EPA/453/R-94/066/B
United States Office of Air Quality
Environmental Protection Planning and Standards
Agency Research Triangle Park, NC 27711
EPA-453/R-94-066-B
March 1995
Air
& EPA
STUDY OF
VOLATILE ORGANIC COMPOUND EMISSIONS
FROM
CONSUMER AND COMMERCIAL PRODUCTS
Comprehensive Emissions Inventory
-------
EPA-453/R-94-066-B
REPORT TO CONGRESS
VOLATILE ORGANIC COMPOUND EMISSIONS
FROM
CONSUMER AND COMMERCIAL PRODUCTS
VOLUME 2
COMPREHENSIVE EMISSIONS INVENTORY
Emission Standards Division
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
December 1994
-------
ACKNOWLEDGEMENT
This study was completed with a great degree of cooperation and assistance from the
Chemical Specialties Manufacturers Association, the Cosmetic, Toiletry, and Fragrance
Association, the Soap and Detergent Association, the National Aerosol Association, the
Automotive Chemical Manufacturers Council, and the Adhesive and Sealant Council.
Most of the studies in Section 4 were conducted under the auspices of the EPA's Air
and Energy Environmental Research Laboratory.
u
-------
TABLE OF CONTENTS
Section Page
1.0 INTRODUCTION 1-1
1.1 The Problem 1-1
1.1.1 Ozone Nonattainment and Small Sources of VOC Emissions 1-1
1.1.2 VOC' s or NOX -- Which Should We Control? 1 -1
1.1.3 Relative Photochemical Reactivity 1-2
1.2 Congressional Response: The Clean Air Act of 1990 1-3
1.2.1 Requirements Under §183(e) - Consumer/Commercial Products 1-3
1.2.2 Scope of Consumer/Commercial Products Under §183(e) 1-4
1.2.3 EPA's Study and Report to Congress - Purpose and Structure 1-4
1.3 Inventory of VOC's in Consumer and Commercial Products 1-5
1.3.1 Role of the Consumer and Commercial Products Inventory 1-5
1.3.2 Elements of the Inventory 1-6
1.4 Adjustments to Inventory Data 1-7
1.4.1 Fate of Consumer Product VOC's in Wastewater and Landfills 1-8
1.4.2 Emissions in Nonattainment Areas 1-8
1,5 Summary of Findings 1-9
1.6 References 1-13
2.0 CONSUMER PRODUCTS SURVEY 2-1
2.1 Background 2-1
2.1.1 Previous Efforts and the Need for the Survey 2-1
2.1.2 Development of the Survey Approach and Format 2-3
2.1.3 Features of the Survey 2-5
2.1.4 Other Uses of the Information 2-6
111
-------
TABLE OF CONTENTS (Continued)
Section Page
2.2 Survey Questionnaire and Mailing List 2-6
2.2.1 The Survey Questionnaire 2-7
2.2.2 Electronic Reporting Option 2-8
2.2.3 Development of the Mailing List 2-12
2.3 Handling of Survey Responses 2-13
2.3.1 Response Handling and Safeguarding of Confidential Information 2-14
2.3.2 Data Entry 2-17
2.3.3 Batch Files 2-24
2.3.4 SAS Data Sets 2-27
2.4 Results of the Survey 2-29
2.4.1 Response Rate 2-30
2.4.2 Information Obtained from the Survey 2-31
2.5 References 2-54
3.0 PRODUCTS AFFECTED BY EXISTING OR ONGOING FEDERAL
REGULATORY PROGRAMS 3-1
3.1 Introduction 3-1
3.2 Category Descriptions and Sources of VOC Emissions 3-4
3.2.1 Architectural and Industrial Maintenance Coatings 3-4
3.2.2 Automobile Refmishing Products 3-5
3.2.3 Aerospace Coatings 3-6
3.2.4 Wood Furniture Coatings 3-6
3.2.5 Ship and Boat Coatings 3-7
3.2.6 Metal Furniture Coatings 3-7
3.2.7 Flat Wood Paneling Coatings 3-8
3.2.8 Large Appliance Coatings 3-9
3.2.9 Magnet Wire Coatings 3-10
3.2.10 Metal Can Coatings 3-11
3.2.11 Metal Coil Coatings 3-12
3.2.12 Other Metal Product Coatings 3-13
3.2.13 Auto and Light Truck Assembly Coatings 3-14
iv
-------
TABLE OF CONTENTS (Continued)
3.2.14 Paper, Film, and Foil Coatings 3-16
3.2.15 Magnetic Tape Coatings 3-19
3.2.16 Business Machine Plastic Parts Coatings 3-19
3.2.17 Automotive Plastic Parts Coatings 3-20
3.2.18 Flexible Packaging Printing 3-20
3.2.19 Rotogravure Publication Printing 3-21
3.2.20 Lithographic Printing 3-22
3.2.21 Letterpress Printing 3-23
3.2.22 Tire Manufacturing Cements 3-24
3.2.23 Miscellaneous Industrial Adhesives 3-25
3.2.24 Metal Cleaning Solvents 3-26
3.2.25 Industrial Cleanup Solvents 3-27
3.2.26 Petroleum Drycleaning Solvents 3-29
3.2.27 Agricultural Pesticides 3-30
3.2.28 Cutback Asphalt Paving Materials 3-31
3.2.29 Synthetic Fiber Spinning Solvents 3-32
3.2.30 Fabric Coatings 3-34
3.2.30 Fabric Printing 3-35
3.5 References 3-35
4.0 PRODUCTS ADDRESSED BY SPECIAL STUDIES 4-1
4.1 Introduction 4-1
4.2 Construction Materials 4-1
4.2.1 Building Materials and Indoor Air Sources 4-1
4.2.2 Roofing Materials 4-8
4.2.3 Asphalt Concrete Paving Materials 4-14
4.2.4 References 4-15
4.3 Foods, Beverages, and Tobacco 4-17
4.3.1 Alcoholic Beverages 4-17
4.3.2 Deep Fat Frying 4-19
4.3.3 Cigarette, Cigar, and Pipe Tobacco 4-24
4.3.4 References 4-32
-------
TABLE OF CONTENTS (Continued)
Section Page
4.4 Small Combustion Sources 4-34
4.4.1 Kerosene Space Heaters 4-34
4.4.2 Camp Stoves and Lanterns 4-39
4.4.3 Fire Starting Materials 4-41
4.4.4 References 4-44
4.5 Miscellaneous Products 4-48
4.5.1 Mold Release Agents 4-48
4.5.2 Products Used in the Manufacture of Fiberglass Boats 4-51
4.5.3 Automotive Repair Parts Washers 4-54
4.5.4 Products Used in the Textile Industry 4-56
4.5.5 References 4-70
VI
-------
LIST OF FIGURES
Figure 2-1 Company Sheet
Figure 2-2 Page One of the Product Sheet
Figure 2-3 Page Two of the Product Sheet
Figure 2-4 Consumer Product Survey Response Handling System
Figure 2-5 Distribution of Survey Questionnaires
2-9
2-10
2-11
2-15
2-30
LIST OF TABLES
Table 1-1 VOC Emissions from Consumer/Commercial Products 1-10
Table 1-2 Sources of VOC Emissions in 1990 (Nationwide) 1-13
Table 2-1 Results of the Consumer Products Survey 2-33
Table 3-1 Emissions from Products Affected by Existing/Ongoing Programs 3-2
Table 4.1-1 Emissions from Products Addressed by Special Studies 4-2
Table 4.2-1 Substantiated Building Material Sources of VOC Emissions 4-5
Table 4.2-2 Emission Factors for Various Building Materials 4-7
Table 4.2-3 Summary of Roofing Area Worked by Roof Type (1989) 4-12
Table 4.2-4 VOC Emissions from Roofing Application 4-14
Table 4.3-1 Volatile Compounds Identified in Alcoholic Beverages 4-19
Table 4.3-2 Major Constituents of Mainstream Smoke 4-28
Table 4.3-3 Estimated Emissions of VOC Constituents of Mainstream Smoke 4-30
Table 4.4-1 Kerosene Heater Characteristics 4-34
Table 4.4-2 Average Kerosene Heater Emission Factors 4-36
Table 4.4-3 Estimated 1990 Emissions from Kerosene Heaters 4-38
Table 4.5-1 Emissions from Mold Release Agents by Industry 4-51
Table 4.5-2 Emissions from Parts Washer Solvents (Garment Screen Printing) 4-66
Table 4.5-3 Platen Adhesive Emissions 4-67
Table 4.5-4 Spot Cleaner Emissions (Garment Screen Printing) 4-69
VII
-------
1.0 INTRODUCTION
1.1 THE PROBLEM
1.1.1 Ozone Nonattainment and Small Sources of VQC Emissions
National air quality monitoring data from 1986 through 1988 indicate that there are
approximately 100 geographic areas which failed to attain the national ambient air quality
standards (NAAQS) for ozone. Ozone is a major component of smog which poses major
health and environmental concerns when present in high concentrations at ground level. It is
a photochemical oxidant which is formed in the atmosphere through a series of complex
chemical reactions between precursor emissions of volatile organic compounds (VOC's) and
oxides of nitrogen (NOX) in the presence of sunlight.
While most of the large, stationary sources of VOC emissions are covered by existing
regulations, an examination of emissions data completed in 1989 by the Office of Technology
Assessment (OTA) indicates that individually small, dispersed sources of VOC's (area
sources) contribute significantly to the continuing ozone nonattainment problem. According
to the OTA report, Catching Our Breath - Next Steps for Reducing Urban Ozone , one area
source of VOC emissions is the use of a wide range of consumer and commercial products.
1.1.2 VOC's or NOX -- Which Should We Control?
Ground-level (tropospheric) ozone is formed through a series of complex chemical
reactions involving VOC's and NOX in the presence of sunlight. Reductions in the amount
of ozone formed can be obtained through reducing the concentrations of VOC and/or NOV
X
available for reaction. A recent report, Rethinking the Ozone Problem in Urban and
j
Regional Air Pollution , published by the National Research Council explains that a key
factor in reducing ozone formation is the ratio of VOC to NOY in the ambient air. When the
A
VOC-to-NOx ratio is greater than 10:1, VOC reductions have little effect because of the
excess concentration of VOC's available for reaction. In such "NOx-limited" scenarios,
1-1
-------
NOX controls may be much more effective than VOC controls alone in reducing ozone
formation. Conversely, in airsheds which are not NOx-limited, VOC controls can be
effective in reducing ozone formation.
Although VOC controls alone may offer little reduction in ozone formation under
some conditions, there are many instances in which the VOC-to-NOx ratio favors VOC
controls. The U.S. Environmental Protection Agency (EPA) does not anticipate
abandonment of efforts to reduce ozone formation through reduction in VOC emissions,
especially in the case of area sources such as consumer and commercial products. However,
this new way of thinking could affect future strategies for stationary and mobile sources for
which NOX and VOC controls could be tailored to specific conditions.
1.1.3 Relative Photochemical Reactivity
Relative reactivity can be described as the propensity of a VOC species to react in the
presence of NOX and sunlight to form ozone. This phenomenon has been subject to
continuing scientific investigation since well before the mid-1970 's when the EPA adopted its
present scheme of classifying compounds as reactive or negligibly reactive. Although there
has been much discussion over the years among atmospheric scientists, there has been no
broad acceptance, within either the scientific or regulatory communities, of any single
reactivity scale which could be used to make reliable predictions of the ozone formation
potential of specific VOC's. This difficulty arises from the fact that the relative reactivity of
a species is affected by VOC-to-NOx ratios, the presence of other VOC's, and radiation
conditions. Each of these factors can vary widely according to the meteorologic and
geographic conditions not only in the immediate airshed but several days downwind as well.
1-2
-------
1.2 CONGRESSIONAL RESPONSE: THE CLEAN AIR ACT OF 1990
1.2.1 Requirements Under §183(e) - Consumer and Commercial Products
Although control of one small source of VOC emissions may contribute little to
overall ambient air quality, VOC reductions obtained through regulation of multiple small
sources could have a beneficial additive effect. Section 183(e) of the Clean Air Act (CAA)
as amended in 1990 requires the EPA to conduct a study of emissions of VOC's into the
ambient air from consumer and commercial products. The objectives of the study are (1) to
determine the potential of consumer and commercial product VOC emissions to contribute to
ozone levels which violate the NAAQS for ozone; and (2) to establish criteria for regulating
consumer and commercial products or classes or categories of products under the authority of
§183(e) of the CAA. In establishing criteria for regulating consumer and commercial
products, the EPA must take into consideration (1) the uses, benefits, and commercial
demand of consumer and commercial products; (2) any health or safety functions served by
the products; (3) those consumer and commercial products that emit highly reactive VOC's
into the ambient air; (4) those products that are subject to the most cost-effective controls;
and (5) the availability of any alternatives to such consumer and commercial products that are
of comparable costs, considering health, safety, and environmental Impacts. On completion
of the study, the EPA must submit a report to Congress that documents the results of the
study.
Upon completion of the report, the EPA must list those categories of products which
are determined, based on the study, to account for at least 80 percent of the total VOC
emissions, on a reactivity-adjusted basis, from consumer and commercial products in areas
which violate the NAAQS for ozone. The EPA must divide the list into 4 groups by
priority. Pursuant to this requirement, the EPA will publish the prioritized category list in
the Federal Register following submittal of the report to Congress. Beginning no later than 2
years following publication of the list, the EPA must regulate one group every two years
until all 4 groups are regulated.
1-3
-------
1.2.2 Scope of Consumer and Commercial Products under S183(e)
According to the definition in §183(e), "the term 'consumer or commercial product'
means any substance, product (including paints, coatings, and solvents), or article (including
any container or packaging) held by any person, the use, consumption, storage, disposal,
destruction, or decomposition of which may result in the release of volatile organic
compounds. The term does not include fuels or fuel additives regulated under section 211,
or motor vehicles, non-road vehicles, and non-road engines as defined under section 216."
The EPA has determined that the statutory definition of consumer or commercial
product is much broader than just the "traditional" consumer products (e.g., personal care
products, household cleaning products, household pesticides, etc.). Instead, consumer and
commercial products include virtually all VOC-emitting products used in the home, by
businesses, by institutions, and in industrial manufacturing operations. Among these
products are a wide range of surface coatings, metal cleaning solvents, graphic arts inks,
industrial adhesives, agricultural pesticides, asphalt paving materials, and many other
products used in industrial manufacturing processes, many of which have been previously
regulated by the EPA and/or by the States.
1.2.3 EPA's Study and Report to Congress — Purpose and Structure
The primary purpose of the study and report to Congress is to educate the EPA and
Congress on consumer and commercial products as contributors to ozone nonattainment and
to identify opportunities for reduction of VOC emissions from the use of these products. In
addition, some information obtained from the EPA studies was used to establish criteria for
regulation of consumer and commercial products and utilized during the process of exercising
the criteria to develop the regulatory agenda.
In order to prepare the report, the EPA conducted several individual studies. Some
of these studies pertain to specific categories of products for which the EPA has little or no
existing information; five "generic" studies focus on topics which do not relate to any
particular category of products. The 14 individual studies address the following topics:
1-4
-------
(1) inventory of VOC emissions from consumer and commercial products; (2) fate of
consumer and commercial product VOC's in landfills; (3) fate of consumer and commercial
product VOC's in wastewater; (4) aerosol products and packaging systems; (5) economic
incentives to reduce VOC emissions from consumer and commercial products; (6) underarm
antiperspirants and deodorants; (7) hair care products; (8) aerosol spray paints; (9) adhesives
and sealants; (10) household cleaning products; (11) nonagricultural pesticides (including
antimicrobials); (12) automotive aftermarket products; (13) air fresheners; and (14)
architectural and industrial maintenance coatings.
The report to Congress is comprised of five volumes which document the results of
the generic studies, plus one volume (the executive summary) which contains (1) a summary
of findings of the EPA's study of consumer and commercial products; (2) a discussion of
relative photochemical reactivity as it applies to §183(e); (3) criteria developed by the EPA
for regulating products under §183(e); and (4) summaries of each of the 14 individual
studies. The six volumes of the report are:
Volume 1: Executive Summary
Volume 2: Comprehensive Emissions Inventory
Volume 3: Fate of Consumer and Commercial Products in Landfills
Volume 4: Fate of Consumer and Commercial Products in Wastewater
Volume 5: Economic Incentives to Reduce VOC Emissions from Consumer Products
Volume 6: Aerosol Products and Packaging Systems
1.3 INVENTORY OF VOC's IN CONSUMER AND COMMERCIAL PRODUCTS
1.3.1 Role of the Consumer and Commercial Products Inventory
The purpose of the inventory effort was to develop estimates of annual VOC
emissions in ozone nonattainment areas for every consumer and commercial product category
subject to §183(e). These estimates, along with other information, were used to develop a
prioritized list of categories to be regulated as required by §183(e). Seven criteria for
1-5
-------
regulating consumer and commercial products have been developed based on the factors
listed in §183(e). Three of these criteria (emissions of highly reactive compounds,
availability of lower-VOC alternatives, and annual VOC emissions) will be evaluated based
on the inventory data. These criteria are presented in Volume 1 of the Report to Congress.
The emission estimates and per capita emission factors presented in this report can
also be used by State and local agencies in developing emission inventories for specific mixes
of products and categories within their jurisdiction.
1.3.2 Elements of the Inventory
Because of the wide variety of products subject to §183(e), emission estimates were
developed independently for three major subdivisions of the universe of consumer and
commercial products.
1.3.2.1 Traditional Consumer Products (including institutional uses)
These are products which most people associate with the term "consumer products."
This group includes such products as personal care products, household cleaning products,
household pesticides, automotive maintenance and detailing products, and many others.
Institutional and commercial uses of these or similar products are also considered within the
scope of traditional consumer products. Emission estimates for these categories were
obtained through a 1993 census survey of consumer product manufacturers. This segment of
the inventory is discussed in detail in Section 2.
1.3.2.2 Industrial Products Affected by Existing or Ongoing Regulatory Programs
The statutory definition of "consumer or commercial product" is much broader than
just the traditional consumer products and includes virtually all VOC-emitting products used
in the home, by businesses, by institutions, and in industrial manufacturing operations. This
segment of the consumer and commercial products inventory includes those products which
are used industrially and are affected by existing or ongoing regulations developed by the
1-6
-------
EPA and/or by the States. These products include a wide range of surface coatings, metal
cleaning solvents, adhesives, inks, agricultural pesticides, asphalt paving materials, and a
host of others. Estimates of VOC emissions from these products were obtained from
background documentation developed for the respective regulations or guidance documents.
Emission estimates for previously regulated categories are based on "residual" emissions
(i.e., emissions which continue after controls have been applied). Estimates for categories
associated with ongoing standards development are 1990 "baseline" emission estimates from
those categories (i.e, emissions before any controls have been applied). This segment of the
Inventory is discussed in detail in Section 3.
1.3.2.3 Products Addressed by Special Studies
Many consumer and commercial products fall outside the scope of the consumer
products survey and the categories affected by existing or ongoing regulatory programs.
This third segment of the inventory is comprised of products which do not fall within the
first two segments. It includes such products as foods, beverages, and tobacco; small
combustion sources such as kerosene heaters, camp stoves, artificial fireplace logs, and
commercial explosives; products used in the roofing and textile industries; and a wide range
of miscellaneous products including but not limited to, mold release agents, automotive parts
washers, and fiberglass boat manufacturing products. Emission estimates for these products
were obtained through special studies conducted by the EPA. These studies were limited to
searches of the literature for relevant scientific investigations in which VOC emissions from
these products were characterized and quantified. This segment of the inventory is discussed
in detail in Section 4.
1.4 ADJUSTMENTS TO INVENTORY DATA
In accordance with §183(e), an effort was undertaken to develop estimates of VOC
emissions from consumer and commercial products in ozone nonattainment areas. Because
the "raw" data from the consumer products survey and the other sources were primarily for
nationwide mass emissions based on the VOC content of the products, the following factors
1-7
-------
were considered in an effort to develop realistic estimates which could be used to satisfy the
requirements of §183(e).
1.4.1 Fate of Consumer Product VQC's in Wastewater and in Landfills
Historically, inventories of VOC emissions from consumer and commercial products
have been based on the assumption that all VOC's contained in these products eventually
volatilize, enter the ambient air, and are thus available to react to form ozone. However, the
VOC's in some products such as soaps, laundry detergents, household cleaners,
mouthwashes, disinfectants, etc. may be combined with water during and/or following use
and enter the wastewater stream. In order to ensure that the inventory results reflected actual
VOC emissions rather than VOC content, the EPA initiated an investigation to identify
information on the fate of consumer product VOC's that enter the wastewater stream. This
topic is discussed in detail in Volume 4 of the Report to Congress. Based on this
information, and information provided by the major consumer product industry trade
associations 3>4»5>"; final emission estimates for several product categories reported in this
inventory reflect a "percent VOC content emitted" of somewhat less than 100 percent.
Another area of concern was that it may be possible for a portion of the unused
product (and VOC content) to remain in the container packaging following disposal in
landfills. A study was undertaken by EPA to determine whether some adjustment of the
emission estimates should be made to account for this phenomenon. Based on the study,
presented in Volume 3 of the Report to Congress, the EPA determined that, because of the
scarcity of information and the variability of landfill operating procedures, it was not possible
to adjust the emission estimates to account for fate of the VOC's in landfills. Consequently,
for the purpose of this inventory study, the assumption was made that any remaining VOC
content in the disposed products is eventually emitted to the air.
1.4.2 Emissions in Nonattainment Areas
Section 183(e) primarily focuses on VOC emissions in ozone nonattainment areas.
Because much of the emissions data collected reflects nationwide emissions, some adjustment
1-8
-------
was necessary to obtain estimates for nonattainment areas. Since emissions from traditional
consumer products are roughly proportional to population, the nationwide estimates from the
consumer products survey were scaled down based on the distribution of population in ozone
nonattainment areas. In 1990, approximately 148 million of the 248 million persons in the
U.S. resided in ozone nonattainment areas (59.7 percent). Based on this distribution, the
nationwide emission estimates for traditional consumer products were multiplied by a factor
of 0.6 to reflect emissions in nonattainment areas.
For many categories of industrial products, estimates were developed based on actual
locations of emission sources. Consequently, nonattainment area emission estimates were
developed for those products directly. For any categories for which specific locations of
emission sources were unknown, the population scaling method discussed above was
employed.
1.5 SUMMARY OF FINDINGS OF THE INVENTORY STUDY
Based on estimates developed for the many categories of products investigated in this
study, VOC emissions from the use of consumer and commercial products in ozone
nonattainment areas totalled 3,332,145 tons in 1990. Table 1-1 presents a summary of the
estimates which were developed during this study. More detail is provided in Sections 2, 3,
and 4.
Following Table 1-1 is a discussion of the relative contribution of consumer and
commercial products among all sources of man-made VOC emissions nationwide.
1-9
-------
TABLE 1-1
VOC EMISSIONS FROM CONSUMER/COMMERCIAL PRODUCTS
IN OZONE NONATTAINMENT AREAS (1990)
Product Category
Traditional Consumer Products
Personal care products
Household products
Automotive aftermarket products
Adhesives and sealants
FIFRA-regulated products
Coatings & related (except AIM)
Other traditional consumer products
Architect & Indust Maint (AIM) Coatings
Industrial Adhesives
Tire manufacturing cements
Platen adhesives (textile industry)
Miscellaneous industrial adhesives
Industrial Solvents
Metal cleaning solvents
Industrial cleanup solvents
Petroleum drycleaning solvents
Synthetic fiber spinning solvents
Textile industry equipment cleaning
Textile industry spot cleaners
Automotive repair - parts washers
174,115
55,095
106,469
45,467
121,464
89,405
5,194
26,400
2,092
201,600
36,000
150,000
54,600
46,200
68
848
2,607
Emissions in
Nonattainment
Areas (tons/yr)
597,209
315,000
230,092
290,323
1-10
-------
TABLE 1-1 (Continued)
VOC EMISSIONS FROM CONSUMER/COMMERCIAL PRODUCTS
IN OZONE NONATTAINMENT AREAS (1990)
Product Category
Industrial Coating Operations
Automobile refinishing products
Aerospace paints and coatings
Wood furniture coatings
Ship and boat coatings
Metal furniture coatings
Flat wood paneling coatings
Large appliance coatings
Magnet wire coatings
Metal can coatings
Metal coil coatings
Other metal product coatings
Auto and light truck assy coatings
Paper, film, and foil coatings
Magnetic tape coatings
Business machine plastic part coatings
Automotive plastic part coatings
Flexible packaging printing
Rotogravure publication printing
Lithographic printing
Letterpress printing
Fabric coating
Fabric printing
Mold release agents
55,000
107,500
60,000
15,100
63,000
20,000
15,600
4,800
45,000
21,600
218,400
75,000
65,000
5,500
5,500
16,500
150,000
20,000
600,000
28,200
21,000
25,200
75,400
Emissions in
Nonattainment
Areas (tons/yr)
1,713,300
1-11
-------
TABLE 1-1 (Continued)
VOC EMISSIONS FROM CONSUMER/COMMERCIAL PRODUCTS
IN OZONE NONATTAINMENT AREAS (1990)
Product Category
Asphalt Paving and Roofing Materials
Cutback asphalt paving materials
Asphalt concrete paving materials
Roofing - built-up
Roofing - elastomeric
Roofing - modified bitumen
Other Consumer/Commercial Products
Fiberglass boat manufacturing
Alcoholic beverage use
Smoking tobacco
Kerosene space heaters
Camp stoves and lanterns
Artificial fireplace logs
Agricultural pesticides
ALL §183(e) CATEGORIES
128,400
360
7,126
9,123
2,276
12,100
11,600
37
39
6
154
15,000
Emissions in
Nonattainment
Areas (tons/yr)
147,285
38,936
3,332,145
To put the VOC contribution from consumer and commercial products into
perspective, a 1990 estimate of all man-made VOC emissions was derived by modifying an
"I
existing 1985 estimate. The old estimate was modified by grouping all of the emission
sources which fall within the scope of §183(e) and replacing the sum of their individual 1985
estimates with the new estimate developed in this study. To compare consumer and
commercial products with other large sources of man-made VOC emissions, it was necessary
1-12
-------
to modify the estimate developed for nonattainment areas to reflect nationwide emissions
from consumer and commercial products. Employing the population distribution factor (0.6)
discussed in Section 1.4.2, the nonattainment area estimate (3,332,145 tons/yr) for all
categories under §183(e) was scaled up to a nationwide estimate of about 5,500,000 tons/yr,
or about 26 percent of all man-made VOC emissions. This scaled-up estimate is rough and
should be used only for this comparison exercise. The principal source categories and their
respective VOC emission estimates are presented in Table 1-2.
TABLE 1-2 SOURCES OF VOC EMISSIONS IN 1990 (NATIONWIDE)
Emission Source Category
Mobile Sources
Consumer/ Commercial Products
Petroleum Marketing
Fuel Combustion (Stationary Sources)
Forest, Agricultural, and Other Burning
Petroleum Refineries
Organic Chemicals Manufacturing
Industrial Manufacturing
TOTAL FOR ALL SOURCES
Nationwide
Emissions (tons/yr)
7,920,000
5,500,000
2,460,000
2,300,000
990,000
820,000
550,000
400,000
20,940,000
Share of Total
(percent)
37.8
26.3
11.8
11.0
4.7
3.9
2.6
1.9
100.0
1.6 REFERENCES
1. U.S. Congress, Office of Technology Assessment, Catching our Breath: Next Steps
for Reducing Urban Ozone, OTA-O-412, Government Printing Office, Washington,
D.C., July 1989.
2. National Research Council, Committee on Tropospheric Ozone Formation and
Measurement, Rethinking the Ozone Problem in Urban and Regional Air Pollution,
Washington, D.C., 1991
3. Letter from D. Fratz, Chemical Specialties Manufacturers Association, to B. Moore,
U.S. EPA, Office of Air Quality Planning and Standards, Comments on raw survey
data concerning compounds reported, percent VOC emitted, and market share
reporting, March 25, 1994.
1-13
-------
4. Letter from J. Graf, Cosmetic, Toiletry, and Fragrance Association, to B. Moore,
U.S. EPA, Office of Air Quality Planning and Standards, Comments concerning
survey - percent VOC emitted and market share reporting, May 6, 1994.
5. Letter from R. Sedlak, Soap and Detergent Association, to B. Moore, U.S. EPA,
Office of Air Quality Planning and Standards, Comments on percent VOC emitted
and market share reporting, May 16, 1994.
6. Letter from R. Sedlak, Soap and Detergent Association, to B. Moore, U.S. EPA,
Office of Air Quality Planning and Standards, Further comments on percent VOC
emitted from laundry and dishwashing products, August 8, 1994.
7. U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park,
North Carolina, Alternative Control Techniques - Control Techniques for Volatile
Organic Compound Emissions from Stationary Sources (EPA-453/R-92-018),
December 1992.
1-14
-------
2.0 CONSUMER PRODUCTS SURVEY
2.1 BACKGROUND
One important segment of the inventory of VOC's from consumer and commercial
products are the "traditional consumer products" described in Section 1.3.2.1. In order
to develop an accurate, reliable emission estimate for this segment, the EPA and the
consumer products industry undertook a massive effort to assemble accurate formulation
and sales information for over 24,000 individual consumer products. The following
sections describe the survey process and the results obtained from that effort.
2.1.1 Previous Efforts and the Need for the Survey
There are four major studies that were previously conducted and are believed to
be the best, publicly available, comprehensive inventories of VOC content in consumer
and commercial products prior to the current effort. The four studies are listed below:
1. Photochemically Reactive Organic Compound Emissions from Consumer
and Commercial Products, performed by Science Applications International
Corporation (SAIC) for EPA Region H (EPA-902/4-86-001, November
1986)
2. Compilation and Speciation of National Emissions Factors for
Consumer/Commercial Solvent Use, performed by SAIC for EPA's Office
of Air Quality Planning and Standards (EPA-450/2-89-008, April 1989)
3. Analysis of Regulatory Alternatives for Controlling Volatile Organic
Compound (VOC) Emissions from Consumer and Commercial Products in
the Afew York City Metropolitan Area (NYCMA), performed by Pacific
Environmental Services, Inc. (PES) for the New York State Department of
Environmental Control (NYDEC) (January 1990)
4. Expansion of the New York Study: Evaluation of VOC Emission Reduction
Alternatives from Selected Consumer and Commercial Products, performed
by PES for EPA's Air and Energy Engineering Research Laboratory
(February 1990)
The first three studies are represented as being VOC emissions inventories,
although they are actually inventories of the VOC content of the consumer and
2-1
-------
commercial products. The third study identified brands of products sold in the New York
City area. The fourth study compared brands sold in four other cities to those reported
in the third study. The number of product categories, the geographic coverage, and the
species of interest vary from study to study, due to the different objectives for each
study.
The formulation data used in the first two studies are dated. Many of the
formulations date back to the chlorofluorocarbon (CFC) propellant era or shortly after the
CFC propellant ban in 1978. The formulations used in these two studies were also
primarily generic formulas. The sales data for these two studies are also somewhat old,
dating generally from 1981 to 1985. The PES study uses formulation data gathered
directly from manufacturers and distributors and includes individual species. Both the
formulation and usage data for the PES study are from 1988. None of the listed studies
was able to match brand-specific sales data with specific product formulations. Both the
second SAIC and the two PES studies indicate that some data are available, but that such
data are inadequate to develop the comprehensive emissions inventory required under
§183(e).
Market share data were not used in developing the typical formulations nor in
apportioning total sales volume to specific brands in any of the existing inventories. The
NYDEC study performed by PES contains a great deal of brand-specific formulation
data; however, these data were obtained by a voluntary letter survey of suppliers and
distributors of commercial and institutional product brands in the New York area. These
brands were rarely the same as the nationally marketed retail consumer brands for which
market share information is available. Formulation information for nationally marketed
retail consumer products is, therefore, one of the main items missing from previous
inventories. Market share and total sales volume data for these products are available
from a variety of market research organizations.
A second area of concern is the total sales volume and market share data for
products marketed for commercial and institutional use. The NYDEC study's letter
survey obtained many formulation and sales data for individual products in this category.
2-2
-------
These data were extrapolated through the assumption that the total market volume was
five times the volume represented in the survey responses because only a 20 percent
response was obtained from the survey. A more defensible extrapolation method must be
implemented to represent the total market volume for commercial and institutional use.
Available market research data appear to address only the retail consumer market.
Many sources of market share data and sales volume data are available, but the
boundaries of the market represented by the data are not always clear. It is also very
possible that widely different sales volumes will be reported by different market research
firms. This is the third major consideration for assessing the accuracy of inventories of
the consumer/commercial products categories.
None of these existing inventories are adequate for regulatory assessment and
development purposes. In addition, no adequate speciated inventories exist for any
significant consumer product category.
2.1.2 Development of the Survey Approach and Format — A Cooperative Effort
The following discussion summarizes the events that led to the development of the
Consumer and Commercial Products Survey.
2.1.2.1 Events Prior to the Clean Air Act Amendments of 1990
In anticipation of the Clean Air Act Amendments, the EPA conducted a
symposium in November 1989 on regulatory approaches for reducing VOC emissions
from the use of consumer products. The purpose of the symposium was to initiate a
dialogue among the EPA, the States, and industry toward working cooperatively in
addressing this source of emissions. Approximately 100 companies and trade associations
participated in the symposium. State environmental agencies from California, Michigan,
New Jersey, New York, North Carolina, Texas, and Virginia were represented as well.
The proceedings of that symposium were documented in an EPA report (EPA-450/3-90-
008, January 1990).
2-3
-------
One of the key issues raised by industry representatives at the symposium was the
lack of a reliable, speciated inventory of VOC emissions from traditional consumer
products. The industry demanded that the EPA employ a census survey of consumer
product manufacturers and distributors to develop such an inventory.
2.1.2.2 Development of the Survey Questionnaire and Approach
In November of 1990 the EPA and the Cosmetics, Toiletries, and Fragrances
Association (CTFA) met twice with the Chemical Specialties Manufacturers Association
(CSMA) at CSMA's headquarters in Washington, D.C. to discuss the plan for developing
the inventory. It was suggested that the best approach for a number of consumer product
categories would be through a survey of the manufacturers of those products.
Information for this consumer products survey would be collected by the EPA via
questionnaires distributed under the authority of Section 114 of the CAA. This approach
was suggested by the consumer products industry, specifically the CSMA.
Refinements to the survey approach were made over the next two years by the
work group consisting of EPA, CSMA, CTFA, the Soap and Detergent Association
(SDA), several other associations, and representatives of several companies (L&F
Products, Sherwin-Williams, Procter & Gamble, Drackett, Gillette, S.C. Johnson Wax,
United Industries, Aeropres Corporation, CCL Custom Manufacturing, Helene Curtis,
Inc., and Carter-Wallace, Inc.). This work group met on several occasions to develop a
joint protocol for the consumer products inventory effort.
In December 1991, the survey questionnaire developed by the work group was
sent to nine test respondents. The responses from this test were analyzed and the results
presented at the March 1992 meeting of the work group. Revisions to the questionnaire
were made and sent to the work group members for final endorsement.
2-4
-------
2.1.2.3 Approval to Conduct the Survey
It was recognized early in the planning of the survey that, due to the large number
of survey questionnaires necessary, Office of Management and Budget (OMB) approval
would be required through the submission of an information collection request (ICR).
The CSMA offered to endorse the ICR and state that the industry and EPA reached a
consensus that the Section 114 approach is the most suitable vehicle for gathering the
required information. The ICR was submitted to OMB in July 1992, and approval was
granted by OMB in December 1992.
2.1.3 Features of the Survey
The survey focus was to establish a reasonably accurate inventory of the VOC
content of consumer and commercial products. This is the first step necessary to
determine the VOC emissions from these products. Separate studies have been conducted
to estimate the fraction of VOC content emitted to the atmosphere for various products.
The product categories included in the survey were selected to ensure the most
comprehensive database possible without causing an unreasonable economic burden. The
list of product categories (along with the category codes used with each) are given in
Appendix A, under Attachment A to the survey form (Enclosure 1).
Listing of individual VOC ingredients (speciation) was necessary to develop the
most useful, accurate, and reliable data base of VOC content in consumer and
commercial products. If speciation were not provided by the manufacturers in the course
of the inventory, it would have to be developed or assumed by EPA, and would be
challenged as inaccurate. Therefore, the goal of the survey was to create an accurate
estimate of total VOC for each product category with small ranges of uncertainty for each
species in each category. Speciation data for each category creates greater flexibility,
allowing the data to be used in a number of ways:
1. To perform quality assurance checks on the submitted formulation data.
2-5
-------
2. To estimate the fraction of the VOC content which escapes to the ambient ab-
using the known fate pathways identified in other EPA studies.
3. To develop hazardous air pollutant and indoor air pollution inventories.
2.1.4 Other Uses of the Information
The primary use of the survey was to provide the basic information (VOC
content) that was used to develop VOC emission estimates for consumer and commercial
products.
A secondary use of the inventory could be in the preparation of a national
hazardous air pollutant (HAP) inventory. The CAA requires the EPA to list major
sources of hazardous air pollutants, including urban area sources. Some of the major
HAP's in consumer and commercial products include methylene chloride, trichloroethane,
toluene, xylene, and methanol. Some additional toxic materials are present in smaller
quantities. These include hexane, ethylene glycol, diethanolamine, chlorobenzene,
methyl ethyl ketone, and methyl isobutyl ketone. The consumer products portion of the
HAP inventory may be particularly useful for indoor air pollution assessments.
A third use of the inventory derived from the survey will be to provide per capita
emission factors for the States to use in State Implementation Plan (SIP) inventories.
Some differences in per capita use of specific products by geographic region or season
may exist, but identification of these variables is beyond the scope of this study.
2.2 SURVEY QUESTIONNAIRE AND MAILING LIST
Survey packets were mailed to approximately 3,700 respondents from a mailing
list developed with the assistance of several trade associations. The survey packets
consisted of a cover letter, the survey questionnaire form (Enclosure 1), a computer
floppy diskette with a computer program that emulated the questionnaire (Enclosure 2),
and supporting documents explaining EPA's regulatory authority and procedures for
2-6
-------
protecting confidential business information (Enclosures 3 through 5). Appendix A
contains an example of a survey packet without the accompanying computer diskette.
Appendix B contains the original mailing list.
2.2.1 The Survey Questionnaire
The purpose of this section is to give an overview of the information requested in
the survey questionnaire. The survey questionnaire form and instructions explain the
information being requested and should be referred to for greater detail. The survey
questionnaire consists of two "sheets", each requesting different information.
The Company Sheet requests general information on the company and/or division
providing the survey information. The Company Sheet also allows the company to
indicate that the information request does not apply to them either because they do not
produce any of the products listed as included in one of the survey categories (Attachment
A to the Survey Form) or that they are not identified on the product label as the "owner"
of the brand name. Figure 2-1 shows the Company Sheet.
The Product Sheet consists of a two-page form that requests product-specific
information. The information requested includes the Company and Division reporting the
information, the product category code taken from a list of categories listed in
Attachment A to the survey form, the product or product group name, information on
product packaging and sizes, the sales of the product in 1990 (in pounds), the total
amount of reportable VOC (RVOC) in the product (percent of product net weight), a list
of each RVOC ingredient (defined in Attachment B of the survey form) that comprises at
least five percent of the total product weight, the percentages of methylene chloride and
1,1,1-trichloroethane in the product (exempt as VOC but of interest as toxic air
pollutants), and which classes of information should be considered confidential business
information (CBI). The list of RVOC ingredients is requested in Item 11A on the survey
questionnaire. If the responding company did not have this information, they were
instructed to provide an alternate contact (company) that could provide that information.
This can occur when a company, such as a large chain store, sells it's brand of a product
2-7
-------
(e.g., spray paint) that is actually produced and packaged by second firm (e.g., a contract
packager). Figure 2-2 shows page 1 and Figure 2-3 shows page 2 of the Product Sheet.
2.2.2 Electronic Reporting Option
The computer version of the survey form consisted of an introductory screen that
appeared at the beginning of the program, two data entry screens emulating the two pages
of the product sheet, and a closing screen used at termination of a session with the
program. Additional functions allowed the person using the software to get context-
specific help (help related to the current data being entered), to search the data files for a
specific record (information on a specific product), print the data in the data files by
either record number or by alphabetical listing by product name, and to create an export
disk for submittal of the entered data. The computer files generated by the software were
designed to be sent on computer disk instead of on individual paper forms for each
product. The Company Sheet, however, should have accompanied each submission,
whether on disk or paper. This was done intentionally to avoid problems when the data
on the disk is unreadable or when questions arise on data that was entered, that a contact
person could be reached. An additional reason was to ensure that a signature from a
company employee always accompanied a data submittal.
2-8
-------
FIGURE 2-1 COMPANY SHEET
1. Company Name:
Division Name:
2. Person to Contact:
3. Address:
4. Phone:
5. Type of Business (mark with "X" for all that apply):
Manufacturer
Retailer
Distributor
Private Label
Custom Contract
Contract Packager Packager
6. Indicate Number of Product Sheets Attached:
or Zero Product Sheets Are Attached Because:
This company does not manufacture, distribute, or sell any of the
reportable products listed in Attachment A.
This company does manufacture, distribute, or sell some of the reportable
products listed in Attachment A, but this company is not the party
responsible to report because it is not named on the product label.
7. Certification. The company or division officer who is responsible for environmental
compliance or government affairs must sign the certification statement below:
"I hereby certify that, to the best of my knowledge and belief, all information
entered on this Company Sheet and any attached Product Sheets is complete
and accurate."
NAME SIGNATURE
TITLE DATE SIGNED
2-9
-------
FIGURE 2-2 PAGE ONE OF THE PRODUCT SHEET
Page 1 of 2
1. Company Name :
Division Name :
2. Product Category Code (see Attachment A):
Description (from Attachment A and/or additional descriptions):
3. Product Form :
Aerosol
Spray Pump I I Liquid
Gel
Solid
Other
4. Full Product or Product Group Name
5. If the data on this Product Sheet represent more than a single Stock Keeping Unit (SKU), enter the
number of SKU's represented (see instructions on grouping products):
6. Major Customer Type :
Household
Comm/lnst
7. Product Size -by Volume :
to 55 gal. I 1 >55 gal.
or
0 to 32 fl.oz.
-by Weight :
0 to 1 Ib.
to 5 Ib.
Industrial
>32 fl.oz. to 5 gal. I I >5
5 to 20 Ib.
>20 Ib.
8. Net Product Weight sold in the U.S. for 1990 (pounds) :
9. Total Reportable VOC (weight % of net products):
Do not include methylene chloride or 1,1,1-trichloroethane in this total.
If formulation information is not available to your company, go to item 11B.
10. Total Number of Reportable VOC ingredient compounds :
2-10
-------
FIGURE 2-3 PAGE TWO OF THE PRODUCT SHEET
Page 2 of 2
11 A. List the largest reportable VOC ingredient compounds and their weight %'s for all reportable VOC that
are 5% or more of the total net product weight. Use the criteria listed in Attachment B to determine
if the compound is an RVOC. Please list these compounds in order of highest to lowest weight percent,
if possible. Do not report individual components of ingredients which are fragrance materials.
Reportable VOC Ingredient Compounds
Name
CAS No.
Weight % in Final
Product
1.
2.
3.
4.
5.
6.
7.
8.
All Other Reportable VOC
Total Reportable VOC %
(should match item 9)
Additional information is requested separately for two other compounds: methylene chloride and 1,1,1-trichloroethane.
Do not include these compounds in total RVOC.
Methylene chloride
1,1,1-tnchloroethane
75-09-2
71-55-6
11B. Enter name and address of company with formulation information if you do not have the data to
complete item 11 A. This requirement does not apply to ingredients which are fragrance materials.
12.
Indicate which items contain Confidential Business Information :
Sales
Total RVOC %
Ingredients
Ingredient Wt. %s
2-11
-------
2.2.3 Development of the Mailing List
The mailing list for the survey questionnaires was derived from a number of sources, most
notably CSMA, CTFA, SDA, the Motor Equipment Manufacturers Association (MEMA), the
Adhesives and Sealants Council, and the California Air Resources Board (CARB). There was
a degree of overlap between the lists provided by the various sources, and although methods
were used to eliminate as many duplications as possible, there were some that remained in the
final version of the mailing list. These became apparent only after responses to the survey were
returned. There were over 9,000 entries on the mailing list before any deletions of duplicates
were performed.
The list provided by CSMA included their membership, as well as the list of companies to
whom they occasionally send promotional material (i.e., potential CSMA members and other
interested parties). The list provided by CTFA included names listed in two documents that they
publish: CTFA - Who's Who 1992 (their membership directory) and the Cosmetic Industry ON
CALL 1992, which is put out jointly with the American Academy of Dermatology. The list from
the Adhesives and Sealants Council was taken from their membership directory. The
membership lists often consisted of sales managers or administrative officers who were not the
best contact person for receiving the survey questionnaire. In many cases, addressing survey
packets to these individuals resulted in delayed responses and may have reduced the overall
response rate (see discussion in Section 2.4). The CTFA "On Call" list gives the technical
and/or emergency contact person for companies that produce cosmetics and cosmetic ingredients.
The list from CARB was found to be the least reliable, with many companies out-of-business
and/or not responding. The lists provided by MEMA and SDA were given after the preliminary
list from the other sources was compiled and reviewed by them. They then submitted potential
respondents that were not already included on the list.
After all contributions to the survey mailing list were completed, the list was printed in
alphabetical order to account for obvious duplications. This reduced the number of listings by
one-half. A second reduction was accomplished by doing a search of the data file by substrings
(unique parts) of each company's name to look for similarities in other records of the data file.
Both of these strategies were tempered by not eliminating possible subsidiaries or divisions of
2-12
-------
the parent company in different locations. This conservative approach did lead to some
duplication; however, the result was more often justified by the fact that many responses were
received from divisions or subsidiaries which were not included on the mailing list.
The final mailing list was comprised of 3,610 potential respondents. After the mailing was
completed, a number of additional contacts were added to the mailing list through forwarding
of survey packets to companies that had moved or been sold, and to divisions requesting
additional forms to be sent to other divisions or subsidiary companies. The final number of
"contacts" included in the mailing list data file is 3,802.
2.3 HANDLING OF SURVEY RESPONSES
The data collected from the Consumer and Commercial Products Survey were entered into
®
dBase database files either by the company providing the information or by EPA contractor
®
staff working on the survey project. The SURVEY software, developed using Clipper Version
5.0 and provided on the disk enclosed with the survey questionnaire, was the usual method for
this entry into the data files. This software was also used by the EPA contractor staff when the
responding company provided their response via paper forms. In some cases, the responding
company provided information in a format other than those used in the database files or hard
copy forms provided with the survey questionnaires; these were handled on a case-by-case basis.
All responses to the survey (either on disk or hard copy) were handled as confidential business
information (CBI) using a CBI Plan approved by the document control officer (EPA-DCO) in
the Emission Standards Division (ESD) of the Office of Air Quality Planning and Standards
(OAQPS), U.S. EPA, Research Triangle Park, North Carolina, and only after a notice of EPA's
need for CBI handling was published in the Federal Register (FR Vol. 58, No. 137, 38758, July
20, 1993). The CBI plan is presented in Appendix C.
The discussion in this section covers the methods used to handle an individual response to
the survey. The discussion includes the following steps in the information flow:
1) data entry from a hard copy form,
2) creation of batch files containing response data from several responding companies,
2-13
-------
3) quality assurance (QA) checks performed on the batch data,
®
4) building of an interim SAS data set,
®
5) additional QA checks performed on the SAS data set, and
6) migration of this data to the final data set used to create the summary data tables presented
in this report.
2.3.1 Response Handling and Safeguarding of Confidential Business Information
There were several procedures used in handling responses to the survey that helped assure
protection of CBI as outlined in Appendix C, and for the processing of survey data (Figure 2-4).
When an envelope with a potential response was received, it was held in the CBI-approved file
cabinet until the Document Control Officer (DCO) could open and log the response into the
system designed for that purpose. The paper forms (hard copy) of the survey were placed into
a data entry queue while those responses received on diskette were tested for the presence of
viruses and to assure that the file structure corresponded with those used in the data entry
program. All materials contained within an envelope was placed in a CBI file folder for the
responding company and given a CBI log-in number. If additional submissions were received
for that company and division they were appended to that same file.
The log-in system was built around the database file used to generate the mailing list. That
file was updated to track responses received, survey forms returned as undeliverable, and the
processing of the data in the survey response. Upon log-in, CBI cover sheets and a CBI control
record were added to the file. Hard copy survey responses were keyed into a. separate set of
data files for each file folder, with the data diskette kept with that file. Several data entry
personnel performed these tasks simultaneously on separate personal computers (PC's). Control
of CBI files was maintained by following the standard CBI handling procedures discussed in
Appendix C.
®
All disks with data to be read into SAS were checked for the presence of viruses and for
proper file structure. These checks were conducted regardless of whether the data was entered
by the project staff or by the respondent company. This data was then read into batch files for
2-14
-------
Paper Forms
CBl File
Data Entry
Queue
Batch
Queue
DATA ENTRY
5AS
Queue
Hoidino
BATCH OPERATIONS
Figure 2-4. Consumer and Commercial Product Survey Response handling system.
more efficient transfer of the data into SAS , and more effective quality assurance.
Below is an outline of the responsibilities of the various project staff in handling of the
survey responses and the data derived from those responses. A more detailed discussion is
presented later.
2-15
-------
2.3.1.1 Document Control Officer (DCO)
The Document Control Officer was responsible for the envelopes containing survey
responses from the time they arrived until they were archived at the end of the project. The
DCO opened the envelope with a response and performed the log-in using the software
developed for that purpose. These responsibilities included the following:
1) Update the file based on the information on the respondent's Company Sheet;
2) Enter the Postmark date on the envelope;
3) Enter the date the envelope was received;
4) Assign a CBI Control Number to the file and either attach that number to the disk or to a
blank disk added to the file;
5) Set up a file for the response and all materials associated with that response;
6) Place the file into the proper queue (Data Entry, Batch, or Holding) in the file cabinet or
give to the data review person for determination on anomalous submissions; and J
date a file moves from one queue to the next.
Responses were not processed unless all the above steps could be completed in that same day.
The DCO assured that the responses being handled by data entry were secured (filed and locked)
daily, and that the CBI Log was updated.
2.3.1.2 Data Entry
There were a number of data entry personnel working at any one time. When a file was
checked out from the data entry queue, the data entry person was required to sign the control
record for that file and leave it with the DCO. After completing the data entry, the file was
returned with the data diskette to the DCO for filing in the "batch" queue.
2.3.1.3 Batch Operator
The Batch Operator's duties included checking the format and contents of all data on disks
®
to ensure that it met the requirements for entry to the SAS data set. Disks were checked-out
2-16
-------
from the DCO with the entire file. If any anomalies on the disk were noted that could not be
rectified immediately, the file and disk were returned to the DCO to be stored in the batch queue
until the problem could be resolved. After work with a disk was complete, the file and disk
were returned to the DCO and filed under "Holding." The diskette containing the batch files
®
was held in the SAS queue.
2.3.1.4 SAS Operator
The responsibilities of the SAS Operator were to maintain the SAS data sets, generate
reports, ensure that the information in the data sets passed QA checks, and perform any analyses
as directed. There were a number of tasks that this person was required to perform that was
dictated by the nature of the responses received. The primary data handling function of this
®
person was the reading of batch files into the SAS data sets.
2.3.1.5 Data Reviewer
The task of data review was conducted by the project manager and a data entry person.
Any disks that contained a virus were investigated to determine if an uncontaminated disk or
other version of the data could be acquired. Data formats other than those constructed for the
project were viewed using various software to determine if it could be read into the data files
or in some other way used in the project. Additional review and/or revisions to the data entered
into batches was conducted if there were problems identified by the SAS Operator.
2.3.2 Data Entry
As mentioned previously, the data entry step was performed either by the responding
company where they provided data via computer diskette, or by the EPA contractor staff when
the responding company provided data via hard copy form. In cases where the responding
company provided a response by some alternate format (flat ASCII files, other data formats,
etc.), every attempt was made to enter the data into the proper format. Sometimes, however,
the data was unusable and the file was never entered into the final data set. The following
2-17
-------
discussion covers the data entry step, including the automated QA checks performed by the data
entry program.
The data entry program was included along with the required data files in a self-extracting
file on a 360 kilobyte double-sided, double-density 5.25 inch diskette. This format was used so
®
that a responding company with any IBM compatible, personal computer could use the
software. A company unable to use that format was requested to call the Survey technical
support line for assistance in acquiring another disk. When the INSTALL batch file (included
on the Survey diskette) was executed, it created a directory on the computer's hard drive called
"SURVEY," extracted all the files needed from the floppy disk, and copied these files onto the
hard disk in the SURVEY directory. At that point the software was ready for use.
The software was designed to gather the information from the PRODUCT SHEETS of the
survey questionnaire (Appendix A). The information from the COMPANY SHEET was not
included as part of the electronic data gathering to ensure that a signed, hard copy form was
returned, which gave authentication to the data submittal and a method to contact the respondent
if the data on a data disk was unreadable. Appendix D presents the data dictionaries for the data
files used by the Survey software.
Each data item asked for in the Survey software is presented below along with a discussion
of validity checks performed internally by the software for that item. These validity checks
allow only a certain range of possible entries to be allowed. Many data items were mandatory
and trying to skip over that item or leave a value at zero would not be allowed by the validity
checks.
2.3.2.1 Company Name and Division Name
Company Name information was mandatory and the program would not progress to the next
data entry item nor to the second page of the data entry screen unless some entry was made for
the Company Name. The Division Name was not mandatory but was requested in the event that
one company would be providing information for many divisions, either in one submittal or by
separate submittals from each division. The information in these two fields, along with the
2-18
-------
Product Name and Product Form information, were included in the indexes used with the data
files, and in the linkages between the three data files used for storing the survey data (the
Product file, the RVOC Ingredient file, and the Alternate Contact file).
The Company Name and Division Name entered in the first record were held in the
computer memory and automatically entered when a second record was added to the data files.
This process was continuous, whereby the Company Name and Division Name for the third
record was automatically entered from the information in the second record. In this way, any
change to these fields resulted in the "new" company and division names being read into records
subsequently added to the data files.
2.3.2.2 Product Category Code
The Category Code was limited to those valid codes listed in Attachment A to the survey
questionnaire (Appendix A). The Survey software read this information into an array hi the
computer memory. Any entries attempted in this field were checked against the array for
validity. If the data entry person attempted to enter blanks or an invalid code, the software
presented a series of pop-up menus that allowed a correct category name and code to be chosen
and entered into Product file. Categories were added to the list based on the information
available to estimate the market share of the responding companies, and upon the nature of the
responses collected from the survey. A discussion of these revisions to the list of categories is
included in Section 4.
2.3.2.3 Product Form
There were six valid options for entry as the Product Form: "A" for Aerosols, "P" for
Spray Pump, "L" for Liquid, "G" for Gel, "S" for Solid, or "O" for Other. In the software
version, this was the entry of the appropriate letter while on the hard copy form check-off boxes
were used. If "Other" was chosen, the data entry person was asked to provide a more detailed
description of the product form.
2-19
-------
2.3.2.4 Full Product or Product Group Name
This item was intended to be a unique entry so that no other product could have the same
Company Name, Division Name, Product Form, and Product Name. The check for a valid
entry involved first reading the information for all of the items listed above in the Product file
at start-up into an array in the computer memory; then checking an attempted entry against that
array. As new products were entered, the information was added to the array. If the validity
check was passed, the Product Name was entered into the Product file. If the name was edited,
any records linked to that Product Name in the RVOC Ingredient file or the Alternate Contact
file, and in the validity check array were also changed. Blanks were not allowed for the Product
Name.
Companies were allowed to group information for a number of products with similar
formulations into one product sheet/record. The stipulations for grouping products were that the
products all had to be in the same category (have the same four-digit category code), have the
same product form, and not differ by more than five percent in total RVOC. This was done to
reduce the number of separate product sheets a company would need to submit and thereby
reduce the overall burden of reporting.
2.3.2.5 Stock Keeping Units (SKU's)
Stock Keeping Units were used as a barometer of how many package sizes and grouped
products would be represented by any product sheet. The data entry person was instructed to
enter the number of SKU's represented for the product sheet. It was suggested by the trade
associations as being a useful and easily identifiable number for responding companies. This
information was not mandatory. The default value was one.
2.3.2.6 Major Customer Type
The data entry person was instructed to check off all customer types that applied to the
listed product or product group. The purpose of the question was to determine if the product
or products being used by any one group had differences in formulation that would affect
2-20
-------
emissions of VOC's. This information will be useful to EPA in future determinations of
effective regulation of individual product categories. The software presented this question as
a series of logical (Yes/No) choices; the hard copy form presented check-off boxes. No choices
in this set of logical fields were required.
2.3.2.7 Product Size
The data entry person had the opportunity to answer yes or no to a set of product size
ranges in a manner similar to how Item 6 above was done. There was an additional choice,
however, where the data entry person could change from weight ranges to volume ranges by
pressing the < Alt> and keys simultaneously, which changed the value in the Product
file. Both sets of ranges were presented as check-off boxes on the hard copy form. No choices
for size ranges were required by the software.
2.3.2.8 Net Product Weight Sold in the U.S. for 1990 (Pounds)
The net product weight sold in the U.S. for 1990 (Item 8) was a required entry. The data
entry person was instructed to enter the total number of pounds of the product sold in the U.S.
for the year 1990. The responding companies were given the option to use a fiscal year with
the stipulation that the fiscal year cover the major portion of calendar year 1990. In many cases,
the responding companies requested using a more recent year because the data were more
accessible. This was allowed as long as the responding company documented any divergence
from the survey instructions or any adjustments made to the data in a cover letter.
2.3.2.9 Total Reportable VOC
The total reportable VOC percent (Item 9) was the first data item requested in the second
page of the data entry screens in the Survey software, although on the hard copy form Item 9
and Item 10 were found on the first page. The data entry person was prompted to answer yes
or no as to whether they had formulation data for the product before any information was
requested for the second data entry screen. If the answer to this logical question was yes, the
second data entry screen presented requests for Item 11 A; if the answer was no, the screen
2-21
-------
requested information for Item 1 IB. Regardless of the response to this query, Items 9, 10 and
12 were requested; however, requests for Item 11A and 11B information were mutually
exclusive.
The responding company was instructed to evaluate the list of ingredients for products
containing RVOC, as defined in Attachment B to the survey questionnaire (Appendix A), and
enter the total weight percent that those ingredients represented. The instructions for the survey
questionnaire requested that individual ingredients be listed in Item 11A for those comprising
five percent or greater of the total product weight. This total should have included all RVOC
ingredients and not only those that were greater than five percent and required to be listed in
Item 11 A. A zero entry was allowed in this field.
2.3.2.10 Total Number of VOC Ingredient Compounds
The intention of this data item was for the responding company to provide the total number
of ingredients they have identified as RVOC, regardless of the percent weight that any individual
ingredient may have. The entry in this item should have included all RVOC that comprised the
total RVOC entered in Item 9 and not only those to be listed individually in 11 A. A zero was
allowed if the total RVOC percent entered in Item 9 was zero, and/or no items were previously
entered in Item 11 A.
2.3.2.11 List of RVOC Ingredients or Alternate Contact
Two options were available for the responding company to answer Item 11: the first was
to give the list of all RVOC ingredients that comprise five percent or more of the final product
weight, and the second was to give an alternate contact company that could provide the RVOC
ingredient list. The first option was available as Item 11A and the second as Item 11B on the
Survey questionnaire. These options were both shown on the hard copy form but were mutually
exclusive in the Survey software. The only exception to this exclusivity were the requests for
information on methylene chloride (MeCl) and 1,1,1-tricnloroethane (TCA). The request for
what information was done as described for Item 9 in Section 2.3.2.9 above.
2-22
-------
An array was created in the computer's memory after a response to Item 10 was registered
for storing RVOC ingredient information. After the entry of RVOC ingredients was complete,
the data was transferred as separate records in the RVOC Ingredient file for individual
ingredients. If data was previously entered in the RVOC Ingredient file for the current
Company/Division/Product/Form combination, that information was read into the array for
editing and replaced afterward. Internal QA checks on the ingredient array included confirming
that the number of ingredients listed did not exceed the number entered in Item 10, and that the
total RVOC percent in the ingredient list did not exceed the value reported in Item 9. If either
of these checks were failed, the data entry person was prompted to either change the entries to
the ingredient list, or to Item 9 or Item 10 as appropriate. Another QA check was to determine
the average weight percent of the remaining unlisted ingredients (average of "Other"). This
check was performed as described in the formula below:
Average "Other" RVOC - (Item 9 Entry ~ Total RVOC Usted)
(Item 10 Entry - Number of RVOC Listed)
Until the QA checks were passed, data entry was halted at the RVOC ingredient list. The
entries for MeCl and TCA were not included in the RVOC ingredient list, however, the QA
checks were failed if values for these items plus those for the RVOC list were greater than 100
percent.
If the information for an alternate contact was being entered, the information was stored in
a discrete record in the Alternate Contact file for the Company/Division/Product/Form
combination from the Product file. The request for MeCl and TCA information was also
included if Item 11B information was being entered.
2.3.2.12 Claims of Confidentiality
Confidentiality claims could be made in four logical fields that were synonymous with the
check-off boxes in the Survey questionnaire: claiming sales, the total RVOC percent, the
ingredient list, and/or the ingredient weight percents as Confidential Business Information. This
2-23
-------
information was not required by any validity checks. This information was included at the
request of OMB, even though all responses to the survey were to be treated as confidential.
2.3.3 Batch Files
Data for single companies were provided on separate diskettes either by the responding
company or through entry of the data from paper forms by the EPA contractor staff. These
diskettes were held in CBI files with all other materials sent with the data submissions. The
large number of diskettes, some listing only a few products were then combined with the data
from several other companies into "batch" files so that QA analysis could be performed more
®
effectively, the data could be read into the SAS data set more efficiently, and the original data
files would remain unchanged. The following discussion gives the details of how the batch files
were compiled, the procedures for performing QA analysis, the procedures for changing the data
failing the QA analysis, and how those changes were documented. The discussion of how the
®
data were transferred to the SAS data sets is presented in Section 2.3.4.
2.3.3.1 Compilation of Batch Files
As mentioned above, the data from the three data files for individual companies were
combined into similar files that contained data for several companies. There were two additional
fields added to these batch files that contained the disk identification number (DISK_ID) and the
identification number for the batch (BATCHJGD). Disk numbers were assigned to data diskettes
at the time the survey responses were opened and logged into the CBI tracking system used for
the project (detailed in the CBI Plan in Appendix C). Batch numbers were assigned sequentially
through the software that read individual data diskettes into the batch files. The batch files were
named such that the batch number was incorporated into the batch file name. Product file data
from the individual data disks was read into the PROnnn.DBF file, where nnn refers to the batch
number. In a similar manner, the data from the RVOC Ingredient file and Alternate Contact
file were read into the RVOnnn.DBF and PROnnn.DBF files, respectively.
Two checks were performed on the data diskettes, however, before any data was read into
batch files. The first data check was for the presence of computer viruses, and the second was
2-24
-------
a check for readability and proper file format. Many companies submitted data in a format other
than the dBase format provided for use with the Survey software, which would cause problems
when attempting to read the file into a batch. If a problem occurred during the generation of
a batch and the batch files were unusable, that batch number was abandoned and the diskette that
caused the problems was reviewed and revised. Revisions were performed on copies of the
original data diskettes that were kept unchanged in the file for that data submission.
Batches were created by reading the data files from individual CBI file folders, all from the
same type of disk (either 3.5" or 5.25") onto a disk of the other size. This avoided writing data
to the hard drive of the computer running the batch creation software, and the need to then erase
that file from the hard drive and/or lock-up the computer. The disks containing the batch files
could then be stored along with the other CBI file folders as specified in the CBI plan (Appendix
C). The data read into a batch was not ordered in any way other than availability at the time
that the batch was being created.
2.3.3.2 Quality Assurance Checks for Batch Files
®
After the data was transferred to batch files, a dBase program (QA_RVOC) was executed
that performed some of the QA and validity checks used in the data entry program and wrote
exceptions to a series of text files for later review. This program performed six checks on the
data in the Product file and the RVOC Ingredient file:
1) If the category code was missing in the Product file;
2) If there were any duplications for a Company/Division/Product/Form in the Product file;
3) If there were no matching records in the RVOC Ingredient file for records found in the
Product file;
4) If the RVOC percent listed in the Product file matched the total RVOC for all ingredients
listed in the RVOC Ingredient file for that Company/Division/Product/Form combination;
5) That the pounds sold in 1990 listed in the Product file was greater than 1; and
6) If there were records in the RVOC Ingredient file with no match in the Product file.
2-25
-------
If problems requiring revisions to the data in the batch files were found, the original batch
was not edited directly. A new batch number was assigned by adding 1000 to the original batch
number and the files (PROnnn, RVOnnn, and ALTnnn) were copied to a new diskette with the
new batch number incorporated into the file name. Therefore, if the original batch number was
54, the original product data for the batch was contained in the PRO54.DBF file; and the batch
file that was created for revisions to the data was PRO1054.DBF, with a batch number of 1054.
If additional revisions to the data were determined to be necessary, the next cycle of revisions
would be found in the 2054 batch file. Whenever a change to a particular record was made, it
was documented on a form that was filed with the other documents in the CBI file for that
companies data submission, in the log book used to track the creation of batches, and in a data
review log book. This redundancy allows the tracking of all changes to the data if ever an audit
of data from a particular company or product needs to be performed in the future.
2.3.3.3 Revisions to Batch files
®
Two additional dBase programs were written to correct the most common errors found in
the batch files. These programs were executed on the "1000" series of batches because they
performed some automated additions, deletions, and revisions to data in the batch files. These
programs documented these changes in reports particular to the program being executed.
The first program (DUP_RVOC) marked any duplicated records for deletion in the RVOC
ingredient file (RVOnnnn.DBF) for a particular Company/Division/Product/Form combination
when the ingredient name was identical for a second record. This type of error seemed to be
an artifact of incomplete deletion of records if the data entry person made additions and/or
deletions to the list of RVOC ingredients in Item 11A in the SURVEY software. The records
marked for deletion were not removed permanently from the data file until after the person
performing the data review had checked the data file and verified that deletion was indeed
warranted.
The second program (FIX_MISS) was written to add missing RVOC ingredient records to
the RVOnnnn.DBF file if none existed for a record found in the PROnnn.DBF file. The first
step was to determine the average RVOC percent by dividing the total RVOC percent in the
2-26
-------
RVOC field by the number of RVOC ingredients as entered in the PROnnnn.DBF file for the
Company /Division/Product/Form combination. If that value was less than five percent, a record
in the RVOnnnn.DBF file was created with the appropriate Company/Division/Product/Form
data entered, "Other" listed as the ingredient name, 0 as the RVOC Number (Appendix D) and
the total RVOC from the Product file entered as the RVOC percent. If the average value was
greater than five percent, two records were added for the Company/Division/Product/Form listed
in the Product file, one that listed "Other" as the ingredient name with 0 as the RVOC percent,
and the second listed as "Unknown" with the total RVOC percent from the Product file entered
as the RVOC percent.
Additional revisions to the batch were performed by the person performing the data review
as needed and based on the reports generated by the QA_RVOC program. These changes were
documented on standardized sheets, one for each disk number, so that these error sheets could
be filed with the other original information for the data submission.
2.3.4 SAS Data Sets
®
There were three steps involved in the development of the SAS data sets used to generate
the table presented in Section 2.4. The first step involved the reading of the data in the batch
files into a temporary data set and performing rudimentary, automated QA analysis on the data
from that batch. The second step involved the appending of the data in the temporary data set
into a set of "validation" files upon which additional QA was performed. Data were read into
the final master data set after passing all validation checks and were analyzed with the results
used to create the table shown in Section 2.4. No revisions to the data were performed on the
master data sets. The discussion below gives additional detail on these three steps.
2.3.4.1 Reading of Batch Files
®
The SAS data sets were created from the batch files that were compiled as described in
Section 2.3.3 above. As batch files became available, they were read individually into a set of
temporary data sets, where a number of QA checks were performed and had to be passed before
2-27
-------
the data for the batch were added to a validation file. The QA checks performed on the
temporary data sets included those listed below.
1) Uniqueness
• only one product record for a combination of
Company
Division
Product Name
Product Form
• only one RVOC record per product (as defined above) with a unique RVOC
number
• only one alternate contact per product (as defined above)
2) Checks of entered CAS registry numbers in the RVOC Ingredient file for validity.
3) Checks that the Product Form entry is valid: either A, P, L, G, S, or O.
4) Checks for a valid Category Code as specified in the Survey.
5) Checks for a valid record in the Product file for each listed RVOC ingredient in the RVOC
file.
6) Check for a mandatory RVOC record when the number of RVOC ingredients is greater than
zero, the average RVOC percent is greater than five, and there is no Alternate record for
the product.
7) Check that the total RVOC percent listed in the RVOC list for a product matches the RVOC
percent entered in the Product file.
Whenever a problem with a batch was encountered that did not allow it to be read into the
®
SAS data sets, the batch was revised by the data review person as a "1000" series batch as
®
previously described. After the data from a batch was read into SAS , the diskette holding that
®
batch data was filed in holding along with all the original data submission files. All the SAS
®
data sets were stored on a Bernoulli disk drive (a high capacity, removable storage media) that
was removed from the computer running SAS for OS/2 , and placed in a locked CBI file
cabinet when not in use.
2-28
-------
2.3.4.2 Validation Data Sets
There were three validation files used, one for each of the three data files used in the
Survey software and batch files: PRVALID for the Product data, RWALID for the RVOC
ingredient data, and ALVALID for the Alternate contact data. The validation files were created
to have various QA checks performed upon them. Their contents were compared to those of
batch files being read into the temporary data sets previously described. These comparisons
®
ensured that no duplicated data were read into the SAS data sets.
When revisions to the data in a batch file were required after the batch data was read into
a validation file, the 1000 series batch file was generated by the data review person, the SAS
operator erased the old batch data, and appended the 1000 series batch data to the validation file.
If this process were repeated, the second revision to the batch would be in the "2000" series,
and that data would replace the 1000 series data in the data sets. Simple editorial changes to
®
data were performed, when feasible, directly on the SAS data sets and documented.
2.3.4.3 Master Data Sets
All revisions were conducted on the batch files and validation files. After all QA tests were
satisfactorily passed and the data were believed to be valid, the master data sets were created.
No revisions or editorial changes were enacted upon the master data sets. The analyses used
to create the table shown in Section 2.4 were performed on the master data sets.
2.4 RESULTS OF THE SURVEY
An overview of the response rate to the survey and a summary of the survey responses are
discussed in the following section. The total number of questionnaires mailed was 3,610;
however, a number of companies had responses from more than one division and some
companies forwarded copies to companies that were not on the original mailing list. These
additional sources of survey responses brought the total number of questionnaires being tracked
to 3,802.
2-29
-------
2.4.1 Response Rate
The response rate is illustrated graphically in Figure 2-5. Out of the 3,802 survey
questionnaires accounted for in the tracking system, only 1,283 questionnaires (approximately
one-third) are included in the No Response category. An additional 352 (9.3 percent) were
returned as undeliverable by the U.S. Postal Service. Approximately one-quarter of the survey
questionnaires were responded to by the company indicating that they did not manufacture,
distribute, or sell any of the products found in one of the product categories (Company Sheet
Only). Responses via either product sheets or data diskettes were provided by the remaining
companies. Thirty responses (0.8 percent) were obviously inappropriate and could not be used
in the survey without lengthy follow-up. The remaining responses were complete with product
information. An additional 35 completed responses were provided for a company by a third
party (e.g.,consulting firm, custom packager, etc.). There were, therefore, 1,173 complete
responses (including product formulation data) from companies manufacturing consumer and
commercial products.
Response
^ | by A 11ernate
I Co Sheet On Iv
964 / / D 9%
25 4%
/ Va i id Response
Figure 2-5. Distribution of survey
questionnaires
2-30
-------
Not all of the 1,173 complete responses, however, could be used in the final analysis. A
number of problems were encountered when the information was attempted to be entered into
the data entry system from both hard copy product sheets and data diskette submissions. In all
cases, attempts were made to contact the company when problems were identified, but these
attempts were not always successful. In many cases, information vital to the project was not
provided, such as the pounds of product sold in 1990. In a number of other cases, products
were inappropriately grouped (i.e., products from more than one category were reported
together). These problems and a number of others were not resolved and 96 such responses
were not included in the final data set. Therefore, the final data set included 1,077 complete
responses.
2.4.2 Information Obtained from the Survey
The compiled survey data are presented in several ways. The following sections describe
the outputs of the survey which have been generated to date. Other analyses of the data may
be generated by the EPA as the need arises.
2.4.2.1 Total RVOC Emissions by Individual Product Category
The primary purpose of the survey was to develop 1990 estimates of VOC emissions from
traditional consumer products, including commercial and institutional uses. Table 2-1, the
principal output of the survey, presents RVOC emission data for each product category
surveyed. Specifically, this table provides information for each category on (1) number of
products reported; (2) product sales reported (tons); (3) RVOC content reported (tons); (4)
estimated market coverage (i.e. what percentage of the market for a particular category was
captured by the survey); (5) adjusted product sales (reported sales scaled up based on market
coverage); (6) adjusted RVOC content (reported content scaled up based on market coverage);
(7) RVOC content emitted (i.e., the percentage of the product's RVOC content which enters the
ambient air after fate adjustments); (8) tons of RVOC emitted nationwide; (9) pounds of RVOC
emitted per 10,000 persons; and (10) RVOC emissions in ozone nonattainment areas (tons).
2-31
-------
Items 1, 2, and 3 were obtained directly from the survey responses. Item 4 (market
coverage) was estimated based on information furnished by CSMA, CTFA, and SDA following
their reviews of the initial results of the survey and lists of companies responding in each
19^
category 1'Z>J. Adjusted product sales (item 5) and adjusted RVOC content (item 6) were
obtained by dividing reported sales and reported RVOC content by the estimated market
coverage percentage. The adjustments were made in the following manner:
= AdjustedTonnage
Estimated Market Coverage
Percent RVOC emitted was obtained by multiplying the RVOC content by a fate adjustment
factor. The adjustment factor ranged from 1 percent to 100 percent and was based on
information supplied by CSMA, CTFA, and SDA*'^'"'' ' . The fate adjustments were made
to account for products which enter the wastewater stream and subsequently biodegrade rather
than being emitted to the air. A fate adjustment factor of 100 percent means that all the RVOC
content of the product is emitted to the air. Conversely, an adjustment factor of 1 means that
only 1 percent of the RVOC's are emitted.
Probably the most useful of the items in Table 2-1 are items 9 and 10. Data on emissions
in ozone nonattainment areas will be used by the EPA in determining which categories will be
targeted for regulation under §183(e). The per capita emission factors (in this case, pounds of
RVOC emitted per 10,000 persons) will enable the EPA, States, and local environmental
agencies to develop emission estimates for specific combinations of products based on
population. This is by far the most comprehensive and accurate set of emission factors ever
developed for consumer products.
2-32
-------
J
3
H
G .°
o 3 53 "w*
'i g 2 §
1 § %
ts^at
u c ^S -^
O S ^ e
^ § l£
* U jg "
§ « 1 ^
•5-300
2 t« R-H
ct OT ^ C-
^£ T3 'D
ll 1
•3 g £
SB- 3
e- oi g.
Product Category
NAL CARE PRODUCT!
ire Products
O u
Cfl W
BC .!=
S£
E
r-
—
—
0
*
1
en
in
ON
O
vO
en
ON
ts
ts
E
u
c
u
I
•a
§
£
1
CO
o
o
o
o
8
o
VO
in
ON
O
vo
Tf
Brilliantines
ts
o
t-
ts
Tf
n
o
S
•t
ON
(-
8
ON
I—
•*
en
oo
>n
ON
O
t-
ts
ON
t—
ON
Curl revitalizers
S
ts
ON
ts
TJ-
>n
o
i
in
tn
vo
vo
in
in
ON
en
vO
•"t_
i
00
•*
>n
en
V
£
tn
O
£
t^
0
ts
o
Tt-
O
vo
en
t~-
ts
en^
en
»n
ON
t
en
£
en
ts
1
.S-
&
%
£
oo
0
f~
•*
VO
oo
r~
O
00
t—
r~
1
VO*
in
ON
ON
en
t-
p~
en
vO
ON
Dyes - Temporary
S
oo
t-
ON
Tf
0
en
•<»•
a
s
R
S
00
vo
en
>n
ON"
t~-
^•*
•n
ON
m
vo"
in
O
r~
r-
Finishing spritzes
T— i
ts
O
•*•
R
•*
1
m
ON
•>t
00
en
en
^^
VO
Grooming creams
ts
en
3
»
—
i?
ts"
R
ts
•*
ts
ts
r-
•*
oC
^^
in
ON
g
es"
00
ON
•
-------
I
S
§e
- 1 8 f
w z ^
M S
1 8 £ ^
;ss § §
O ^ cfl ^
8jlf
*c3 J^
T3 •<_»
£ W C ""wi*
&U
"8 « « ^
» » o c
^ r^ art O
•5"1 £M C/5 H*
«i
S!
en
00
00
ft
en
ts
g
es
OS
en
en"
8
Os
fS
00
s
so
8
00
en
vo
8
vO
m
1
a1
1
s
00
•"I-
tN
«
t--"
R
Os
so
^
Os
Os
in
Os
VO
Os
r~
SO
00
5
CA
1
oo
»n
ts
0
o
o
8
o
Os
es
m
Os
o
00
es
-
Thickeners
S
ts
ts
^^
tn
VO
oo
en
8
SO
00
en
SO
in
_^
in
Os
r-
VD
en
00
en
—
Cfl
o
1
ts
z
^
in
8
i—
i/-)
S
ts
m
Os
00
00
00
in
ts
in
en
1
0.
a
o
Wri
'5
1
O
%
*•-
1-H
0
oo
""*
»— i
i
fS
^\
f)
ON
OO
I
«s
>n
o
•s.
GO
V0
f)
^^
i
1
£
ed
U
(U
M
ants and Antiperspirant
•§
°
00
en
m
r~-
oo"
8
so
Os
00
oo
en
C*^
8
so
OS
00,
oo"
rt
SO
en
ts
00
ts
Underarm deodorants
5
§
r-j
.«,.
ir*
'•O
IN
in
8
IN
in
i—
a
in
m
,C
en
8
(S
in
in
in
t-
tn
oo
CO
01
Unuerarm aniiperspirani
S
2
^.
~*
vO
8
VO
f-^
in
00
8
r-
VD
VD
r-
so
Foot deodorant sprays
1
8
en
en
8
•t
en
Os
SO
m
Os
ts
en
so
SO
-
1
Feminine hygiene deodo
s
m
t-
•n
oo
8
in
00
en
VO
en
X
00
in
en
in
S
Other deod/antiperspiran
ts
|
*•*
in
>— i
I
»5
^f
&\
?J
S
-------
I
S
en £
O .2 <2 /n
ON
r-
*^
fS
vO
es
-
t
8
U.
en
in
en
>n
P
oo
8
00
m
m
r-J
S
t^
00
ON
00
en
rs
S
13
1
I
a
o
s
1
o
£
^
1
^
S
«M
o
^-^
1— 1
r-
^i.
1
S;
2
All Powder Products
V)
CO
(S
-------
w 6
'« ts S g
'§ J -a 2 £
t* u .s c
8jl£
*6 Jfe
3o|i?
i>n
T3 *-
Q G r* <^
w 3 O C
•o* *- w ^~v
1 -s M^
0 1 o f
55|£
-
M W ^ W
"^ "—5 O o
O _5 ON t—i
*5 5 E
CU Q^ ^-"
Product Category
are Products
U
S
ON
O
S
en
Cuticle softeners
S
in
o
o
o
8
o
vO
ON
O
en
in
.-
Manicure preparations
1
v>
o
t--
VO
8
VO
VO
00
en
ON
en
cs
00
^
o
*
U)
1
O.
£
a
o
c
1
O
V)
i
•n
^
P
00
2
ON
00
vH
s
-*
06"
oo
o
•s
s
1
I
ed
U
Z
and Body Treatments
_
1
S
en
en
51
VO
«n
8
VO
>n
S
IT)
ON
ON
00
t~
~*
0
Astringents
§
00
rs
o
en
o
_
o
ON
S
S
VO
00
en
vo
S
X
u
"o
o
U
i
-
_
::
8
_H
•"•
S
U")
ON
0
*
*-• •
o
£
•f
•u
i
I
1
f:
o
§
8
00
cs
oo
•<*•
>n
ON
es
cs
en
r~
ON
es
00
Foundations & fixatives
i
CS
o
en
8
en
en
in
ON
en
•n
ON
ON
_o
.S-
i
t—
00
£
8
«0
O\
&
en
—
>n
ON
S
ON
vO
en
vo
en
Moisturizers
i
~
„
00
8
00
<—
00
ON
t-
.n
cs
00
Skin lighteners
t-
en
-------
18
1
I
I
I
% .3
°w •J3 S S
•a a J? f2
a o < c
w z
si -«
o o c *o
||1|
u "8 r>
VO
VC
ts
3
8
2
**!
ON
VO
en
o
-— •
,j
Skin protectants
VO
VO
—
O
-1
o
vO
vc
VO
vO
00
00
*M
(S
Depilatories
t~
VO
t-
CN
?5
8
tM
00
00
ON
oo
fS
00
g
c
o
i
cS
O-
8
o.
60
C
00
VO
ON
(S
Tf
5
8
"*
ON
ON
5
8
*—
—
c
'o
1
"5
S
1
ON
vO
r~
t—
o
§
8
ON
en
en
CS
(S
^^
00
Sunscreens
p.
$
-"
S
§
8
S
00
en
V)
en
CN
i
fN
-
S
2
's
s
Other factal/bod>
a
VO
ve
**•>
^
J5
*
|
1
S
00
9\
\£)
**
0
*^
•v
Os
i
42
S
b
S
H
All Facial/Body
are Products
U
S~*
en
ON
S
«
00
o.
2
oC
"^
en
en
es
es
00
O
en
8
Mouthwashes
— «
R
2
M
8
V)
i
ON
VO
ON
vO
00
en
00
00
>/•)
o
Breath fresheners
ts
®
«n
*»
i—
V)
§
tn
ON*
oo
en
00
*M
O
v
-------
1
§
u
8 -S
O 3
ia z
.8 2
u jL'
(2
.S b
"8 3
•a o /—s
1111
*•* W
II
*l I
8. I
roduct Category
£
ol
IS
ON
aq
OS
Os
ts
VO
uoride rin
vn
o
CJ
00
Os
tal care prod
g
Os
0.
1
9
E
in
S
55
Os
Os
8
Os
8
S
8
Os
«*»
1
S?
OO
ve1
1C
€
1
w
I
I
U
1
us
VO
•o
§
8
Os
I
8
Os
8
V)
PI
o
en
Rubbing alcoho
Os
r
eo
SI
i
VI
8
**!
oT
oo
Os
•t
oo
VO
-------
1
W £J
o 3 s "w*
w 'rt 8 S
1|
§ ° I §
» 5 « c
fl'fi
T3 O /^~v
5 u W) c
Pit
"lit
1 § !•&
1 all
Q, -
l/D T3 ^
1 1 1
1 1 i
D. pi Z
S)
V
3
tj
1
CN
c»
SEHOLD PRODUC
Surface Cleaners
1 s
ffi a:
1
;
?
>n
o
u-i
00
ts
00
VO
§
s
(S
s
VD
VO
f-
r-
IA
U-
§
General purpose c
o
fS
cs
Ov
r-
S
i
>n
8
-
^>
wT
00
r-
cs
$
00
00
2
(S
vo"
£
g
o
s
(N
§
—
S
s
8
(S
oo
^
1
oo
C4
S
PI
t—
2
Oven cleaners
S
cs
s
s
g
0
^1
r^
ts
o
>D
^
vO
U^
(S
Tt
VD
M
CS
«
e
V
a
"o
1
1
1
0
o\
ts
vo
ro
g
S
M
?2
00
tf)
i
5
o_
««
oo
S
8
^
I-N
z
c
ca
U
•s
OJ
Iry Products
i
j
s
-
i
—
In
vO
Q
O
00^
rf
00
g
5
?
oC
00
ri
(S
Detergents
S
0
0
o
~
-
o
d
s
-
"1
2
i
s
-------
f
I
a
w C
8 3 a ^
11 U
w z
1 § S?
1 2 § §
llrfl
* J.sb
811?
^ ^ "S Ai
&< 6 w~
l^lf
"O* r^ O h*
< u
U O „ O
w 3 o e
P "* £
< B-
ill!
0||?
> o &•{£
i.?*
J3 §•£
O D- 3
£ c2 5.
Product Category
o
o
o
*"*
en
cs
R
^r
en
t-
t-
VO
tn
^
Presoaks
en
0
IT!
£
1
CN
CN
tn"
S
vo'
oo
tn
r-
tN
oo
tn
oo
S
0
1
I
1
(X
cs
^
—
2
,
tn
ts
00
IN
O\
|