. Corrections and enhancements to the spatial fractions are also summarized
in Appendix A.
SPATIAL ALLOCATION FACTOR FILE FORMAT
The Spatial Allocation Factor File is an EBCDIC file containing one record type. Each record contains all of
the spatial allocation information for a particular SCC in a specific county. The detailed format is contained
in Table 3-5.
A89-380
41
-------
TABLE 3-4. SUMMARY OF SPATIAL FACTOR ADJUSTMENTS FOR THE 1985 NAPAP INVENTORY*
Problem
Areas affected
Adjustments
Missing counties
Alexandria, VA; Chesapeake, VA;
Fairfax, VA; Nansemond, VA;
Portsmouth, VA; Richmond, VA;
Barbour, WV
Population and housing fractions
available from a previous project
Cibola, NM
Adjust population, housing, and
total land area fractions for
Valencia to account for new
county boundary.
Sabine, LA; Menominee, WI
Calculate land area fractions and
insert into population field.
Missing grid cells
Queen Amies, MD
Population and housing fractions
available from a previews project
Alexander, IL
Insert missing population and
housing fraction into single
missing grid cell.
Monroe, FL; Dare, NC
Calculate population fractions for
missing grid cells from town
populations and latitude/longitude
data.
County code
assignments
Saginaw, MI
Verify correct county code and
adjust in spatial fraction file.
Lincoln, OR; Linn, OR; Jackson,
SD; Washabaugh, SD
Insert land area fractions into
the population field; zero out
housing values.
Location data
Pulaski, VA
Normalize in SAFP.
Rappahannock, VA
Assign missing population and
housing fractions to appropriate
grid cell.
Adams, WI
Substitute land area fractions
in population field.
Massachusetts county
to APCD conversion
Berkshire, Central Massachusetts,
Merrimack Valley, Metropolitan
Boston, Pioneer Valley, and
Southeastern Massachusetts
Calculate new land area fractions;
zero out all other land use
spatial fraction fields.
Algorithms
Modify all problematic algorithms
for future processing.
'A mere complete discussion of spatial factor adjustments and enhancements is presented in Appendix A,
A89-380
42
-------
TABLE 3-5. FILE FORMAT FOR THE EBCDIC 1985 SPATIAL ALLOCATION FILE
Record position
Column Variable
First Last width Format name Description
1
2
2
I
STATE
AEROS State Code
3
6
4
I
COUNTY
AEROS County Code
7
9
3
I
see
Area Source Category Code
10
12
3
I
NGRID
Number of Grids in County
13
16
4
I
NROW
Row of Grid Cell
17
20
4
I
NCOL
Column of Grid Cell
21
25
5
F5.4
NSAP
Spatial Allocation Facta- for
Grid Cell*
'Implied decimal, with four digits to the right of the decimal place and one to the Ml
Spatial allocation factors are dimensionless fractions representing the portion of a county's land
use for a specific land use category in a grid cell.
Note: Columns 13 to 25 are repealed up to NGRID times.
A89-380
43
-------
SECTION 4
SPECIES ALLOCATION FACTORS
INTRODUCTION
One of the primary end uses of the 1985 NAPAP inventory is as input to models such as the Regional Acid
Deposition Model (RADM). RADM includes a chemical reaction mechanism to simulate the transformation
of primary emissions into stable product species during atmospheric transport. Certain pollutants in the
NAPAP annual inventory represent composites of various reported compounds in an aggregated format too
general to be useful as input to the model. As a result, three pollutants - total hydrocarbons (THC), oxides
of nitrogen (NOJ, and total suspended particulates (TSP) - must be resolved into component species or
groups of species which share similar reaction chemistry prior to RADM input Specifically, the NAPAP
modeler's inventory includes NO and N02, 32 classes of hydrocarbon species, and 15 particulate fractions
based on alkalinity and size distribution.
The allocation of pollutants to species classes is accomplished using speciation profiles, which provide a
breakdown of THC, TSP, and NO, emissions to a typical set of component classes. The profiles are
generally process-specific, and are matched with emissions records at the level of point source SCC or
NEDS/NAPAP area source category.
The earliest speciation efforts, using the Regional Model Data Handling System (RMDHS), could
accommodate only ten hydrocarbon classes and no particulate speciation was attempted. With the
development of the Flexible Regional Emissions Data System, a significant increase in the number of
allowable speciated components was achieved. The final version of the base year 1980 inventory contained
28 hydrocarbon species, as well as a preliminary allocation of particulate matter into four alkalinity classes.
In support of the expanded RADM requirements for the 1985 inventory, and the increasing need for
representative speciation data for numerous EPA research efforts, Radian Corporation performed a
comprehensive review and update of existing hydrocarbon and particulate speciation data (Shareef, et al.,
1988). These data were reviewed by Alliance and subsequently manipulated into a format consistent with
NAPAP inventory processing needs. In addition, extensive modifications to FREDS were undertaken to
accommodate an increased number and expanded format for the 1985 species classes.
A89-380
44
-------
The remainder of this section describes in detail the development of the speeiation factors used in the 1985
NAPAP modeling inventory, including sources of data, quality assurance, and procedures used to convert data
to a format suitable for NAPAP. Also included is a section describing the preprocessing of hydrocarbon
emissions using data from the speeiation files.
HYDROCARBON/NOx SPECIATION FACTOR DEVELOPMENT
Methodology
Regional models such as RADM employ algorithms to simulate the complex transformation processes
occurring in the atmosphere. These reaction mechanisms are applied to classes of organic compounds, each
of which may include a number of individual species with similar reaction properties and reaction products.
For example, several C4 through C6 olefine may be satisfactorily expressed by a single set of reaction
statements and, therefore, be represented by a single species class in the model. The RADM species classes
used in the 1985 NAPAP modeling inventory are depicted in Table 4-1.
Since early in the NAPAP inventory development effort, it has been recognized that flexibility in speeiation is
desirable to accommodate the needs of different modeling chemistries. Speciated VOC data are also required
for other EPA activities including ozone and air toxics studies. Therefore, hydrocarbon species data are coded
as a set of species profiles, each of which provides a breakdown of the representative component species for
a source's hydrocarbon emissions. The profiles can then be manipulated by the end user to represent any
desired classification scheme.
For the NAPAP inventory, species profiles and class assignments are used to create speeiation factors for
hydrocarbons by means of a computer program called the Pollutant Splits System (PSPLIT). PSPLIT is a
Fortran program that can divide hydrocarbon emissions into over 600 discrete species based on Source
Classification Code (SCC) specific profiles. The individual species are subsequently reaggregated into up to
32 classes based on the requirements of the selected transport model.
Three data files are required as input to PSPLIT. The Species-Profile file contains the weight percent
contribution of each species to each profile. Each individual species is assigned to a reactivity class in the
Species-Class file. Finally, profiles are matched to NAPAP emissions records using the SCC-Profile Index
file. PSPLIT performs the appropriate processing steps to calculate the amount of total hydrocarbon that
contributes to each class, for every source category in the inventory. The resulting mole factors are expressed
in moles/kilogram and when multiplied by the mass of total hydrocarbon from a source, they yield the
number of moles of each class emitted.
A89-380
45
-------
TABLE 4-1. HYDROCARBON SPECIES CLASSES, 1985 NAPAP
RESOLVED INVENTORY
Class number Description
1
Methane
2
Ethane
3
Propane
4
Alkanes (0.25 - 0.50 reactive)*
5
Alkanes (0.50 - 1.00 reactive)
6
Alkanes (1.00 - 2.00 reactive)
7
Alkanes (> 2.00 reactive)
8
Alkane/aromatic mix
9
Ethene
10
Propene
11
Alkenes (primary)
12
Alkenes (internal)
13
Alkenes (primary/internal mix)
14
Benzene, halobenzenes
15
Aromatics (< 2.00 reactive)
16
Aromatics (> 2.00 reactive)
17
Phenols and cresols
18
Styrenes
19
Formaldehyde
20
Higher aldehydes
21
Acetone
22
Higher ketones
23
Organic acids
24
Acetylene
25
Haloalkenes
26
Unreactive
27
Others (< 0.25 reactive)
28
Others (0.25 - 0.50 reactive)
29
Others (0.50 - 1,00 reactive)
30
Others (> 1.00 reactive)
31
Unidentified
32
Unassigned
"Reactivity is defined with respect to rate constant range (with OH), 104 ppm"1 min
A89-380
46
-------
PSPLIT is also used to disaggregate NO, emissions into NO and NOz components. The species profiles input
to PSPLIT contain the weight fractions of NO and N02, which are carried through the program and become
the speciaxion factors for these two components,
Speciatlon File Development
The source of the hydrocarbon species profiles used in the 1985 NAPAP inventory is Radian Corporation's
updated Air Emissions Species Manual (Shareef, et al, 1988). This document is largely based on the Volatile
Organic Compound (VOC) Specks Data Manual - Second Edition (EPA, 1980), which in turn formed the
basis for earlier NAPAP speciation data gathering efforts. Revisions to the 1980 VOC Data Manual consisted
of identifying and replacing poor quality and/or outdated profiles, and adding profiles to fill data gaps
primarily in the area of air toxics.
New profiles were developed as a result of a literature search. Key sources of information included: a study
conducted for the California Air Resources Board; source assessment and background information documents;
Section 114 responses and trip reports from standard development activities; and work conducted by the
Atmospheric Research and Exposure Assessment Laboratory (AREAL). Data from the Northeast Corridor
Regional Modeling Project (NECRMP), current and past NAPAP work, Environment Canada, and State
agencies were also reviewed. The effort resulted in the addition of nearly 200 new profiles, 80 percent of
which covered source categories not characterized in the 1980 manual. An additional five profiles were
developed from a speciation test program to characterize high priority emission sources. These profiles
represent the following source categories: gasoline marketed, degreasing, dry cleaning, auto body repair, and
graphic arts/printing.
In addition to profile development, Radian assigned each NEDS point source SCC and area source category to
a hydrocarbon profile. Ideally, a separate profile would be needed to represent each category. However, the
number of categories is significantly larger than the number of available profiles. Source categories without
original profile assignments were, therefore, assigned to existing profiles using engineering judgement. In
addition, 43 industry-specific average profiles were developed and applied to source categories for which a
satisfactory assignment based on engineering judgement was not possible. Finally, for those sources not
expected to report hydrocarbon emissions, a "zero" profile was generated; this profile is an overall average
based on all the other profiles in the data base. The zero profile was intended for use only with SCCs
characterized by "zero" or "negligible" VOC emission factors for which a State or other agency may have
reported nonzero emissions.
The assignment of the nearly 650 individual hydrocarbon species to classes consistent with RADM chemistry
was a collaborative effort coordinated by the National Center for Atmospheric Research (NCAR) representing
NAPAP Task Group HI (Atmospheric Transport and Modeling). Species classification was an iterative
A89-380
47
-------
procedure involving a number of individuals from the modeling community. During the process, the number
of classes desired increased considerably in an attempt to accommodate as many different mechanisms as
possible. Constraints on the number of classes which could be processed through NAPAP software limited
the final class total to 32, although a 50-category scheme was also proposed. The NCAR classification
scheme is to be used both for RADM and the Acid Deposition and Oxidant Model (ADOM); a separate
classification mechanism based on carbon bond classes is being developed for the Regional Oxidant Modeling
for Northeast Transport (ROM.NET) program.
NO, speciation fractions represent NO and N02 fractions recommended by AP-42 (EPA, 1985), and were
adapted from the 1980 data with little modification. In previous applications, profiles that were applied to an
SCC that was not expected to have NO, emissions were assigned profile splits of zero for both NO and N02.
It was of concern that NO, emissions could be lost during the FREDS speciation step if a source with NO,
emissions was misassigned to an SCC that was represented by zero splits for NO and N02. To ensure that
no NO, emissions are lost, a default split of 95 percent NO and 5 percent N02 was substituted for every
profile which did not have an explicit N0/N02 split.
Quality Control Checking
Prior to PSPLIT execution, the three component speciation files were subjected to a quality control review
involving manual and automated data checks. These screening checks were designed primarily to detect
missing, invalid, duplicate, or inconsistent data elements within the computer files. No attempt was made to
verify the quality of the data upon which the speciation files were based. All findings were reported to
Radian for review and correction.
One of the Radian profiles, corresponding to diesel vehicle exhaust emissions, was identified as too general
for use in the NAPAP speciation methodology. The profile was based on analyses of diesel fuels and
included species identified sequentially as "C2 compounds" through "C43 compounds". Neither the structure
nor the chemical makeup of these 42 species were provided, and assignment to RADM classes was therefore
not possible. The profile for diesel vehicle exhaust emissions developed for the 1980 inventory was modified
based on information supplied by EPA (Black, F. personal communication with M. Saegar, Alliance), and
used as a substitute for the 1985 inventory application.
The original 1980 profile was modified by adding the species o-xylene to represent the appropriate weight
percent of high carbon number aromatics in diesel exhaust. The weight percent of n-pentadecane in the
profile was decreased by the weight percent assigned to o-xylene.
A89-380
48
-------
A total of five area source categories (66 through 70) added late in the 1985 annual inventory development
were not assigned to profiles by Radian. NAPAP data processing software requires that all source categories
be assigned to profiles regardless of whether or not the category is likely to include hydrocarbon emissions.
Therefore, the vehicular source ammonia categories (66, 67, and 68) and the new livestock waste management
ammonia categories (69 and 70) were all assigned to the global default (0) hydrocarbon species profile.
The data processing requirements of the 1985 inventory made it necessary to revise the original PSPLIT
computer code. This effort resulted in the creation of a new version of the program designated PSPLIT_2A.
The new version differs from earlier versions as follows:
• The maximum allowable number of species per profile was increased from 75 to 325;
• The maximum allowable number of species in all classes was increased from 500 to 640;
• The maximum allowable number of SCC - Profile index records was increased from 1,800 to
4,000;
• The maximum allowable number of profiles was increased from 175 to 320;
• The maximum allowable number of species-level profile records was increased from 2,000 to
4,500; and
• The maximum allowable number of species classes was increased from 30 to 32.
In addition, the format of the output speciation factors was increased from four digits to five digits to
increase the precision of species emission estimates.
FREDS Input File Format
The output of PSPLIT_2A for the RADM modeling inventory (known simply as the PSPLIT File) contains
mole factors for the 32 hydrocarbon classes and weight percents for NO and N02 for each point and area
source category in the annual file. The file is formatted as illustrated in Table 4-2, allowing it to be read by
the FREDS Speciation Module. The file header information (record type 1) occupies the first two lines and
contains codes which correspond to each of the 32 classes, 16 per line. The remainder of the file consists of
a series of two-line speciation records (type 2), with 16 hydrocarbon speciation factors per line.
PARTICULATE SPECIATION FACTOR DEVELOPMENT
Methodology
Speciation factors for pollutants not handled by PSPLIT can be input to FREDS using an independent
Speciation Factor File (SFF). For the 1985 modeling inventory, the SFF contains factors for 15 subspecies of
particulate matter. Since particle size is an important criterion in TSP transport mechanisms, particulates are
A89-380
49
-------
TABLE 4-2. PSPUT FILE FORMAT
Record position
First Last
Column
width
Format
Variable
name
Description
Header
Record No. 1
1
5
5
I
CLASS 1
Pseudo-SAROAD code
for THC
Class
No.
1
6
10
5
I
CLASS2
Pseudo-SAROAD code
for THC
Class
No.
2
11
15
5
I
CLASS3
Pseudo-SAROAD code
for THC
Class
No.
3
16
20
5
I
CLASS4
Pseudo-SAROAD code
for THC
Class
No.
4
21
25
5
I
CLASS5
Pseudo-SAROAD code
to THC
Class
No.
5
26
30
5
I
CLASS6
Pseudo-SAROAD code
for THC
Class
No.
6
31
35
5
I
CLASS7
Pseudo-SAROAD code
for THC
Class
No.
7
36
40
5
I
CLASS8
Pseudo-SAROAD code
for THC
Class
No.
8
41
45
5
I
CLASS9
Pseudo-SAROAD code
for THC
Class
No.
9
46
50
5
I
CLASS10
Pseudo-SAROAD code
for THC
Class
No.
10
51
55
5
I
CLASS 11
Pseudo-SAROAD code
for THC
Class
No.
11
56
60
5
I
CLASS 12
Pseudo-SAROAD code
for THC
Class
No.
12
61
65
5
I
CLASS 13
Pseudo-SAROAD code
for THC
Class
No.
13
66
70
5
I
CLASS 14
Pseudo-SAROAD code
for THC
Class
No.
14
71
75
5
I
CLASS 15
Pseudo-SAROAD code
for THC
Class
No.
15
76
80
5
I
CLASS16
Pseudo-SAROAD code
for THC
Class
No.
16
81
113
33
-
—
Unused
114
115
2
I
TYPE
Record Type (=11)
Header
Record No. 2
1
5
5
I
CLASS 17
Pseudo-SAROAD code
for THC
Class
No.
17
6
10
5
I
CLASS 18
Pseudo-SAROAD code
for THC
Class
No.
18
11
15
5
I
CLASS19
Pseudo-SAROAD code
for THC
Class
No.
19
16
20
5
I
CLASS20
Pseudo-SAROAD code
for THC
Class
No.
20
21
25
5
I
CLASS21
Pseudo-SAROAD code
for THC
Class
No.
21
26
30
5
I
CLASS22
Pseudo-SAROAD code
for THC
Class
No.
22
31
35
5
I
CLASS23
Pseudo-SAROAD code
for THC
Class
No.
23
36
40
5
I
CLASS24
Pseudo-SAROAD code
for THC
Class
No.
24
41
45
5
I
CLASS25
Pseudo-SAROAD code
for THC
Class
No.
25
46
50
5
I
CLASS26
Pseudo-SAROAD code
to THC
Class
No.
26
51
55
5
I
CLASS27
Pseudo-SAROAD code
for THC
Class
No.
27
56
60
5
I
CLASS28
Pseudo-SAROAD code
for THC
Class
No.
28
61
65
5
I
CLASS29
Pseudo-SAROAD code
to THC
Class
No.
29
66
70
5
I
CLASS30
Pseudo-SAROAD code
to THC
Class
No.
30
71
75
5
I
CLASS31
Pseudo-SAROAD code
for THC
Class
No.
31
76
80
5
I
CLASS32
Pseudo-SAROAD code
to THC
Class
No.
32
81
113
33
-
Unused
114
115
2
I
TYPE
Record Type (=12)
(continued)
A89-380
50
-------
TABLE 4-2 (continued)
Record position
—.. Column Variable
First Last width Format name Description
Speciation Record No. 1
1
2
2
I
STATE
AEROS State Code
3
6
4
I
COUNTY
AEROS County Code
7
10
4
A
PLANT ID
Plant Identification Code*
11
13
3
A
POINT Id
Point Identification Code-
14
21
8
I
see
Area Source Category or Point Source SCC
22
26
5
F5V3"
TSF1
Speciation Factor, THC Class 1°
27
31
5
F5V3
TSF2
Speciation Factor, THC Class 2'
32
36
5
F5V3
TSF3
Speciation Factor, THC Class 3"
37
41
5
F5V3
TSF4
Speciation Factor, THC Class 4'
42
46
5
F5V3
TSF5
Speciation Factor, THC Class 5'
47
51
5
F5V3
TSF6
Speciation Factor, THC Class 6'
52
56
5
F5V3
TSF7
Speciation Factor, THC Class T
57
61
5
F5V3
TSF8
Speciation Factor, THC Class 8"
62
66
5
F5V3
TSF9
Speciation Factor, THC Class 9*
67
71
5
F5V3
TSF10
Speciation Factor, THC Class 10"
72
76
5
F5V3
TSF11
Speciation Factor, THC Class 11'
77
81
5
F5V3
TSF12
Speciation Factor, THC Class 12"
82
86
5
F5V3
TSF13
Speciation Factor, THC Class 13'
87
91
5
F5V3
TSF14
Speciation Factor, THC Class 14'
92
%
5
F5V3
TSF15
Speciation Factor, THC Class 15'
97
101
5
F5V3
TSF16
Speciation Factor, THC Class 16*
102
105
4
F4V4
HCHOWT
Weight Fraction Formaldehyde in Profile
(dimensionless)
106
109
4
F4V3
TSF33
Speciation Factor for NOz (dimensionless weight
fraction)
110
113
4
F4V3
TSF34
Speciation Factor for N02 (dimensionless weight
fraction)
114
115
2
I
TYPE
Record Type (=21)
Speciation Record No. 2
1
2
2
I
STATE
AEROS State Code
3
6
4
I
COUNTY
AEROS County Code
7
10
4
A
PLANT ID
Plant Identification Code*
11
13
3
A
POINT Id
Point Identification Code*
14
21
8
I
SCC
Area Source Category or Point Source SCC
22
26
5
F5V3b
TSF17
Speciation Factor, THC Class 17*
27
31
5
F5V3
TSF18
Speciation Factor, THC Class 18*
32
36
5
F5V3
TSF19
Speciation Factor, THC Class 19"
37
41
5
F5V3
TSF20
Speciation Factor, THC Class 20*
42
46
5
F5Y3
TSF21
Speciation Factor, THC Class 21'
47
51
5
F5V3
TSF22
Speciation Factor, THC Class 2?
52
56
5
F5V3
TSF23
Speciation Factor, THC Class 23*
A89-380
(continued)
51
-------
TABLE 4-2 (continued)
Record position
First Last
Column
width
Format
Variable
name
Description
57
61
5
F5V3
TSF24
Speciation Factor, THC Class 24c
62
66
5
F5V3
TSF25
Speciation Factor, THC Class 25c
67
71
5
F5V3
TSF26
Speciation Factor, THC Class 26c
72
76
5
F5V3
TSF27
Speciation Factor, THC Class 27°
77
81
5
F5V3
TSF28
Speciation Factor, THC Class 28c
82
86
5
F5V3
TSF29
Speciation Factor, THC Class 29c
87
91
5
F5V3
TSF30
Speciation Factor, THC Class 30*
92
96
5
F5V3
TSF31
Speciation Factor, THC Class 31'
97
101
5
F5V3
TSF32
Speciation Factor, THC Class 32"
102
113
12
-
—
Unused
114
115
2
I
TYPE
Record Type (=22)
"For area source processing, these fields are left blank.
¦"Speciation factors are 5-digit numbers with an implied decimal point to the right of the second digit.
"Units are moles of the THC class per kilogram of total hydrocarbon.
A89-380
52
-------
"speciated" by both alkalinity and size fraction. Each of the four alkaline dust species in the inventory—
calcium, magnesium, sodium, and potassium—is expressed in terms of 0 to 2,5 micron fraction, 2,5 to 10
micron fraction, and total species. In addition, TSP is speciated into three size ranges including 0 to 2.5, 2.5
to 6, and 6 tt) 10 micron components. As with NO, factors, speciation factors for particulates are
dimensionless weight fractions which are multiplied by TSP emissions to yield an estimate of the number of
tons of each species class.
A complete listing of particulate species classes in the 1985 modeling inventory is shown in Table 4-3.
Unlike other pollutants in the inventory, these classes cannot be uniquely defined using five-digit Storage and
Retrieval of Aerometric Data (SAROAD) pollutant codes. In order to distinguish by both species and size
fraction, a two-digit size class code was developed. The code is concatenated to the front of the species
SAROAD code to create a unique seven-digit identifier. When input to FREDS, the two codes are interpreted
as one for the purpose of TSP speciation.
Input File Development
Data for speciated particulate emissions were extremely limited during the early stages of NAPAP inventory
development. For the 1980 inventory, TSP emissions were speciated into four alkaline dust categories;
however, speciation factors were available for only 10 percent of all point source SCCs. Emission factors
were developed by Meteorological and Environmental Planning Limited and the Ontario Research Foundation
and incorporated into the inventory without change. Generic source descriptions used in development of the
factors were matched to point source SCCs at the six-digit level. No attempt was made to speciate
anthropogenic area source categories.
As in the case of hydrocarbons, Radian's Air Emissions Species Manual is the primary source of data for
speciated particulate matter (PM) for the 1985 NAPAP inventory. The Radian study, conducted in support of
NAPAP as well as PM10 source apportionment studies, consisted of an analysis of the profiles contained in
the Receptor Model Source Composition Library (EPA, 1984) and the development of new profiles based on
literature search efforts. The Radian data base consists of 131 "original" profiles, 11 composite profiles
(which represent groupings of several profiles available for the same category), 16 industry-specific average
(default) profiles, and an overall average or "zero" profile.
For each profile, the following data were compiled:
• Composition data for four size ranges: 0 to 2.5 um, 2,5 to 10 um, 0 to 10 um, mid total
particulate. These data are expressed as weight percent of the given size fraction occurring as a
certain species.
• Mass Fraction data for the 0 to 2.5, 0 to 6, and 0 to 10 um size ranges. These represent the
fraction of total particulate mass contained within the specific size range.
A89-380
53
-------
TABLE 4-3. PARTICULATE SPECIES CLASSES, 1985 NAPAP RESOLVED INVENTORY
Size
Species
class code
SAROAD code
Description
10*
12111"
Reactive calcium (Ca), 0-2.5 micron size range
30
12111
Reactive calcium, 2.5-10 micron size range
50
12111
Reactive calcium, total
. 10
12140
Reactive magnesium (Mg), 0-2.5 micron size range
30
12140
Reactive magnesium, 2.5-10 micron size range
50
12140
Reactive magnesium, total
10
12180
Reactive potassium (K), 0-2.5 micron size range
30
12180
Reactive potassium, 2,5-10 micron size range
50
12180
Reactive potassium, total
10
12184
Reactive sodium (Na), 0-2.5 micron size range
30
12184
Reactive sodium, 2.5-10 micron size range
50
12184
Reactive sodium, total
10
11101
Total particulates, 0-2.5 micron size range
20
11101
Total particulates, 2.5-6 micron size range
40
11101
Total particulates, 6-10 micron size range
'Size class codes are defined as follows:
Code
Size range (microns)
10
0 - 2.5
20
2.5 - 6.0
30
2.5 - 10.0
40
6.0 - 10.0
50
Total
"Species SAROAD
codes are defined as foEows:
SAROAD
Species
12111
Calcium (Ca)
12140
Magnesium (Mg)
12180
Potassium (K)
12184
Sodium (Na)
11101
Total particulates
A89-380
54
-------
Reactivity data for the alkaline particulate species (Na» K, Ca, and Mg). The "reactive fraction"
is defined as that fraction of each element which is present as an oxide, carbonate, carbide, or
hydroxide. These are species in which the cation is readily dissociated and available for reaction
with acidic solutions.
In addition, as part of the profile development effort, Radian assigned all NEDS point and area source
categories to particulate speciation profiles. The specific particulate classes required for RADM were selected
by personnel from NCAR and EPA's Atmospheric Research and Exposure Assessment Laboratory (AREAL).
Speciation Factor File Generation and Quality Control
In order to convert the profile data into speciation factors suitable for NAPAP, the composition, mass
fraction, and reactivity data must be combined so as to represent the reactive fraction of a given species
within a given size fraction. For example, to calculate the speciation factor representing the reactive fraction
of species X within size fraction Y for profile Z the following equation is used:
SF„Z = _L_ (WP^vz) (MFyj) (RF^) (4.1)
100
where: WP^ = weight percent of size fraction Y occupied by species X, profile Z;
MFY7 = mass fraction of total particulate occurring within size fraction Y, profile Z; and
RF^z = reactive fraction of species X, profile Z.
Two computer programs were developed to derive speciation factors from Radian data. The first program
utilized the profile, fraction, and reactivity data to create speciation fractions using equation 4.1 above. The
Radian profile data were reformatted and reduced to contain only the four species relevant to NAPAP (Na,
Ca, K, and Mg), at the size ranges desired. These were multiplied by the fraction of total TSP occurring at
each size range, and finally by the reactive fractions of the four species. The resulting "species-profile" file
was used as input to a second program, which matched the speciation factors to source categories using
SCC-profile index files, and wrote the resulting data set in FREDS -compatible EBCDIC format.
Quality control checks of the Radian particulate speciation data consisted of a series of manual and automated
procedures designed primarily to detect problems which would prevent successful FREDS execution. Missing
or invalid data elements, duplicate SCC-profile combinations and point source SCCs missing profile
assignments were detected in early versions of the data. These were reported and subsequently corrected by
Radian. Profile assignments for new NAPAP area source categories were made by Alliance. The PM default
profile (0) was employed for the mobile source ammonia categories (66-68). Livestock waste management
categories for turkeys aid sheep (69 and 70) were assigned to profile 90003, (food ami agriculture-industry
average) consistent with Radian's assignments for the similar categories 70 through 75.
A89-380
55
-------
During early tests of speciation data, discrepancies between the mass of the speeiated size fractions and the
total species masses were discovered. Of the 159 particulate profiles, 20 were identified in which the sum of
the 0 to 2.5 micron fraction and the 2.5 to 10 micron fraction exceeded the total species estimate.
Examination of the profiles indicated that the problem was associated with the calculation erf the total species
fraction. In the absence of an algorithm with which to recalculate the fraction, the total species values were
replaced with fractions representing the minimum allowable total species value. The 0 to 10 micron fraction
was used for 15 of the 20 profiles; the remainder were calculated on a case-by-case basis. Although the
procedure resulted in internal consistency for the profiles, it raises some concern as to the validity of other
total mass fractions in the profiles.
Speciation Factor File Format
An example of Speciation Factor File format is shown in Table 44. The SFF can accommodate 20
speciation factors, 10 per line, and is formatted in a manner similar to the FSPLIT file. Each speciation
factor is expressed as an eight-digit integer which is read into the FREDS Speciation Module with all digits
to the right of the decimal. Seven digits are allowed for the species class codes in the file header to
accommodate particulate size class/SAROAD code combinations.
HYDROCARBON PREPROCESSING
Background
Another important application of speciation data for the NAPAP inventory is the "preprocessing" of reported
hydrocarbon emissions. While annual NEDS data report source-level emissions of Volatile Organic
Compounds (VOC), the VOC estimates are often calculated on different bases, making quantitative
comparisons of raw data prone to inaccuracy. For example, a source reporting "VOC" may or may not
include methane in its estimate. Also, depending on the method used to estimate emissions, the presence of
aldehydes may not be accounted for. Flame Ionization Detection (FED) is relatively insensitive to aldehydes;
sources reporting hydrocarbons calculated using FID can, therefore, be assumed to lack a mass corresponding
to the expected weight fraction of aldehydes.
To ensure internal consistency in handling organic compound emissions estimates, NAPAP processing software
includes a Hydrocarbon Preprocessor which accepts a single reported annual hydrocarbon value and derives
estimates of VOC and total hydrocarbons (THC) for each source in the inventory. The NAPAP Annual
Emissions Inventory is the output of the Hydrocarbon Preprocessor. By definition, VOC and THC differ only
by the expected mass of methane. Since methane is included in the speciation profiles used by FREDS, it is
the THC value which is used in subsequent NAPAP hydrocarbon speciation steps.
A89-380
56
-------
TABLE 4-4. PARTICULATE SPECIATION FACTOR FILE FORMAT
Record position
Column Variable
First Last width Format name Description
Header Record No. 1
1
7
7
I
CLASS 1
Pseudo-SAROAD code for TSP
Class
No. 1
6
14
7
I
CLASS 2
Pseudo-SAROAD code for TSP
Class
No. 2
15
21
7
I
CLASS3
Pseudo-SAROAD code for TSP
Class
No. 3
22
28
7
I
CLASS4
Pseudo-SAROAD code for TSP
Class
No. 4
29
35
7
I
CLASS5
Pseudo-SAROAD code ft* TSP
Class
No. 5
36
42
7
I
CLASS6
Pseudo-SAROAD code for TSP
Class
No. 6
43
49
7
I
CLASS7
Pseudo-SAROAD code for TSP
Class
No. 7
50
56
7
I
CLASS8
Pseudo-SAROAD code for TSP
Class
No. 8
57
63
7
I
CLASS9
Pseudo-SAROAD code for TSP
Class
No. 9
64
70
7
I
CLASS 10
Pseudo-SAROAD code for TSP
Class
No. 10
71
103
33
-
-
Unused
104
105
2
I
TYPE
Record Type (=11)
Header Record No. 2
1
7
7
I
CLASS 11
Pseudo-SAROAD code for TSP
Class
No. 11
8
14
7
I
CLASS 12
Pseudo-SAROAD code for TSP
Class
No. 12
15
21
7
I
CLASS13
Pseudo-SAROAD code for TSP
Class
No. 13
22
28
7
I
CLASS 14
Pseudo-SAROAD code for TSP
Class
No. 14
29
35
7
I
CLASS 15
Pseudo-SAROAD code for TSP
Class
No. 15
36
42
7
I
CLASS 16
Pseudo-SAROAD code for TSP
Class
No. 16
43
49
7
I
CLASS 17
Pseudo-SAROAD code to TSP
Class
No. 17
50
56
7
I
CLASS 18
Pseudo-SAROAD code for TSP
Class
No. 18
57
63
7
I
CLASS 19
Pseudo-SAROAD code for TSP
Class
No. 19
64
70
7
I
CLASS20
Pseudo-SAROAD code for TSP
Class
No. 20
71
103
33
-
-
Unused
104
105
2
I
TYPE
Record Type (=12)
Speciation Record No.
1
1
2
2
I
STATE
AEROS State Code
3
6
4
I
COUNTY
AEROS County Code
7
10
4
A
PLANT ID
Plant Identification Code*
11
13
3
A
POINT ID
Point Identification Code*
14
21
8
I
see
Area Source Category or Point Source SCC
22
29
8
F8V8"
TSF1
Speciation Factor for Class 1*
30
37
8
F8V8
TSF2
Speciation Factor for Class 2*
38
45
8
F8V8
TSF3
Speciation Factor for Class 3"
46
53
8
F8V8
TSF4
Speciation Factor to Class 4'
54
61
8
F8V8
TSF5
Speciation Factor for Class 5*
62
69
8
F8V8
TSF6
Speciation Factor for Class 6*
70
77
8
F8V8
TSF7
Speciation Factor to Class T
(continued)
A89-380
-------
TABLE 4-4 (continued)
Record position
Column Variable
First Last width Format name Description
78
85
8
F8V8
TSF8
Speciation Factor for Class 8"
86
94
8
F8V8
TSF9
Speciation Factor for Class 9t
95
102
8
F8V8
TSF10
Speciation Factor for Class 10"
103
103
1
-
-
Unused
104
105
2
I
TYPE
Record Type (=21)
Speciation Record No. 2
1
2
2
I
STATE
AEROS State Code
3
6
4
I
COUNTY
AEROS County Code
7
10
4
A
PLANT ID
Plant Identification Code*
11
13
3
A
POINT Id
Point Identification Code*
14
21
8
I
see
Area Source Category or Point Source SCC
22
29
8
F8V8"
TSF11
Speciation Factor for Class 11'
30
37
8
F8V8
TSF12
Speciation Factor for Class 12"
38
45
8
F8V8
TSF13
Speciation Factor for Class 13e
46
53
8
F8V8
TSF14
Speciation Factor for Class 14*
54
61
8
F8V8
TSF15
Speciation Factor for Class 15'
62
103
41
-
-
Unused
104
105
2
I
TYPE
Record Type (=22)
'For area source runs, these fields are left blank,
"Speciation factors are 8-digit integers which are read into FREDS with all 8-digits to the right of
the decimal point.
'Speciation factors are dimensionless weight fractions.
A89-380
58
-------
Hydrocarbon preprocessing is performed at the source category (SCC) level, as is speciation, based on the
mass fractions of formaldehyde and methane reported in the hydrocarbon species-profile file. Adjustment is
controlled by a set of formaldehyde and methane flags which specify whether or not augmentation is to be
performed for the SCC in question (1 = augment; 0 = do not augment). The flag settings depend on whether
or not the source type is expected to have accounted for aldehydes and/or methane in its original emissions
estimate.
An example of this methodology in practice is illustrated in Figure 4-1, on a hypothetical source which
reports hydrocarbon emissions of 100 tons. The source category corresponds to a profile containing
10 percent methane and 5 percent formaldehyde by weight An intermediate adjustment factor is calculated
using the following equation:
formaldehyde x formaldehyde\ + / methane x methane \
weight % flag / \weight % flag /
Adjustment factor = 100 (4 J)
The adjustment factor represents the total mass fraction of species unaccounted for in the input hydrocarbon
estimate. Actual THC can therefore be calculated as:
Actual THC - Input hydrocarbon .
Actual ihc — 1 . (Adjustment factor) (4J)
VOC is then calculated by removing the mass fraction of methane from THC:
Actual VOC = Actual THC x 100 ' (meU^e weiKht %> (4.4)
Figure 4-1 illustrates the four possible THC/VOC outcomes depending on the settings of the formaldehyde
and methane flags.
File Development and Quality Assurance
Profile-specific weight percents for formaldehyde and methane were extracted from the Radian hydrocarbon
speciation profiles discussed previously. Alliance conducted a review of these profiles and developed a set of
augmentation recommendations which formed the basis for the aldehyde and methane flag values (Battye,
1987). Several general rules and assumptions were used in making the recommendations list, which are
summarized below:
• Where a profile does not contain formaldehyde, augmentation for formaldehyde would have no
effect Therefore, a no-augment code generally was entered in these cases. The same rule was
used for profiles that do not contain methane.
A89-380
59
-------
r \
GIVEN:
Input hydrocarbon = 100 tons/year
Weight percent methane in profile = 10%
Weight percent formaldehyde in profile = 5%
GOVERNING EQUATIONS:
r/ formaldehyde\ / formaldehyde\"| r/ methanex x / methenex]
Adjustment factor = U we'9hl % ' 1 «'gg )J weiqht ->/J \ flag
[5 x (formaldehyde flag) ] + [10 x (methane flag)]
100
Ac|ua|THC= Input hydrocarbon
100 tons/year
1 - (adjustment factor) 1 - (adjustment factor)
Actual VOC = Actual THC x100" (meth^ weight %) _ Actua| THC x (0.9)
RESULTS:
Methane
flag
Aldehyde
flag
Adjustment
factor
Calculated emissions,
_ _tons/year_ _
THC VOC
0
0
0.00
100.0
90.0
0
1
0.05
105.3
94.7
1
0
0.10
111.1
100.0
1
1
0.15
117.6
105.9
AS8-471/A
Figure 4-1.
Example of VOC and THC calculation, FREDS Hydrocarbon Preprocessor.
60
-------
• For all point source combustion categories, VOC emissions were assumed to be based on flame
ionization detector measurements of total hydrocarbons, in which methane would not have been
broken out from other hydrocarbons and formaldehyde would not have been detected. A similar
assumption was made for VOC emissions from metallurgical processes. Thus, for these categories,
it was recommended that emissions be augmented for formaldehyde, but not for methane.
• For petrochemical vents containing formaldehyde, it was assumed that a detection technique was
used that is suited to formaldehyde. It was also assumed that the formaldehyde content of
petrochemical process streams was known, and hence that fugitive VOC emissions estimates include
formaldehyde where appropriate. It was assumed that VOC emissions estimates for petrochemical
processes did not include methane. Thus, it was recommended that petrochemical process vent
emissions and fugitive emissions be augmented for methane, but not for formaldehyde.
• For area source categories in general, emissions estimates were assumed to be derived from detailed
compound-specific measurements. Thus, it was assumed that VOC missions estimates for these
categories include formaldehyde and do not include methane. Augmentation was therefore
recommended for methane and not for formaldehyde.
Following EPA review, the flap were merged to the profile-specific weight pcrcents of formaldehyde and
methane to form the augmentation file used by the Hydrocarbon Preprocessor. The file is illustrated in Table
4-5.
Preliminary testing uncovered a problem with the flag settings for ihe profile representing residential natural
gas combustion, AP-42 listed both a methane and a non-methane hydrocarbon emission factor for this source
type (5.3 and 2.7 lb/million ft5, respectively), and a methane augmentation was recommended. The Radian
data, however, indicated that the profile was 100 percent methane. The resulting adjustment factor of 1.00
would cause a division-by-zero condition in equation 4.3, and could not be processed by NAPAP software.
The methane flag for this profile was therefore changed to zero prior to final inventory processing.
A89-380
61
-------
TABLE 4-5, HYDROCARBON AUGMENTATION FLAG FILE
Profile Methane Formaldehyde Formaldehyde Methane
number wt. pet. wt, pet. flag flag
0
7.36
1.55
0
0
1
11.00
42.00
1
0
2
0.00
48.70
1
0
3
56.00
8.00
1
0
4
7.60
7.60
1
0
5
82.80
0.00
1
0
7
70.00
30.00
1
0
8
11.60
0.00
1
0
9
11.60
0.00
1
0
11
45.30
0.00
1
0
12
15.80
0.00
1
0
13
73.30
0.00
1
0
14
0.00
0.00
1
0
16
11.10
0.00
1
0
23
0.00
0.00
1
0
24
21.30
0.00
1
0
25
56.00
8.00
1
0
26
15.70
0.00
1
0
29
36.00
51.00
1
0
31
2.90
0.00
1
0
35
0.00
0.00
1
0
39
13.30
0.00
1
0
47
0.00
0.00
1
0
51
20.00
20.00
1
0
66
0.00
0.00
0
0
68
0.00
0.00
0
0
72
0.00
0.00
0
0
76
0.00
0.00
0
0
78
0,00
0.00
0
0
79
0.00
1.70
1
0
85
0.00
0.00
0
0
87
0.00
0.00
0
0
88
0.00
0.00
0
0
89
0.00
0.00
0
0
90
0.00
0.00
0
0
100
0.00
0.00
0
0
121
0.00
0.00
1
0
122
80.40
0.00
1
0
127
0.00
0.00
0
0
166
63,00
0.00
0
0
182
0.00
0.00
0
0
183
0.00
0.00
0
0
195
100.00
0.00
1
0
197
0.00
0.60
0
0
202
98,70
0,00
1
0
203
70,00
0,00
0
1
217
40.90
0.00
1
0
219
0.00
0.00
0
0
220
0.00
0.00
0
0
221
0.00
0.00
0
0
222
0.00
0.00
0
0
(continued)
62
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt. pet, wt. pet. flag flag
223
0.00
0.00
0
0
225
0.00
0.00
0
0
226
0.00
0.00
0
0
227
0.00
0.00
0
0
228
0.00
0.00
0
0
229
0.00
0.00
0
0
230
0.00
0.00
0
0
271
0.00
0.00
0
0
272
0.00
0.00
0
0
273
0.00
0.00
0
0
274
0.00
0.00
0
0
275
0.00
0.00
0
0
277
0.00
0.00
0
0
282
0.00
0.00
0
0
288
0.00
0.00
0
0
289
0.00
0.00
0
0
290
0.00
0.00
0
0
291
0.00
0.00
0
0
292
0.00
0.00
0
0
296
6.20
0.00
0
0
297
8.80
0.00
0
0
299
0.00
0.00
0
0
301
0.00
0.00
0
0
304
0.00
0.00
0
0
305
2.60
0.00
0
0
307
9.82
0.00
1
1
316
28.60
0.00
1
0
321
3.30
0.00
1
0
332
17.10
21.80
0
0
333
37.66
0.00
0
0
1001
76.69
0.81
1
0
1002
22.40
0.00
1
0
1003
0.00
0.00
0
0
1004
0.00
0.00
0
0
1005
0.00
0.00
0
0
1006
0.00
0.00
0
0
1007
0.00
0.00
1
0
1008
0.00
0.00
0
0
1009
0.00
0.00
0
0
1010
46.31
0.00
0
0
1011
37.60
0.00
0
0
1012
61.30
0.00
0
0
1013
0.00
0.00
0
0
1014
0.00
0.00
0
0
1015
0.00
0.00
0
0
1016
0.00
0.00
0
0
1017
0.00
0.00
0
0
1018
0.00
0.00
0
0
1019
0.00
0.00
0
0
1020
0.00
0.00
0
0
1021
0.00
0.00
0
0
(continued)
63
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt. pet, wt. pet. flag flag
1022
0
00
0.00
0
0
1023
0
00
0.00
0
0
1024
0
00
0.00
0
0
1025
17
23
0.00
0
0
1026
0
00
0.00
0
0
1027
0
00
0.00
0
0
1028
0
00
0.00
0
0
1029
0
00
0.00
0
0
1030
4
28
0.00
0
0
1031
0
00
0.00
0
0
1032
0
00
0.00
0
0
1033
0
00
0.00
0
0
1034
0
00
0.00
0
0
1035
0
00
0.00
0
0
1036
0
00
0.00
1
0
1037
18
35
0.00
0
1
1038
3
73
0.00
0
1
1039
12
94
0.00
0
1
1040
0
00
0.00
0
1
1041
0
00
0.00
0
1
1042
0
00
0.00
0
1
1043
0
00
0.00
0
1
1044
0
00
0.00
0
1
1045
0
00
0.00
0
1
1046
4
88
0.00
0
1
1047
0
00
0.00
0
1
1048
0
00
0.00
0
1
1049
16
07
0.00
0
1
1050
0
00
0.00
0
1
1051
0
00
0.00
0
1
1052
0
00
0.00
0
1
1053
0
00
0.00
0
1
1054
0
00
0.00
0
1
1055
0
00
0.00
0
1
1056
0
00
0.00
0
1
1057
52
30
0.00
0
1
1058
24
38
0.00
0
1
1059
1
79
0.00
0
1
1060
0
00
0.00
0
1
1061
0
00
0.00
0
1
1062
0
00
0.00
0
1
1064
12
46
0.00
0
1
1065
0
51
0.00
0
1
1066
21
73
0.00
0
1
1067
34
72
0.00
0
1
1068
0
00
0.00
0
1
1069
0
00
0.00
0
1
1070
86
71
0.00
0
0
1071
0
00
0.00
0
0
1072
0
00
0.00
0
1
1073
0
00
0.00
0
1
(continued)
64
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt, pet. wt, pet. flag flag
1074
0.00
0.00
0
1
1075
0.00
0.00
0
1
1076
0.00
0.00
0
1
1077
0.00
0.00
0
1
1078
0.00
0.00
0
1
1079
0.00
0.00
0
1
1080
0.00
0.00
0
1
1081
0.00
0.00
0
1
1082
0.00
0.00
0
1
1083
0.00
0.00
0
1
1084
38.39
0.00
1
0
1085
20.00
0.00
1
0
1086
0.00
0.00
0
0
1087
0.00
0.00
0
1
1088
0.00
0.00
0
0
1089
0.00
0.70
1
0
1090
0.00
0.00
0
1
1091
0.00
0.00
0
1
1092
0.00
0.00
0
1
1093
0.00
0.00
0
1
1094
0.00
0.00
0
1
1095
0.00
0.00
0
1
1096
0.00
0.00
0
1
1097
9.38
15.49
1
0
1098
9.57
15.01
1
0
1099
10.95
14.14
1
0
1100
0.00
0.00
0
0
1101
10.07
0.74
1
1
1103
0.00
0.00
0
0
1104
0.00
0.00
0
0
1105
0.00
0.00
0
0
1106
0.00
0.00
0
0
1107
0.00
0.00
0
0
1108
0.00
0.00
0
0
1109
0.00
0.00
0
0
1110
0.00
0.00
0
0
1111
0.00
0.00
0
0
1112
0.00
0.00
0
0
1114
0.00
0.00
0
0
1115
0.00
0.00
0
0
1116
0.00
0.00
0
0
1118
0.00
0.00
0
0
1119
0.00
0.00
0
0
1120
0.00
0.00
0
0
1121
0.00
0.00
0
0
1122
0.00
0.00
0
0
1123
0.00
0.00
0
0
1124
0.00
0.00
0
0
1125
0.00
0.00
0
0
1126
0.00
0.00
0
0
1127
0.00
0.00
0
0
(continued)
65
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt, pet, wt. pet. flag flag
1128
0
00
0.00
0
0
1129
0
00
0.00
0
0
1130
0
00
0.00
0
0
1131
0
00
0.00
0
0
1132
0
00
0.00
0
0
1134
0
00
0.00
0
0
1135
0
00
0.00
0
0
1136
0
00
0.00
0
0
1137
0
00
0.00
0
0
1138
0
00
0.00
0
0
1139
0
00
0.00
0
0
1140
0
00
100.00
0
0
1141
0
00
0.00
0
0
1142
0
00
0.00
0
0
1144
0
00
0.00
0
0
1145
0
00
0.00
0
0
1146
0
00
0.00
0
0
1147
0
00
0.00
0
0
1148
0
00
0.00
0
0
1149
0
00
0.00
0
0
1150
0
00
0.00
0
0
1151
0
00
0.00
0
0
1152
0
00
0.00
0
0
1153
0
00
0.00
0
0
1154
0
00
0.00
0
0
1155
0
00
0.00
0
0
1158
0
00
0.00
0
0
1159
0
00
0.00
0
0
1160
0
00
0.00
0
0
1162
0
00
0.00
0
0
1163
0
00
0.00
0
0
1164
0
00
0.00
0
0
1165
0
00
0.00
0
0
1166
0
00
0.00
0
0
1167
0
00
0.66
1
1
1168
0
00
0.00
0
0
1171
0
00
0.00
0
0
1172
0
00
0.00
0
0
1173
0
00
0.00
0
0
1174
0
00
0.00
0
0
1175
0
00
0.00
0
0
1176
0
00
0.00
0
0
1178
0
00
0.00
1
0
1185
0
00
0.00
1
0
1186
2
45
0.00
1
1
1187
0
00
0.00
0
0
1188
0
00
0.00
0
0
1189
0
00
0.00
1
0
1190
0
00
0.00
0
0
1191
0
00
0.00
0
0
1192
0
00
0.00
0
0
(continued)
66
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt, pet. vt. pet. flag flag
1193
0.00
0.00
0
0
1194
0.00
0.00
0
0
1195
0.00
0.00
0
0
1196
0.00
0.00
0
0
1197
0.00
0.00
0
0
1198
0.00
0.00
0
0
1199
0.00
0.00
0
0
1200
0.00
0.00
0
0
1201
4.40
12.20
0
1
1202
0.00
0,00
0
0
1203
10.97
1.42
1
1
1204
0.04
0.00
1
1
9001
25.35
15.19
1
0
9002
42.45
7.70
1
0
9003
9.02
1.82
0
1
9004
5.07
0.03
0
1
9005
0.00
0.00
0
1
9006
0.00
0.00
0
1
9007
43.35
0.00
0
1
9008
0.00
0.00
1
0
9009
29.10
0.00
1
0
9010
0.00
0.35
1
0
9011
18.60
1.60
1
0
9012
13.01
8.88
1
0
9013
0.00
0.00
1
0
9014
0.00
0.00
0
1
9015
53.81
0.00
1
0
9016
0.00
0.00
0
0
9017
0.00
0.00
0
0
9021
0.00
0.00
0
0
9022
59.71
0.00
1
0
9023
0.00
0.00
0
0
9024
2.01
0.00
0
0
9025
3.02
0.00
0
0
9026
10.01
2.73
0
0
9027
1.30
0.00
0
0
9028
0.00
1.47
0
0
9029
0.00
0.00
0
0
9030
0.00
0.00
0
0
9031
0.00
0.00
0
0
9032
0.00
0.00
0
0
9033
0.00
0.00
0
0
9034
0.00
0.00
0
0
9035
0.00
0.00
0
0
9036
0.00
0.00
0
0
9037
0.00
0.00
0
0
9038
0.00
0.00
0
0
9039
0.00
0.00
0
0
9040
0.00
0.00
0
0
9041
0.00
25.00
0
0
(continued)
67
-------
TABLE 4-5 (continued)
Profile Methane Formaldehyde Formaldehyde Methane
number wt. pet. wt. pet. flag flag
9042 0.00 0.00 0 0
9043 0.00 0.00 0 0
9044 0.00 0.00 0 0
9046 0.00 0.00 0 0
9047 0.00 0.00 0 0
68
-------
SECTION 5
REFERENCES
GENERAL REFERENCES
Demmy, J.L., W.M. Tax, and T.E. Warn. Area Source Documentation for the 1985 National Acid
Precipitation Assessment Program Inventory. EPA-600/8-88-106 (NTIS PB89-151427), U.S. Environmental
Protection Agency, Research Triangle Park, NC. December 1988.
Lebowitz, L.G., and A.S. Ackerman. Flexible Regional Emissions Data System (FREDS) Documentation for
the 1980 NAPAP Emissions Inventory. EPA-600/7-87-Q25a (NTIS PB88-129499). U.S. Environmental
Protection Agency, Research Triangle Park, NC. November 1987.
Middleton, P. Analysis of Emissions Databases for Regional Models. Atmos. Environ. 21:1497-1509. 1987.
Modica (Lebowitz), L.G., D.R, Dulleba, R.A. Walters, and J.E. Langstaff. Flexible Regional Emissions Data
System (FREDS) Documentation for the 1985 NAPAP Emissions Inventory. EPA-600/9-89-047 (NTIS
PB89-198816). U.S. Environmental Protection Agency, Research Triangle Park, NC. May 1989.
Sellars, F.M., T.E. Fitzgerald, Jr., J.M. Lennon, LJ. Maiocco, N.M. Monzione, and D.R. Neal, Jr. National
Acid Precipitation Assessment Program Emission Inventory Allocation Factors, EPA-600/7-85-035 (NTIS
PB86-104247), U.S. Environmental Protection Agency, Research Triangle Park, NC. September 1985.
U.S. Environmental Protection Agency. Compilation of Air Pollutant Emission Factors, Volume t Stationary
Point and Area Sources. APA2. Fourth Edition. (GPO No. 055-000-00251-7), U.S. Environmental
Protection Agency, Research Triangle Park, NC. September 1985.
Wagner, J. K., R. A. Walters, L. J. Maiocco, and D. R. Neal. Development of the 1980 NAPAP Emissions
Inventory. EPA-600/7-86-057a (NTIS PB88-132121), U.S. Environmental Protection Agency, Research
Triangle Park, NC. December 1986.
Zimmerman, D., W. Tax, M. Smith, J. Demmy, and R. Battye. Anthropogenic Emissions Data for the 1985
NAPAP Inventory. EPA-600/7-88-022 (NTIS PB89-151419), U.S. Environmental Protection Agency, Air and
Energy Engineering Research Laboratory, Research Triangle Park, NC. November 1988.
TEMPORAL FACTOR REFERENCES
Electric Power Research Institute. The EPRI Regional Systems. EPRI P-1950-SR. Mo Alto, CA. 1981.
Engineering-Science. Emissions Inventories for Urban Airshed Model Application in the Philadelphia AQCR.
Prepared for the U.S. Environmental Protection Agency. EPA-450/4-82-005 (NTIS PB83-187823). Research
Triangle Park, NC. April 1982.
Klemm, H. A. and R. J. Brennan. Emissions Inventory for the SURE Region. Alliance Technologies
Corporation for the Electric Power Research Institute, EA-1913. Palo Alto, CA. 1981.
A89-380
69
-------
Sellars, F., et al. Northeast Corridor Regional Modeling Project Annual Emission Inventory Compilation and
Formatting. GCA Technology Division for the U.S. Environmental Protection Agency. EPA-450/4-82-013a-r.
Research Triangle Park, NC. 1982.
U.S. Department of Commerce. Local Climatological Data Published Stations (1980 data), National Climatic
Data Center, National Oceanic and Atmospheric Administration. Asheville, NC. 19®).
U.S. Department of Commerce. State, Regional, and National Monthly and Annual Temperatures Weighted
by Area. Historical Climatology Series 4-1. National Oceanic and Atmospheric Administration. Asheville,
NC. 1983a.
U.S. Department of Commerce. State, Regional, and National Monthly and Seasonal Heating Degree Days
Weighted by Population (1980 Census). Historical Climatology Series 5-1. National Oceanic and
Atmospheric Administration. Asheville, NC. 1983b.
U.S. Department of Energy. Energy Data Reports, Washington, D.C., 1979.
U.S. Department of Energy. Petroleum Marketing Monthly. Energy Information Service. Washington, D.C.
1984.
U.S. Environmental Protection Agency. 1979a. Documentation of the Regional Air Pollution Study (RAPS)
and Related Investigations in the St. Louis Air Quality Control Region. EPA-600/4-79-076 (NTIS
PB80-138241). Research Triangle Park, NC. 1979.
U.S. Environmental Protection Agency, 1979b. Procedures for the Preparation of Emission Inventories for
Volatile Organic Compounds; Volume II-Emission Inventory Requirements for Photochemical Air Quality
Simulation Models. EPA-450/4-79-018 (NTIS PB80-202229), Research Triangle Park, NC. 1979.
U.S. Environmental Protection Agency. Emission Inventories for Urban Airshed Model Application in Tulsa,
Oklahoma. EPA-450/4-80-021 (NTIS PB81-156986). Research Triangle Park, NC. 1980.
U.S. Environmental Protection Agency. Technical Tables to the National Air Pollutant Emissions Estimates,
1940-1984. EPA-45G/4-85-014 (NTIS PB86-121100). Office of Air Quality Planning and Standards,
Research Triangle Park, NC. January 1986.
U.S. Department of Labor. Supplement of Employment and Earninp: Revised Establishment Data, Bureau
of Labor Statistics, Washington, D.C. 1981.
U.S. Department of Transportation, Federal Highway Administration. Highway Statistics. 1980 Annual.
SPATIAL FACTOR REFERENCES
Beaulieu, T.A., and L.G. Modica (Lebowitz). Documentation of Spatial Allocation Factor Procedures far the
1980 NAPAP Emissions Inventory. EPA-600/7-88-024a (NTIS PB89-159479), U.S. Environmental Protection
Agency, Research Triangle Park, NC. December 1988,
U.S. Bureau of the Census, Census of Population and Housing, 1980: Master Area Reference File (MARF)2
[machine-readable data file]/prepared by the Bureau of Census. Washington: The Bureau [producer and
distributer]. 1983.
U.S. Bureau of the Census, Census of Population and Housing, 1980: Summary Tape File 3 Technical
Documentation. Prepared by the Data User Services Division, Bureau of the Census. Washington: The
Bureau, pp. 437. 1982.
A89-380
70
-------
SPECIATION FACTOR REFERENCES
Compilation of Air Pollutant Emission Factors, Volume I: Stationary Point and Area Sources. AP-42.
Fourth Edition (GPO No. 055-000-00251-7), U.S. Environmental Protection Agency, Research Triangle Park,
NC, September 1985.
Core, JJE., JJ. Shah, J.A. Cooper. Receptor Model Source Composition Library. Nero & Associates aid
NEA, Inc. for the U.S. Environmental Protection Agency, EPA 450/4-85-002 (NT1S PB85-228823).
November 1984.
Shareef, G., W. Butler, L. Bravo, and M. Stockton. Air Emissions Species Manual Volume I - Volatile
Organic Compound (VOC) Species Profiles. EPA-450/2-88-003a (NTIS PB88-225792), U.S. Environmental
Protection Agency, Research Triangle Park, NC. April 1988.
Shareef, G„ and L. Bravo. Air Emissions Species Manual Volume II - Particulate Matter (PM) Species
Profiles. EPA-450/2-88-003b (NTIS PB88-225800). U.S. Environmental Protection Agency, Research
Triangle Park, NC. April 1988.
Volatile Organic Compound (VOC) Species Data Manual - Second Edition. Report No. EPA-450/4-80-015
(NTIS PB81-119455), U.S. Environmental Protection Agency, Research Triangle Park, NC. July 1980.
A89-380
71
-------
APPENDIX A
ADJUSTMENTS TO SPATIAL FRACTIONS FOR
1985 NAPAP PROCESSING
A-l
-------
APPENDIX A
ADJUSTMENTS TO SPATIAL FRACTIONS FOR 1985 NAPAP PROCESSING
Based on quality control procedures, adjustments were made to the existing spatial fractions for use with the
1985 NAPAP inventory. Corrections applied to the spatial fraction file are discussed below.
MISSING COUNTIES
Census-based spatial allocation factors (SAFs), i.e., population and housing, for the six Virginia
counties/independent cities (Nansemond is the only Virginia county equivalent which is not an independent
city) and Barbour, WV were available from another project. These, data were input to the spatial fraction
file. Since population is used as the default surrogate for NAPAP processing, these data are sufficient to
spatially allocate emissions from these counties and independent cities.
To generate spatial fractions for Cibola, New Mexico, fractions developed for Valencia county were adjusted
to account for the new county boundary, since Cibola was part of Valencia in 1980. Using the total
population, housing, and land area fractions for Valencia along with the corresponding spatial fractions,
population, housing and land area totals by grid cell were determined. These surrogates were then
reaggregated separately for Cibola and Valencia and new population, housing, and land area fractions were
calculated for each county. The remainder of the spatial category fields were assigned values of zero.
The remaining two missing counties, Sabine and Menominee, required a more detailed effort to generate
gridded spatial fractions. Since population fractions could not be estimated for Sabine and Menominee, land
area fractions were calculated. Since population is the default surrogate for NAPAP, these spatial fractions
were inserted into the population field of the spatial fraction file.
MISSING GRID CELLS
For spatial allocation of emissions to missing grid cells for 1985 NAPAP processing, gridded population and
housing fractions were calculated. Population and housing fractions for Queen Annes, MD for the two
missing grid cells were available from a previous project. These values summed exactly to the fraction
missing. Alexander, IL had only one missing grid cell, therefore, the calculated missing fractions for
population and housing were assiped to that grid cell.
A89-380
A-2
-------
In Monroe, FL, and Dare, NC, the calculated missing fractions required allocation over several grid cells. To
determine, as accurately as possible, the population distribution over these grid cells, population totals for each
grid cell were determined using available town populations in each county. The grid cell coordinates for each
town were calculated from the latitude and longitude of each town and the fractional population of each
missing grid cell was estimated.
The population fraction for each town/grid was calculated by multiplying the fractional population of each
missing grid cell by the missing population fraction from the SAFP processing messages. Spatial fractions for
the same grid cell were combined. For Monroe County, SAFs for grid cells outside the NAPAP grid were
kept such that the county's emissions could be correctly allocated.
COUNTY CODE ASSIGNMENTS
Incorrect county code assignments in the Saginaw County spatial data resulted from the presence of two
different county codes in the land use file: 4600 and 4780. In the FIPAEROS file (and hence, census spatial
file), Saginaw is coded as county 4600. In the emissions file, however, Saginaw is coded as 4780. This was
further verified using the AEROS Manual (U.S. EPA, 1986). To assure correct spatial allocation of
Saginaw's emissions, records with a county code of 4780 were deleted and records with a county code of
4600 were recoded to 4780 since these contained the complete census and land use fractions for each grid
cell in Saginaw. The resultant spatial fractions are compatible with the area source emissions file.
For the Oregon and South Dakota counties, population and housing fractions were dropped in SPACEMERGE
due to incorrect county code assignments. In Oregon, the county code assignments for Lincoln and Linn
were switched in the land use county-to-grid file such that grid cells belonging to Lincoln were assigned to
Linn, and vice-versa. As a result, when the census spatial surrogates were matched to the land use file, the
census data was dropped due to no matching grids. Since the census data was dropped, population and
housing values could not be retrieved. Thus, when the county codes for Lincoln and Linn were interchanged
to the correct codes, the census fields still contained zero values. Since population is used as a default
surrogate for NAPAP, values of land area were inserted into the population field to serve as the default
surrogate for these counties.
In the census data, Jackson and Washabaugh counties were treated as a single entity (coded as Jackson -
0880). In the land use and emissions files, however, these were treated as separate counties. Thus, when the
land use and census surrogates were merged, the census grids corresponding to Washabaugh county were
dropped. The census-based spatial surrogates for Jackson are relative to the total of both counties and are,
therefore, incorrect Since the information necessary to correct the factors was not available, population
values in both counties were substituted by land area fractions; housing fractions were set to zero.
A89-380
A-3
-------
LOCATION DATA
Spatial factor problems due lo incorrect location data in the census files were noted for two Virginia counties
(Pulaski and Rappahannock) and Adams, WI. In Pulaski, an independent city (Manassas) was assigned the
NEDS code for this county. Manassas, however, is located in a different county (Prince William), in a
different part of the State. The effect of this on Pulaski is census spatial surrogates off by more than
20 percent To correct this, the grids located in Pulaski will be normalized by SAFP, thereby removing
Manassas from the Pulaski county total.
In Rappahannock, latitude and longitude data for a subcounty census record corresponds to a grid cell outside
the county border. As a result, the population and housing values for this grid cell were dropped in
SPACEMERGE and the resultant county total for these surrogates is off by approximately 20 percent. To
correct this, Rappahannock grid cells with zero population and housing values were identified. Two such
cases existed, however, only one of the grids contained a town. The missing population and housing fractions
were assigned to that grid cell.
Latitude and longitude data contained in the census data county and subcounty records for Adams, WI do not
correspond to grid cells in the county. As a result, when the land use and census surrogates were merged,
the census data was dropped by SPACEMERGE. The missing census population and housing fractions cannot
be retrieved, however, land use fractions were available and, therefore, land area fractions were substituted in
the population field to act as the default surrogate.
MASSACHUSETTS
A problem with the assignment of Massachusetts grid cells to Air Pollution Control Districts (APCDs) was
discovered for categories spatially allocated by land use surrogates. Examination of the data files and
software at various points in the spatial development process revealed a problem with the resolution of the
spatial factors relative to the resolution of the reported emissions for Massachusetts. The land use spatial
factors were developed relative to NAPAP county/grid cell information; i.e„ the fraction of each county's land
use contained within each grid cell. Massachusetts' emissions data are typically reported by APCD. When
converting from county to APCDs, the land use fractions were not correctly transformed from county/grid cell
resolution to APCD/grid cell resolution.
To correct this problem, a new set of land area spatial factors was developed for Massachusetts. The
fractions were combined to form all unique APCD, column, and row combinations necessary for input to the
Spatial Allocation Factor Preprocessor (SAFP). To incorporate the newly developed Massachusetts land area
fractions
A89-380
A-4
-------
into the Spatial Fraction File, the old Massachusetts land area fractions were set to zero, and the new
fractions merged into the file. Because information was not readily available to accurately correct the
remaining land use surrogates, all other land use surrogates to be utilized by FREDS were set equal to the
new land area fractions.
ALGORITHMS
Problems discovered with algorithms in CREATE5A, SPACEMERGE, and SAFP are summarized in
Table 3-3. Minor problems with algorithms include: calculating column and row numbers for border grids
for a subset of the NAPAP grid (CREATE5A), calculating column and row from grid number
(SPACEMERGE), and normalization of county-level factors not summing to 100 percent (SAFP). The
necessary corrections and changes to the spatial processing software were applied for future spatial factor
processing.
NEW AREA SOURCE CATEGORIES
For 1985 NAPAP processing, ten new area source categories were created. Spatial allocation of emissions
from these categories was accomplished by assigning surrogates to each category based on the existing 14
NAPAP surrogates. For each category, a surrogate was assigned based on activity levels used for
characterization of emissions (Demmy, et al., 1987) and engineering judgment. The SCCs and surrogates
corresponding to each category were included in the surrogate selection file prior to execution of SAFP. The
assigned surrogates for each SCC are given in Table 3-2.
REFERENCES
U.S. Environmental Protection Agency. AEROS Manual Series, Volume V: AEROS Manual of Codes (3rd
ed.). EPA-450/2-76-0Q5b. Research Triangle Park, NC. April 1986.
Demmy, J.L., W.M. Tax, and T.E. Warn. Area Source Documentation for the 1985 National Acid
Precipitation Assessment Program Inventory. EPA-600/8-88-106 (NTIS PB89-151427). U.S. Environmental
Protection Agency, Research Triangle Park, NC. December 1988.
A89-380
A-5
-------
APPENDIX B
DEVELOPMENT OF ALLOCATION FACTORS FOR
CANADIAN ANTHROPOGENIC SOURCES
A89-380
B~1
-------
APPENDIX B
DEVELOPMENT OF ALLOCATION FACTORS FOR
CANADIAN ANTHROPOGENIC SOURCES
INTRODUCTION
Environment Canada, provincial air pollution control agencies, and the U.S. Environmental Protection Agency
worked cooperatively to develop an annual inventory of Canadian emissions which would be consistent and
compatible with the U.S. inventory. The Canadian data include emissions for 729 plants and 154 area source
categories for 10 Provinces. The data are derived from a national inventory compiled by Environment
Canada from data collected by provincial air pollution control agencies. Because the NAPAP geographic
domain extends only to 60° north latitude, emissions data from the Yukon and Northwest Territories were
reported to the U.S. EPA but were not processed for use by NAPAP. In addition, to ensure consistency
between the two national inventories, only the 129 area source categories having analogs in the U.S. data
were retained in the NAPAP Canadian file.
In order to allocate Canadian emissions data in a manner consistent with the allocation of U.S. anthropogenic
emissions. Environment Canada also supplied data on spatial and temporal emissions patterns. Point source
annual emissions were assigned to U.S. SCCS to allow speciation using the methodologies developed for the
U.S. NAPAP inventories. The remainder of this section describes the allocation of Canadian point and area
source inventories with particular emphasis on areas in which U.S. and Canadian methodologies differ.
TEMPORAL ALLOCATION FACTORS
Canadian emissions data files were accompanied by separate files containing temporal fractions for point and
area sources. Temporal files contained throughput percentages for months of the year, days of the week, and
hours of the day, respectively. Identifiers in these files permitted the merging of individual temporal
scenarios into profiles compatible with sources in the emissions data files.
Separate temporal factor files were developed for point and area sources. For point sources, most factors
were relative to the plant level. If data were collected at the point level, the profile for the point having the
greatest emissions was used to allocate emissions for the entire plant. Area source factors were developed at
the source category level. For both point and area sources, uniform default profiles were substituted when
A89-380
B-2
-------
temporal data were not available (DesLauriers, M. personal communication with D. Fratt, Alliance, March,
1989),
For FREDS processing, the temporal factors were reformatted to a form similar to the factors used for the
U.S. NAPAP inventories. This required converting monthly throughputs to seasonal factors and weekday
(Monday through Friday) fractions to an average value representing a typical weekday. File formats (Tables
B-l and B-2) are similar to those for U.S. point and area sources (Tables 2-2, 2-3), except that a record
identifier (YWD) is required to link temporal profile records with emissions data. Twelve temporal scenarios
are represented in each profile; one each for a typical weekday, Saturday and Sunday in each season (See
Table 2-4). All files are stored as SAS data sets.
SPATIAL ALLOCATION FACTORS
Spatial allocation surrogates provided by Environment Canada included population, oil homes, gas homes, total
homes, industrial labor force, commercial labor force, agricultural labor force and mining labor force. These
data files were supplied to Environment Canada by Statistics Canada, and represent the 1981 base year.
Census data for 1986 was also provided by Statistics Canada, but only data for population and total homes
were available. To calculate spatial fractions, the census data were summed for each Province to obtain
provincial totals, and the value for each grid cell was divided by the provincial total. Once the fractions were
calculated, each of the Canadian area source categories was matched to the most appropriate surrogate
category based on an analysis of the 1980 allocation factors, creating the Canadian surrogate selection file. A
listing of Canadian area source categories and the assigned spatial surrogates are presented in Table B-3.
For final processing, the Spatial Allocation Factor Preprocessor (SAFP), used for processing the U.S. area
source data, was modified to process the Canadian area source data. SAFP matches a spatial fraction for
each grid cell to each area source category based on the surrogate selection file for later processing into a
gridded, resolved inventory by FREDS. As with the U.S. data, the population surrogate was used as the
default category for Canadian area source categories which either lacked a specific allocation surrogate or for
which the assigned surrogate equaled zero when summed at the Province level. The file format for the
EBCDIC Spatial Allocation Factor File input to FREDS is presented in Table B-4.
SPECIATION FACTORS
Canadian point and area source data were speciated using the data developed by Radian Corporation (Shareef,
et, al., 1988), The Canadian point source data submittals included several SCC process classifications which
had not been used in the U.S. data. The data processing for input into acid deposition models requires that a
TSP and THC speciation profile be associated with every SCC included in the file. Therefore, Environment
Canada requested that EPA make appropriate speciation profile assignments for those SCCs. In some cases.
A89-380
B-3
-------
TABLE B-1. FORMAT OF SAS DATA SET CONTAINING TEMPORAL ALLOCATION
FACTORS FOR 1985 CANADIAN AREA SOURCE EMISSIONS DATA
Variable name
SAS variable length
(bytes)
Description
NUM_DAY
8
Temporal Scenario Code
YWD
8
Profile Identifier
SEA
8
Seasonal Temporal Allocation Factor
DAY
8
Daily Temporal Allocation Factor
HOUR1-HOUR24
8
Hourly Temporal Allocation Factors
A89-380
B-4
-------
TABLE B-2, FORMAT OF SAS DATA SET CONTAINING TEMPORAL ALLOCATION
FACTORS FOR 1985 CANADIAN POINT SOURCE EMISSIONS DATA
Variable name
SAS variable length
(bytes)
Description
NUM_DAY
8
Temporal Scenario Code
YWD
8
Profile Identifier
SEA1-SEA12
8
Seasonal Temporal Allocation Factors
DAY 1-DAY 12
8
Daily Temporal Allocation Factors
HOUR 1-HOUR288
8
Hourly Temporal Allocation Factors
A89-380
-------
iO
11000
11200
11300
11400
11510
.11520
11610
11620
11630
11640
12100
12200
12300
12400
12510
12520
13100
13200
13300
13400
13510
13520
14100
14200
14300
14400
21000
21100
21200
21300
21400
22100
22140
22210
22220
22230
22240
22250
22260
22270
23110
23120
23130
23140
23150
24100
TABLE B-3. SPATIAL ALLOCATION FACTOR SURROGATES FOR 1985
NAPAP CANADIAN AREA SOURCE EMISSIONS CATEGORIES
Surrogate Surrogate
ID Indicator
Emissions Category Description
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
4
Housing
6
Conmercial
Labor
Force
6
Commercial
Labor
Force
6
Conmercial
Labor
Force
6
Conmercial
Labor
Force
6
Conmercial
Labor
Force
6
Commercial
Labor
Force
5
Industrial
Labor
Force
5
Industrial
Labor
Force
5
Industrial
Labor
Force
5
Industrial
Labor
Force
5
Industrial
Labor
Force
5
Industrial
Labor
Force
1
Population
i
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
1
Population
Stationary Fuel Combustion - Residential Natural Gas
Contiustion
Combustion
Combustion
Contiustion
Combustion
Contustion
Combustion
Combustion
Conbustion
Combustion
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Stationary Fue
Hisc Power Generation
Hisc Power Generation
Residential Natural Gas liquids
Residential Kerosene and Stove Oils
Residential Coal
Residential Distillate Oil
Residential Residual Oil
Residential Conventional Wood Stove
Residential Slow Conixjstion Stove
Residential Fireplace
Residential Uood Furnace
Conmercial/Inst Natural Gas
Confeustion - Conmercial/Inst Liquid Petroleum Gas
Combustion - Conmercial/Inst Kerosene Stove Oil
Combustion - Conmercial/Inst Coal
Combustion - Conmercial/Inst Distillate Oil
Combustion - Commercial/lost Residual Oil
Combustion - Industrial Natural Gas
Combustion - Industrial Natural Gas Liquids
Combustion - Industrial Kerosene and Stove Oils
Combustion - Industrial Coal
Conbustion - Industrial Distillate Oil
Combustion - Industrial Residual Oil
Natural Gas
Heavy Fuel Oil
Hisc Power Generation - Light Fuel Oil
Hisc Power Generation - Hisc Diesel Generator
Gasoline Vehicles - Automobiles
Gasoline Vehicles - Light Duty Trucks
Gasoline Vehicles - Heavy Duty Trucks
Diesel Vehicles - Light Duty Vehicles
Diesel Vehicles * Heavy Duty Vehicles
Off-Road Gasoline Vehicles - General
Off-Road Gasoline Vehicles - Snowmobiles
Off-Road Diesel Vehicles - Agricultural
Off-Road Diesel Vehicles - Construction
Off-Road Diesel Vehicles - Mining
Off-Road Diesel Vehicles - Manufacturing
Off-Road Diesel Vehicles - Forestry
Off-Road Diesel Vehicles - PUslic Administration
Off-Road Diesel Vehicles - Pipeline
Off-Highway Mobile Sources - Jet Aircraft
Off-Highway Mobile Sources - Turboprop Aircraft
Off-Highway Mobile Sources - Piston Engine Aircraft
Off-Highway Mobile Sources - Helicopter
Off-Highway Mobile Sources - Small Piston Aircraft
Off-Highway Mobile Sources - Railroad Diesel Oil
(continued)
B-6
-------
TABLE B-3. (Continued)
Category Surrogate Surrogate
ID ID Indicator Emissions Category Description
25110
1
Population
Off-Highway Mobile Sources - Motorships Dockside
25120
1
Population
Off-Highway Mobile Sources - Steamships Docksicte
25210
1
Population
Off-Highway Mobile Sources - Motorships Underwat
25220
Population
Off-Highway Mobile Sources - Steamships Underway
25310
1
Population
Off-Highway Mobile Sources - Gasoline Outboards
27110
1
Population
Gasoline & Diesel Mkting - Refining Storage Transfer
27111
1
Population
Gasoline & Diesel Mkting - Filling Vehicle Tanks
27120
1
Population
Gasoline & Diesel Mkting - Diesel Evaporation
27130
1
Population
Gasoline 4 Diesel Mkting - Station Storage Transfer
27140
1
Population
Gasoline & Diesel Mkting - Vapor Loss at Station (Tar*)
27150
1
Population
Gasoline & Diesel Mkting - Transfer to Cars
27160
1
Population
Gasoline & Diesel Mkting - Spillage at Station
31300
1
Population
Solid Waste Incineration - Multiple Chamber
31500
1
Population
Solid Waste Incineration - Controlled Air
32100
8
Mining Labor Force
Solid Waste Incineration - Wood Waste Disposal
33100
7
Agricultural Labor Force
Slash Burning
47100
8
Mining Labor Force
Forest Fires
47200
4
Housing
Structural Fires
51000
6
Comnercial Labor Force
Bakeries
52110
8
Mining Labor Force
Crude Oil Production - Evaporation of Hydrocarbons
53110
5
Industrial Labor Force
Clay Products Manufacturing - Dryer/Grinder
53120
5
Industrial Labor Force
Clay Products Manufacturing - Storage
53200
8
Mining Labor Force
Coal Industry - Coal Mining
53210
8
Mining Labor Force
Coal Industry - Coal Handling
53220
8
Mining Labor Force
Coal Industry - Overburden Removal (Fugitive)
53300
8
Mining Labor Force
Coal Industry - Transportation
53400
8
Mining Labor Force
Mining & Rock Quarrying
53410
8
Mining Labor Force
Mining & Rock Quarrying - Open Pit Mining
53420
8
Mining Labor Force
Mining & Rock Quarrying - Overburden Removal
53430
8
Mining Labor Force
Mining & Rock Quarrying - Underground Mining
53440
8
Mining Labor Force
Mining & Rock Quarrying - Concentrate Dryers
53450
8
Mining Labor Force
Mining & Rock Quarrying - Concentrate Transport
53500
5
Industrial Labor Force
Sand & Gravel Processing - General
53611
5
Industrial Labor Force
Stone Processing - Crushed Stone - Primary Crushing
53612
5
Industrial Labor Force
Stone Processing - Crushed Stone - Secondary Crushing
53613
5
Industrial Labor Force
Stone Processing - Crushed Stone - Conveying and Handling
53614
5
Industrial Labor Force
Stone Processing - Crushed Stone - Screening
53615
5
Industrial Labor Force
Stone Processing - Crushed Stone - Secondary Crushing (Fugitive)
53621
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Primary Crushing
53622
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Secondary Crushing
53624
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Screening
53625
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Fines Mill
53626
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Storage Pile Losses
53627
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Secondary Crushing (Fugitive)
53628
5
Industrial Labor Force
Stone Processing - Pulverized Stone - Recrushing
53631
5
Industrial Labor Force
Stone Processing - Building Stone - Cutting
{continued)
B-7
-------
TABLE B-3. (Continued)
Category Surrogate Surrogate
ID ID Indicator Emissions Category Description
54100 1 Population Asphalt Production - Drying
54200 1 Population Asphalt Production - Fugitive
55100 1 Population Concrete Batching
55200 1 Population Concrete Batching • Fugitive Material Handling
56000 5 Industrial Labor Force Plastic Fabrication
57100 8 Mining Labor Force Wood Industry - Savfiti 11 Production
57200 8 Mining Labor Force Wood Industry - Plywood and Veneer Production
57300 8 Mining Labor Force Wood Industry - Pulpboard Production
57400 8 Mining Labor Force Wood Industry - Hardwood Production
61110 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Shipping & Receiving
61120 7 Agricultural Labor Force ' Grain Milling/Handling - Terminal Elev - Transfer & Conveying
61130 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Cleaning
61140 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Drying
61150 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Headhouse
61160 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Tripper (Gallery Belt)
61210 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Shipping & Receiving
61220 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Transfer & Conveying
61230 7 Agricultural Labor Force Grain Mi Iling/Handling - Terminal Elev - Headhouse
61240 7 Agricultural Labor Force Grain Milling/Handling - Terminal Elev - Tripper (Gallery Belt)
61310 7 Agricultural Labor Force Grain Milling/Handling - Process Elev - Receiving
61320 7 Agricultural Labor Force Grain Milling/Handling - Process Elev - Precleaning & Handling
61330 7 Agricultural Labor Force Grain Milling/Handling - Process Elev - Cleaning House
61340 7 Agricultural Labor Force Grain Hilling/Handling - Process Elev - Millhouse
61410 7 Agricultural Labor Force Grain Milling/Handling - Priminal Elev - Shipping & Receiving
61420 7 Agricultural Labor Force Grain Milling/Handling - Priminal Elev - Transfer & Conveying
61430 7 Agricultural Labor Force Grain Milling/Handling - Priminal Elev - Headhouse
64000 7 Agricultural Labor Force Fertilizer Application
64100 7 Agricultural Labor Force Fertilizer Application - Ammonia Distributors
66100 7 Agricultural Labor Force Animal Waste - Cattle Dung
66200 7 Agricultural Labor Force Animal Waste - Pig Excrement
66300 7 Agricultural Labor Force Animal Waste - Sheep Manure
70000 1 Population General Solvent Use
71000 1 Population Dry Cleaning
77000 1 Population Application of Surface Coating - Trade/Sales Use
78200 5 Industrial Labor Force Application of Surface Coating - Industrial Use
85100 5 Industrial Labor Force Ferrous Foundries - Induction Furnace (Hot Melt)
85200 5 Industrial Labor Force Ferrous Foindries - Cupola Furnace (Hot Melt)
85300 5 Industrial Labor Force Ferrous Foundries - Electric Arc Furnace (Hot Melt)
NOTE: The surrogates corresponding to Oil Homes and Gas Homes were not used.
B-8
-------
TABLE B-4. FILE FORMAT FOR THE EBCDIC 1985 CANADIAN
SPATIAL ALLOCATION FILE
Record position
First Last
Column
width
Format
Variable
name
Description
1
2
2
I
STATE
AEROS State (Province) Code
3
6
4
-
-
Not used
7
11
5
I
see
Area Source Category Code
12
14
3
I
NGRID
Number of Grids in Province
15
18
4
I
NROW
Row of Grid Cell
19
22
4
1
NCOL
Column of Grid Cell
23
27
5
F5.4
NSAF*
Spatial Allocation Factor for Grid Cellb
'Implied decimal, with four digits to the right of the decimal place and one to the left.
"Spatial allocation factors are dimensionless fractions representing the portion of a province's land use for a specific spatial
category in a grid cell.
Note: Columns 15 to 27 are repeated up to NGRID (900) times.
A89-380
B-9
-------
the process descriptions were not easily matched to specific profiles. In these cases, EPA assiped industry
average default profiles or the overall average default profile. A list of the additional Canadian SCCs and the
profile assignments applied to those SCCs is presented in Table B-5. Similarly, profiles were matched to all
Canadian area source categories in order to facilitate speciation in FREDS. These profile assignments are
shown in Table B-6.
Source category-profile matches for hydrocarbons were input to PSPLIT to create speciation factors suitable
for the use with FREDS. Programs developed to manipulate U.S. particulate data were modified to
accommodate Canadian profile assignments and used to create the particulate speciation factor files.
Unlike the U.S., the Canadian area source annual files contained emission estimates for four alkaline species
(Na, K, Mg, Ca) in addition to TSP. Environment Canada requested that these annual particulate species be
resolved by size fraction, rather than derive all PM species size combinations from TSP. Alliance
subsequently modified its speciation programs to account for this methodology difference. Since some of the
new Canadian area source speciation factors can equal unity, the SFF input format for speciation factors was
charged from 'F8V8' to 'F8V7' to account for the possibility of a nonzero digit to the left of the decimal.
All other PSPLIT and SFF formats remained the same as for the U.S.
Canadian annual area source hydrocarbon values were reported as THC rather than VOC. As a result, the
Hydrocarbon Preprocessor was executed with all methane augmentation flags set to zero, allowing VOC to be
calculated from input THC rather than vice versa.
REFERENCES
Shareef, G., W. Butler, L. Bravo, and M. Stockton. Air Emissions Species Manual Volume I - Volatile
Organic Compound (VOC) Species Profiles. EPA-45Q/2-88-003a (NTIS PB88-225792), U.S. Environmental
Protection Agency, Research Triangle Parte, NC. April 1988.
Shareef, G., and L. Bravo. Air Emissions Species Manual Volume n - Particulate Matter (PM) Species
Profiles. EPA-45G/2-88-003b (NTIS PB88-225800). U.S. Environmental Protection Agency, Research
Triangle Park, NC. April 1988.
A89-3S0
B-10
-------
TABLE B-5. ASSIGNMENT OF HYDROCARBON AND PARTICULATE
SPECIATION PROFILES FOR EXTRA CANADIAN
POINT SOURCE SCCS
Canadian SCC THC profile PM profile
10100104
1178
11201
10200692
3
26101
1020069
3
26101
10201499
0
13501
10299999
0
0
20100291
1001
34001
20100292
1001
34001
20200292
1001
34001
30102299
0
90002
30102702
0
90002
30102703
0
90002
30102799
0
0
30103005
0
90002
30300099
0
0
30300191
0
29101
30300192
0
29101
30300799
0
28401
30300905
16
90004
30501991
0
90013
30600195
9012
26101
30600299
9012
26202
30610099
9012
90014
30700194
9001
90015
30700196
9001
90015
30700201
9001
0
30700208
9001
0
30700290
9001
0
30700210
9001
0
30700399
0
0
30900360
0
90016
31000306
0
0
31000999
0
0
35001001
0
0
40300000
0
0
40301999
9024
0
50100103
122
17106
50100504
122
90001
50300199
9022
90001
A89-380
B-11
-------
see
11100
11200
11300
11400
11510
11520
11610
11620
11630
11640
12100
12200
12300
12400
12510
12520
13100
13200
13300
13400
13510
13520
14100
14200
14300
14400
21000
21100
21150
21200
21300
21400
22100
22140
22210
22220
22230
22240
22250
22260
22270
23110
23120
23130
23140
23150
23210
23230
23240
23250
23260
24100
25110
TABLE B-6. ASSIGNMENT OF HYDROCARBON AND PARTICULATE SPECIATION
PROFILES FOR CANADIAN AREA SOURCE CATEGORIES
Profile
Profile
Profile
Assignment
Assignment
Assignment
THC
TSP
see
THC
TSP
see
THC
TSP
0195
26101
25120
1201
32202
53614
9011
90007
0195
26101
25210
1201
32202
53615
9011
90007
0002
13501
25220
1201
32202
53621
9011
90007
1185
43201
25310
1186
31102
53622
9011
90008
0002
13501
26100
0000
34002
53623
9011
90008
0001
13501
26200
0000
34002
53624
9011
90008
1084
42330
27110
1190
00000
53625
9011
90008
1084
42330
27111
1190
00000
53626
9011
90008
1084
42330
27120
1190
00000
53627
9011
90008
1084
42330
27130
1190
00000
53628
9011
90008
0003
26101
27140
1190
00000
53631
9011
90008
0004
26101
27150
1100
00000
54100
9004
00000
0002
13501
27160
1100
00000
54200
9004
00000
1185
11201
31300
0122
17106
55100
0000
00000
0002
13501
31500
0122
17106
55200
0000
00000
0001
13501
32100
0122
17106
56000
9005
00000
0003
26101
33100
0121
42301
57100
9013
22201
0003
26101
41110
0000
41130
57200
1189
22301
0002
13501
41120
0000
41130
57300
9013
22101
1185
11201
42110
0000
41220
57400
9013
22201
0002
13501
42120
0000
41220
61110
1201
21401
0001
13501
42210
0000
41220
61120
9008
21401
0003
26101
42220
0000
41220
61130
9008
21401
0001
11501
42310
0000
41220
61140
9008
21401
0002
11501
42320
0000
41220
61150
9008
21401
0008
32202
43100
0000
41350
61160
9008
21401
1101
31230
43200
0000
41350
61210
1201
21401
1101
31230
44000
0202
90001
61220
9008
21401
1101
31230
45100
0000
41350
61230
9008
21401
1186
31102
45200
0000
41350
61240
9008
21401
1201
32101
45300
0000
41350
61310
1201
21401
1201
32202
45400
0000
41350
61320
9008
21401
1101
31230
45500
0000
41350
61330
9008
21401
1101
31230
46100
0000
41350
61340
9008
21401
1201
32202
47100
0307
42320
61410
1201
21401
1201
32202
47200
0000
42320
61420
9008
21401
1201
32202
51000
9004
90002
61430
9008
21401
1201
32202
52110
9012
33020
64000
0203
90003
1201
32202
52120
9024
90014
64100
0203
25404
1201
32202
53110
0000
90012
65000
0076
00000
1201
32202
53120
0000
90012
66100
0203
90003
1098
34001
53200
9011
21204
66200
0203
90003
1099
34001
53210
9011
21204
66300
0203
90003
1099
34001
53220
9011
21204
70000
0197
00000
1099
34001
53300
9011
21204
71000
1196
00000
1099
34001
53410
9011
90006
77000
1016
00000
1098
34001
53430
9011
90006
78200
1016
00000
1099
34001
53440
9011
90006
81000
0000
00000
1099
34001
53450
9011
90006
85100
9009
28201
1099
34001
53500
9011
90007
85200
9009
28202
1099
34001
53611
9011
90007
85300
9009
28304
1201
32202
53612
9011
90007
1201
32202
53613
9011
90007
B-12
-------
APPENDIX C
DEVELOPMENT OF ALLOCATION FACTORS FOR NATURAL SOURCES
A89-380
C-l
-------
APPENDIX C
DEVELOPMENT OF ALLOCATION FACTORS FOR NATURAL SOURCES
INTRODUCTION
A significant amount of research on natural source emissions in the United States has been conducted by the
National Oceanic and Atmospheric Administration (NOAA) for NAPAP's Task Group on Atmospheric
Chemistry (Task Group II) and EPA's Atmospheric Research and Exposure Assessment Laboratory (AREAL).
For the 1985 NAPAP Version 2 Emissions Inventory, emissions of alkaline particulate matter were available
in a format readily adaptable to the NAPAP inventory.
Emissions inventories for alkaline particulate matter from natural sources were developed for three categories:
unpaved roads, wind erosion and dust devils. Particulate matter from paved roads are not included in the
1985 NAPAP inventory. Data were provided by Task Group II for total particulate mass and the mass of
total calcium (Ca), magnesium (Mg), sodium (Na) and potassium (K). Data sources for calculation of annual
and resolved natural particulate emissions include field measurements, AP-42 (EPA, 1985), and particulate
matter species profiles (Shareef and Bravo, 1988). Unpaved road emissions were calculated at the county
level by Barnard et al (1986a, 1986b, 1986c, 1987). Wind erosion and dust devil emissions estimates were
provided by Task Group II for major land resource areas (MLRAs). These emissions estimates were resolved
to the county level by EPA. Temporal, spatial and species resolution of the natural particulate emissions were
derived from information provided by Task Group II and EPA.
Environment Canada also supplied natural source particulate emissions estimates from unpaved roads and wind
erosion at the Provincial level. In addition, Environment Canada developed emissions estimates for total and
alkaline particulate matter from paved roads. Temporal, spatial and species resolution were supplied by
Environment Canada. A summary of U.S. and Canadian natural particulate categories included in the 1985
Version 2 inventory is provided in Table C-l.
TEMPORAL ALLOCATION FACTORS
Temporal allocation factors were developed individually for each of the natural source categories in the United
States and Canada. Temporal factors derived for each source category at varying levels of specificity were
A89-380
C-2
-------
TABLE C-1. NATURAL ALKALINE PARTICULATE CATEGORIES INCLUDED IN THE 1985 NAPAP
EMISSIONS INVENTORY (VERSION 2.0)
Canadian
see
Category description
U.S.
901
902
903
Unpaved Road Travel
Wind Erosion, Natural and Agricultural Lands
Dust Devils
41110
41120
42110
42120
42210
42220
42310
42320
43200
Dust - Paved Roads -
Dust - Paved Roads -
Dust Unpaved Roads
Dust Unpaved Roads
Dust Unpaved Roads
Dust Unpaved Roads
Dust Unpaved Roads
Dust Unpaved Roads
Vehicles
Trucks
- Vehicles - Treated Gravel
- Trucks - Treated Gravel
- Vehicles - Untreated Gravel
- Trucks - Untreated Gravel
- Vehicles - Earth Roads
- Trucks - Earth Roads
Agricultural - Erosion from Crops
A89-38G
C-3
-------
concatenated to form a natural source Temporal Allocation Factor File (TAFF) for U.S. natural source
particulates. Canadian natural source temporal profiles were supplied by Environment Canada and are
included in the Canadian area source Temporal Allocation Factor File. The appropriate linker variable was
created in the Canadian annual natural source file for correspondence between the emissions file and the
TAFF.
Unpaved road temporal profiles available in the U.S. area source TAFF were used to temporally allocate U.S.
unpaved road emissions. These factors are SCC-specific and are based on U.S. Department of Transportation
data for light duty vehicles on rural roads.
County-specific temporal profiles for wind erosion emissions in the United States are based on temporal
information supplied by Task Group II. Seasonal fractions were calculated for each of the alkaline species
and total particulates for each county by summing the monthly emissions for a season and dividing the
seasonal total by the annual emissions. The seasonal values for total particulates and the alkaline components
were combined to calculate average seasonal values for all (U.S.) particulate emissions from wind erosion.
The diurnal profile for wind erosion, as recommended by Task Group II, was derived from a figure by
Wigner and Peterson (1987) showing the occurrence of blowing dust by time of day. Daily factors were
assumed to be uniform.
Quality control checks were performed on the wind erosion county specific seasonal profiles to assure the
quality and compatibility of the factors. County codes were matched to those in the U.S. annual natural
source emissions file. Any nonmatches were adjusted to assure complete temporal allocation. In addition, all
seasonal and diurnal fractions were checked to ensure that they summed to unity.
Category-specific seasonal factors for U.S. dust devil emissions were based on data recommended by Task
Group II and from unpublished data from a National Park Service Study (Contract No.
USDICX-0001 -3-0056, W. Malm, NPS Project Manager). The seasonal distribution of dust devils peaks
between June 1 and September 1, decreases linearly from September 1 to October 15, falls to zero between
October 15 and April 15, and finally begins a linear rise from April 15 to June 1. The diurnal profile
recommended by Task Group II from Snow and McClelland (1988) shows maximum activity from 11 a.m. to
3 p.m., a linear decrease from 3 p.m. to 4 p.m., zero activity from 4 p.m. to 10 a.m., and a linear increase
from 10 a.m. to 11 a.m.
The temporal profiles for each category for each scenario were concatenated to form the natural source TAFF.
As a final check on the temporal data, seasonal and hourly fractions were totaled to verify that they summed
to unity.
A89-380
C-4
-------
SPATIAL ALLOCATION FACTORS
The land area surrogate available in the U.S. spatial fraction file was used for spatial allocation of U.S.
natural source emissions. The gridded spatial fractions were matched to the natural source Surrogate Selection
File to generate a natural source Spatial Allocation Factor File (SAFF).
Spatial allocation factors for Canadian natural particulates were provided by Environment Canada. Paved road
particulate emissions were allocated using population as a surrogate and the unpaved road and wind erosion
categories were allocated using the agricultural labor force surrogate.
SPECIATION FACTORS
The NAPAP Task Group n research results related to natural alkaline particulate included estimates of total
mass and the total mass of the alkaline species for limited size ranges. These estimates were divided into the
appropriate size and reactivity fractions to provide estimates that were consistent with the anthropogenic
particulate data. The measured data for unpaved road dust indicated that 24 percent of the total mass was
included in the 10 micron and smaller size fraction. The total TSP mass was estimated by multiplying the
less than 10 micron size fraction by 4.17, the inverse of the 24 percent fraction.
Data that were available in AP-42 suggest that the less than 10 micron size fraction is 36 percent. Estimates
of size fractions for unpaved road dust in size ranges of less than 2.5 micron; less than 5.0 micron; less than
10 micron; less than 15 micron and less than 30 micron are also included in AP-42. These data were plotted
and a linear regression was applied to determine the fraction of particulate in the less than 6.0 micron size
range.
Reactivity fractions were obtained from the Air Emissions Species Manual Volume 11 (Shareef and Bravo,
1988). The AP-42 size fractions were scaled by the ratio of 0.24/0.36 to develop estimates of the size
fraction required for the NAPAP inventory. The same size fractions were applied to both TSP and the
individual alkaline components. These size fractions were used in the speciation of all three natural source
categories. The speciation factors applied to the NAPAP natural source particulate data are listed in
Table C-2.
Size fractions for Canadian natural particulates were available in the Canadian area source Speciation Factor
File (SFF) and the alkaline fractions were obtained from profiles for the natural source categories that were
included in the Air Emissions Species Manual,
A89-380
C-5
-------
TABLE C-2. SPECIATION FACTORS FOR U.S. NATURAL ALKALINE PARTICULATES
0-2-5
Reactive um
Species fraction fraction
TSP NA 0.0207
Sodium (Na) 0.0 0.0207
Potassium (K) 0.0 0.0207
Calcium (Ca) 0.5 0.0207
Magnesium (Mg) 0.5 0.0207
2.5-6.0 6,0-10.0 2,5-10,0
um um um
fraction fraction fraction
0.090 0.129 NA
NA NA 0.2194
NA NA 0,2194
NA NA 0,2194
NA NA 0.2194
A89-380
C-6
-------
REFERENCES
U.S. Environmental Protection Agency. Compilation of Air Pollutant Emission Factors, Fourth Edition, AP-42
Volume 1, (GPO No. 055-000-00251-7), Research Triangle Park, NC. 1985.
Shareef, G. and L. Bravo. Air Emissions Species Manual Volume II - Particulate Matter (PM) Species
Profiles. EPA-450/2-88-003b (NTIS PB88-225800). U.S. Environmental Protection Agency, Research
Triangle Park, NC. April 1988.
Barnard, W.R., GJ. Stensland, and DP. Gatz, 1986a. Alkaline Materials Flux From Unpaved Roads: Source
Strength, Chemistry and Potential for Acid Rain Neutralization, Water Air and Soil Pollution, 30, p.
285-293.
Barnard, WJR., GJ. Stensland, and DP. Gatz, 1986b. Development of Alkaline Emission Flux Estimates
From Unpaved Roads: Problems and Data Needs for Modeling Potential Acid Rain Neutralization,
Paper 86-30.4, Proceedings 79th Annual Meeting of the Air Pollution Control Association, Minneapolis,
MN.
Barnard, WJl., D.F. Gate, and G.J. Stensland, 1986c. Elemental Chemistry of Unpaved Surface Materials,
Paper 86-21.2, Proceedings of the 79th Annual Meeting of the Air Pollution Control Association,
Minneapolis, MN.
Barnard, WJL, DP Gate, and GJ. Stensland, 1987. Evaluation of Potential Improvements in the Estimation
of Unpaved Road Fugitive Emission Inventories, Paper 87-58.1, Proceedings of the 80th Annual Meeting
of the Air Pollution Control Association, New York, NY.
Wigner, K. and R. Peterson. Synoptic Climatology of Blowing Dust on the Texas South Plains, 1947-84.
Journal of Arid Environments, 13, 199-209. 1987.
Snow, J. T. and T. McClelland. Dust Devils at White Sands Missile Range, New Mexico, USA. J. Meteoro,
13, 156-164. 1988.
A89-380
C-7
-------
APPENDIX D
DESCRIPTION OF SEASONAL ELECTRIC UTILITY FACTORS
A89-380
D-l
-------
E. H. PECHAN & ASSOCIATES, INC.
5537 Hempstead Way
Springfield, VA 22151
(703)642-1120
December 6, 1988
Mr. Mark Saeger
Alliance Technologies, Inc.
500 Eastowne Drive
Chapel Hill, NC 27514
Dear Mr. Saeger:
This letter transmits a description of the seasonal eleqtric
utility factors which we have recently developed.
The file itself is in SAS format on the EPA IBM 3090
mainframe. It has the name EXPNAPA. NAPAP. UTILITY . SEASONAL. SASDATA.
The SAS data set name to use in this file is UTILSEAS. The
specific contents are documented in Exhibit 2 to the enclosure.
Feel free to contact roe if you have any questions on this
information or how it was developed.
Sincerely,
Enclosure
cc: David Fratt, Alliance
Rob Lagemann
David Mobley
D-2
-------
Calculating Season Activity Levels for Electric Utilities
Seasonal activity factors for inclusion in the NAPAP
inventory were estimated for electric utilities for the year
1985. The activity factors were based primarily on data
collected by the Department of Energy.
The major source of data contributing to the seasonal
estimates was Energy Information Administration Form 759. This
form collects monthly fuel use data by plant and fuel type (and
prime mover). These monthly data were aggregated to season by
plant and four fuel types (coal, residual oil, distillate oil,
and natural gas). The following seasonal definitions were used:
Winter December, January, February
Spring March, April, May
Summer June, July, August
Fall September, October, November
For each plant, a surrogate measure for "other" fuel was defined
based on the sum of the four fuels indicated above. The "other"
category represents minor fuels used at a few plants such as
petroleum coke and process gas. These minor fuels, included in
NAPAP but not reported in Form 759, were assigned seasonal
allocations averaged from the usage of the four main fuel types.
NAPAP Source Classification Codes (SCC) were used to
determine applicable fuel types based on the first six digits.
The allocation was as follows:
101001-101003 Coal
101004 Residual Oil
101005 Distillate Oil
101006 Natural Gas
101007 Other
The list of applicable NAPAP plants was selected based on
the first three digits of the SCC being 101 (electric utility
boiler) . These plants were then matched to the data derived from
the Form 759 data. Matching was performed using a hierarchical
approach as listed below.
Plant-fuel
State-fuel
D-3
-------
State-all fuels
In the case of Texas, no plant-level data were available in
the NAPAP inventory. As a result, all of the SCCs were matched
at the state-fuel level. Some plants could not be matched on a
state level because, although they reported coal as a fuel, for
example, they reported no coal burned.
Table 1*reports the results of the matching of NAPAP plants
to Form 759 fuel use. For coal, almost 90 percent of total fuel
use from NAPAP was matched at the plant level. Of the remaining
amount, the vast majority was accounted for by the Texas plants
for which NAPAP did not provide plant-level information. For
residual oil, the match ratio was over 98 percent. It was 62
percent for distillate oil and 45 percent for natural gas.
State-level matches were required for less than 0.5 percent of
fuel use for all fuels except distillate oil, which had a state-
level match figure of 8.2 percent.
Most nonmatches occurred due to definitional differences
between NAPAP and EIA 759 data. For EIA 759, electric utilities
are defined as units which provide electricity to the common
grids. The NAPAP definition of electric utilities (SCC 101)
included a number of units which did not meet this criterion.
Even when NAPAP SCC 101 units which did not have the electric
utility Standard Industrial Code (4911) were deleted, some units
remained in NAPAP which would clearly not be classed as utilities
by DOE. The number of such units was small, however, as was
their share of total fuel use.
Table 2 provides an alphabetical list of variables and
attributes contained in the data set below:
EX PNAPA.NAPAP.UTILITY.SEASONAL.SASDATA
The data set name to use in this file is UTILSEAS.
(*) Table D-l.
-------
TABLE D-l. METHODS USED TO ALLOCATE FUEL USE TO SEASONS
Table 1
Methods Used to Allocate Fuel Use to Seasons
Fuel-specific matches
Coal (1000 tons)
Plant-fuel
Texas (see note)
State-fuel
Totals
Number
of Plants
552
74
10
636
Fuels
Used
617,442
72,639
1,584
691,665
% of
Fuels Used
89.3
10.5
0.2
Residual Oil (million gallons)
Plant-fuel 189
Texas (see note) 72
State-fuel 27
Totals 288
Distillate Oil (million gallons)
Plant-fuel 270
Texas (see note) 72
State-fuel 41
Totals 383
Natural Gas (billion cubic feet)
Plant-fuel 341
Texas (see note) 72
State-fuel 24
Totals 437
Other fuels
Plant-fuel 48
Texas (see note) 0
State-fuel 7
Totals 55
Not fuel-specific matches
Other fuels
State total 2
Note: No plant-level data were available from the NAPAP inventory
for Texas
6, 328
90
19
6,437
303
144
40
487
1, 656
1,988
9
3,653
98.3
1.4
0.3
62 . 2
29.6
8.2
45.3
54 .4
0.2
D-5
-------
TECHNICAL REPORT DATA
(Please read Jasiructions on the reverse before completing)
1. REPORT NO. 2,
EPA-600/7-89-010 a
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
The 1985 NAPAP Emissions Inventory: Overview
of Allocation Factors
5. REPORT DATE
October 1989
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Robert A. Walters, Lysa G. Modica, and
David B. Fratt
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Alliance Technologies Corporation
213 Burlington Road
Bedford, Massachusetts 10730
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-4274, Tasks 33/35;
68-02-4396, Task 22
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 11/88-6/89
14. SPONSORING AGENCY CODE
EPA/600/13
is. supplementary notes AEERL project officer is Robert C. Lagemann, Mail Drop 62,
919/541-2709.
16. abstract repDrt documents the development of temporal, spatial, and species
allocation factors for the 1985 National Acid Precipitation Assessment Program
(NAPAP) anthropogenic point and area source emissions inventories. These alloca-
tion factors are used to apportion annual emissions totals into gridded, hourly, spe-
ciated emissions estimates suitable for use as input to atmospheric transport models
such as the Regional Acid Deposition Model. The temporal, spatial, and species
allocation factors are discussed in detail in separate report sections. Each section
contains a description of the methodology for application of the factors, a discussion
of data sources, and documentation of the activities undertaken to create the alloca-
tion factor data sets used in the 1985 NAPAP resolved modeling inventories.
17. KEY WORDS AND DOCUMENT ANALYSIS
a, DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
C. COSATI field/Group
Pollution Allocations
Emission Mathematical Models
Precipitation (Meteorology)
Acidification
Assessments
Inventories
Pollution Control
Stationary Sources
Acid Rain
13B 15 C
14G 12 A
04B
07B.07C
15E
IS. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
Ul
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE y
EPA Form 2220-1 (9-73)
i
------- |