& EPA
             United States
             Environmental Protection
             Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 2771 1
May 1998
             Air
             Economic  Impact Analysis  for the  Proposed
             National  Emission Standard  for Hazardous Air
             Pollutants from  the  Production of Wet-Forn^ed
             Fiberglass Mat
                          U.S. Environmental Protection Agency
                          Region 5, Library (PL-12J)
                          77 West Jackson Boulevard, 12th Floor
                          Chicago.il  60604-3590

-------
     This report has been reviewed by the Emission Standards
Division of the Office of Air Quality Planning and Standards,
EPA,  and approved for publication.  Mention of trade names or
commercial products is not intended to constitute endorsement
or recommendation for use. Copies of this report are available
through the Library Services Office (MD-35),  U.S.
Environmental Protection Agency,  Research Triangle Park, NC
27711, or from the National Technical Information Services,
5285 Port Royal Road,  Springfield, VA 22161.

-------
                                    CONTENTS

Section                                                                        Page

   1      Introduction	1-1

   2      The Supply Side	2-1
          2.1    Types of Products	2-1
          2.2    Production Process	2-2
          2.3    Costs of Production	2-3

   3      The Demand Side	3-1
          3.1    Product Characteristics  	3-1
          3.2    Uses and Consumers 	3-1
          3.3    Substitution Possibilities in Consumption	3-2

   4      Industry Organization	4-1
          4.1    Market Structure  	4-1
          4.2    Manufacturing Facilities  	4-3
          4.3    Companies	4-3
                4.3.1  Size Distribution  	4-5
                4.3.2  Vertical Integration  	4-5

   5      Markets  	5-1
          5.1    Production and Shipment Data  	5-1
          5.2    Foreign Trade	5-1

   6      Economic Impact and Small Business Analysis  	6-1
          6.1    Identifying Small Businesses	6-1
          6.2    Estimated Incremental Emission Control Costs  	6-1
          6.3    Economic Impact	6-4

          References  	R-l
                                        in

-------
                               LIST OF FIGURES




Number                                                                    Page




   3-1    Shares of roll roofing and strip shingles by base type:  1992 	3-3
                                       IV

-------
                                LIST OF TABLES

Number                                                                     Page

   2-1    Production Costs for SIC 3229, Pressed and Blown Glass, N.E.C.: 1992 .... 2-4

   4-1    Market Concentration Measures for SIC 3229, Pressed and Blown
          Glass, N.E.C.: 1992	4-2
   4-2    Relative Size of SIC 3229 versus SIC 3229325: 1992	4-3
   4-3    Sales and Employment of Wet-Formed Fiberglass Mat Facilities: 1997	4-4
   4-4    Sales and Employment Data for Wet-Formed Fiberglass Mat Companies:
          1995 	4-6

   5-1    Value of Product Shipments: 1992-1995  	5-2
   5-2    U.S. Foreign Trade for Fiberglass Mats:  1992-1996 	5-3

   6-1    Employment Data for Affected Companies: 1995 	6-2
   6-2    Estimated Facility-Level Compliance Costs	6-3
   6-3    Ratio of Annual Compliance Cost to Baseline Sales for Wet-Formed
          Fiberglass Manufacturing Companies: 1995 	6-4

-------
                                    SECTION 1
                                 INTRODUCTION
       The U.S. Environmental Protection Agency's (EPA's) Office of Air Quality Planning
and Standards (OAQPS) is developing an air pollution regulation for reducing emissions
during the production of wet-formed fiberglass mats. A National Emission Standard for
Hazardous Air Pollutants (NESHAP) for this industry is scheduled to be proposed in 1998.
The Innovative Strategies and Economics Group (ISEG) of the EPA is responsible for
providing analysis and supporting documents that describe the likely economic impacts of the
proposed standard on directly and indirectly affected facilities. This report provides
information in support of the economic impact analysis (El A) of the proposed NESHAP.
       Fiberglass mat production is classified under Standard Industrial Classification (SIC)
3229325, which is a subset of SIC 3229, Pressed and Blown Glass, Not Elsewhere Classified
(N.E.C).  To the extent possible, other pressed and blown glass products will not be covered
in this report, except in cases where economic data include these products, and more
disaggregated data are unavailable.
       The organization and main findings of this report are as follows.  Section 2 provides a
detailed description of the production of fiberglass mats. The section describes how the
majority of hazardous air pollutants (HAPs) from wet-formed fiberglass manufacturing
facilities are associated with two emission points: the drying and curing oven exhaust, and
the binder application vacuum exhaust (ERG, 1997). The proposed NESHAP will target
emissions from the drying and curing oven exhaust.  The demand side of the industry is the
focus of Section 3.  This section identifies properties of fiberglass that make it a desirable
substrate for asphalt roofing products. Section 4 discusses the organization and market
structure of this industry. This section identifies eight companies operating 14 fiberglass mat
manufacturing facilities that could be affected by the proposed NESHAP.  Section 5 provides
domestic and foreign market statistics for the fiberglass mat industry.  It finds that the value
of shipments for this industry grew steadily from 1992 through 1995.  A small business
analysis is explained in Section 6. The total incremental compliance cost for emission
control is estimated to be $1.564 million per year.  The average incremental cost for facilities
that will have to make improvements in their operations to meet the proposed emission

                                         1-1

-------
standard is $0.223 million per year. However, the Agency estimates that incremental costs
are less than 1 percent of sales for all affected companies.  Although one of the eight affected
companies qualifies for the small business status because it employs fewer than 750
employees, the analysis suggests that the proposed NESHAP will not have a significant
economic impact on this company.
                                        1-2

-------
                                    SECTION 2
                                THE SUPPLY SIDE
       In this section, the supply side of the wet-formed fiberglass mat industry is discussed
 in terms of the types of products, production process, and the costs of production.

 2.1    Types of Products

       The U.S. Department of Commerce classifies fiberglass mat production as part of
 SIC 3229, Pressed and Blown Glass, N.E.C. This industry includes manufacturers of textile
 glass fibers as well as establishments that manufacture machine-made or handmade pressed
 and blown glass and glassware.  Companies in SIC 3229 produce several types of products:

       •  Glass fiber, textile type (SIC 32293)—this includes mats, yarn strand, staple yarn,
          sliver, roving, chopped strand, and milled fiber.  Wet-formed fiberglass mats are a
          subset of this category and are classified under SIC 3229325.  These mats consist
          of a nonwoven material made from chopped rovings that are bound together with
          a chemical binder for use as reinforcement.
       •  Machine-made table, kitchen, art, and novelty glassware (SIC 32295).
       •  Machine-made lighting, automotive, and electronic glassware (SIC 32296).
       •  All other machine-made glassware (SIC 32297)—this includes technical and
          scientific glassware, glass blocks, and lens banks).
       •  Handmade pressed and blown glassware (SIC 32298).
       •  Pressed and blown glass, N.E.C. (SIC 32290).
       The proposed rule only targets the production of fiberglass mats (SIC 3229325). To
the extent possible, other pressed and blown glass products will not be covered in this report,
except where economic data include these products, and more disaggregated data are
unavailable.
                                        2-1

-------
 2.2    Production Process

        The fiberglass mat production process begins with the mixing of short (1 inch) glass
 fibers with water and emulsifiers in large (several thousand gallon) mixing vats to form an
 aqueous slurry of fibers and water. The emulsifiers are added to help disperse the fibers and
 prevent clumping. The  slurry is then pumped to another large vat that acts as a surge tank.
 The slurry is pumped from the surge tank to a third vat that serves as the supply tank for the
 mat-forming machine.

       The mat-forming machine consists of a slurry dispenser and moving wire screen belt.
 The wire screen belt carries the glass fiber mat throughout the production process. The glass
 fiber slurry is dispensed from a slot onto the screen in a uniform curtain. The properties of
 the mat (i.e., density and fiber orientation) can be controlled at this point in the process by
 controlling the width of the slot and the rate at which the slurry is dispensed onto the screen.
 After the slurry is dispensed onto the belt, the screen passes over a vacuum slot into which
 the excess water and emulsion are drawn. This leaves only a layer of fibers on the screen.
 The water and emulsion are then recycled back to the slurry preparation vats.

       Next, the fiber mat passes under a binder dispenser.  An aqueous solution of urea-
 formaldehyde  binder is dispensed  from a slot or a curtain coater onto the mat of fibers in a
 uniform curtain. Just after the binder is dispensed onto the mat, the screen passes over
 another vacuum where the excess binder solution is withdrawn. The excess binder is
 recycled back to the binder storage tanks.

       The mat of fibers and binder passes into a drying and curing oven.  This is a multiple-
 stage oven that uses forced air to carry away excess moisture. In the first stage, the remaining
 moisture is driven from the binder. This causes the binder to migrate to the points where the
 glass fibers cross each other. In the second and third stages of the oven, the binder is cured
 and hardened.  The ovens typically operate at several hundred degrees Fahrenheit (°F). A
 fraction of the  air inside  the oven is recirculated within the oven before  it is vented to an
 exhaust system to remove moisture from the oven system.  After leaving the oven, the
 finished mat is wound into large rolls and prepared for shipment.

       The majority  of HAP emissions from wet-formed fiberglass manufacturing facilities
are associated with two emission points:  the drying and curing oven exhaust, and the binder
application vacuum exhaust (ERG, 1997). The proposed NESHAP will target emissions
from the drying and curing oven exhaust.
                                         2-2

-------
2.3    Costs of Production
       The U.S. Department of Commerce reports cost of production statistics in the 7992
Census of Manufactures. Although fiberglass mats can be separated out by the seven-digit
product code 3229325, the following section reports data for SIC 3229 because more
disaggregated data are not available. However, in 1992 fiberglass mats represented only
5 percent of total value of product shipments reported for SIC 3229 (U.S. Department of
Commerce, 1995a). Therefore, these statistics should be viewed in this context.

       The three primary costs of production for the wet-formed fiberglass mat industry are
capital expenditures; labor expenses; and cost of materials, including the cost of fiberglass
and binding materials.  Table 2-1 provides a detailed breakdown for each of these cost
categories for SIC 3229, Pressed and Blown Glass, N.E.C. for 1992. Labor expenses and
cost of materials together accounted for the bulk of the total cost of production in 1992.
Labor expenses were $1,350.5  million (42.9 percent of total), while the cost of materials was
$1,351 million (42.9 percent of total). In contrast, capital expenditures totaled $448.5 million
and accounted for 14.2 percent of total cost.
                                         2-3

-------
Table 2-1.  Production Costs for SIC 3229, Pressed and Blown Glass, N.E.C.:  1992

Total Cost of Production
Total Capital Expenditures
Building and other structures
New capital expenditures
Used capital expenditures
Retirements
Machinery and equipment
New capital expenditures
Used capital expenditures
Retirements
Total Labor Expenditures
Annual payroll
Fringe benefits
Total Cost of Materials
Materials
Fuels
Electricity
Other
Value
($106)
3,150.0
448.5
34.5
28.9
0.2
5.4
414.0
315.0
4.6
94.4
1,350.5
1,024.6
325.9
1,351.0
929.8
146.2
138.3
136.7
Share of Total Cost of
Production
(%)
100.0%
14.2%
1.1%
0.9%
0.0%
0.2%
13.1%
10.0%
0.1%
3.0%
42.9%
32.5%
10.3%
42.9%
29.5%
4.6%
4.4%
4.3%
Source:  U.S. Department of Commerce, Bureau of the Census.  1995a. 1992 Census of Manufactures,
        Industry Series—Glass Products. MC92-I-32A.  Washington, DC:  Government Printing Office.
                                           2-4

-------
                                     SECTION 3

                                THE DEMAND SIDE
       Fiberglass mats have several desirable properties that make them well suited as
substrates for asphalt roofing products.  This section characterizes the demand side of the
market for fiberglass mats in terms of their characteristics, their uses and consumers, and
substitution possibilities.

3.1    Product Characteristics

       Fiberglass has a number of desirable characteristics and properties:

       •  Strength—fiberglass has enormous tensile strength that can be up to 100 times
          that of base glass.

       •  High heat and fire resistance—its inorganic nature makes it noncombustible and
          able to withstand elevated temperature environments.

       •  Moisture resistance—it does not absorb moisture, thereby resisting rot and
          retaining its maximum mechanical strength in humid environments.

       •  Corrosion and microbial resistance—it is not corrosive, nor is it degraded by most
          chemicals. Furthermore, it is not affected by bacteria, fungus, mildew, or insect
          attacks.

       •  Lightweight—fiberglass can bear more mass per unit than other materials of the
          same weight.

These properties make fiberglass mats well suited as substrates for roofing and other
applications.

3.2    Uses and Consumers

       Starr (1997) identifies several industries that use mats composed of fiberglass in
residential, commercial, and specialty applications:

       •  The roofing industry—The roofing industry first experimented with fiberglass
          mats in approximately  1945 (Richards, 1979). Today, fiberglass mats are widely
          used as substrates for asphalt shingles and roll roofing.

                                          3-1

-------
       •  The composite industry—fiberglass mats are used as reinforcements for
          thermosets, thermoplastics, and cement and gypsum products.
       •  Other industries—fiberglass mats are also used in miscellaneous specialty
          applications such as battery separators and for pipe-wrapping and flooring.

       Through their use in the production of asphalt roofing products, fiberglass mats serve
the residential and commercial roofing markets. Other specialty fiberglass mats are also used
in residential and commercial construction and in the manufacture of automotive and
industrial equipment components (Schuller Corporation, 1998b). However, the proposed
NESHAP targets the production of wet-formed fiberglass mats that are primarily used as
substrates for roofing products. Therefore, the next subsection focuses on the substitutes for
fiberglass mats in the context of their use in manufacturing asphalt roofing products.

3.3    Substitution Possibilities in Consumption
       Organic felts serve as the primary substitute for fiberglass mats in the production of
asphalt shingles and roll roofing products.  Figure 3-1 presents the share of roll roofing and
strip shingles by base type. As shown, the majority of roll roofing products (67.3 percent)
and strip shingles (92.0 percent) use a fiberglass base. As noted earlier, fiberglass mats
exhibit high heat and fire resistance. Fiberglass-based shingles have a higher fire rating
(Class A— effective against severe fire exposures) when compared to organic shingles
(Class C— effective against light fire exposures). When compared to organic-based shingles,
fiberglass-based shingles are also more durable and moisture resistant, so they are more
stable in high heat and humidity climates. However, fiberglass-based shingles stiffen in cold
temperatures and therefore must be manufactured to meet stringent performance standards to
perform reliably in colder climates (Certainteed, 1998).  Consequently, organic-based
shingles may be preferred in geographical regions that experience cold temperatures.

       There are also several substitution possibilities for the final asphalt roofing products
that use organic and fiberglass bases:

       •   Wood shingles—recent technology improvements have increased wood roofing
          attractiveness.  Wood shingles can be treated with chemicals to improve fire
          ratings and resistance to rot and decay.

       •   Cementious roofing tile—these products are made from cellulose fiber-reinforced
          Portland cement.  They resist weathering, insects, fire, and wind uplift.
                                         3-2

-------
                  Roll Roofing3                               Strip Shinglesb
    Fiberglass /            ^^A  Organic      Fibergiass
      base   /           ^^^^M   base          base
     673%  I           ^^^^H  32.7%         92o/o
          Figure 3-1. Shares of roll roofing and strip shingles by base type: 1992.

a 84.9 million squares of roll roofing were shipped in 1992.
b 167.1 million squares of strip shingles were shipped in 1992.

Source: U.S. Department of Commerce.  1995b. 1992 Census of Manufactures, Industry Series—Petroleum
       and Coal Products. MC92-I-29A. Washington, DC:  Government Printing Office.
       •   Clay roofing tile—it is one of the oldest roofing materials and offers resistance
           qualities similar to cementious roofing tile.

       •   Metal roofing—the most common types are steel and aluminum panels. These
           panels are typically coated with zinc or zinc aluminum mixture to increase
           corrosion resistance.

       •   Slate roofing—slate roofing comes in premium clear pieces and ribbon slate.
           Ribbon  slate is not as durable as premium clear slate.  Although slate is highly
           durable, older slate may require additional maintenance to repair cracked pieces.

       These alternatives are significantly more expensive than fiberglass-based asphalt
shingles and roll roofing. However, they typically have a longer lifespan and often last for
30 to 50 years as compared to fiberglass shingles that last for approximately 15 to 25 years
(Friedman, 1998).
                                           3-3

-------
                                    SECTION 4

                           INDUSTRY ORGANIZATION
       This section describes the structure of the market for fiberglass mats, the
characteristics of the facilities that manufacture these mats, and the characteristics of the
companies that own these facilities.

4.1    Market Structure
       Market structure influences the behavior of producers and consumers in an industry.
If an industry is perfectly competitive, then individual producers are not able to influence the
price of the outputs they sell or the inputs they purchase. This condition is most likely to
hold if the industry has a large number of firms, the products sold are undifferentiated, and
entry and exit of firms are unrestricted. Product differentiation can occur both from
differences in product attributes and quality and from brand name recognition of products.
Entry and exit of firms are unrestricted for most industries except, for example, in cases
where the government regulates who is able to produce, one firm holds a patent on a product,
one firm owns the entire stock of a critical input, or a single firm is able to supply the entire
market.
       When compared across  industries, firms in industries with few firms, more product
differentiation, and restricted entry are more likely to be able to influence the price they
receive for a product by reducing output below perfectly competitive levels. This ability to
influence price is referred to as  exerting market power.  At the extreme, a single monopolistic
firm may supply the entire market and hence set the price of the output. On the input market
side, firms may be able to influence the price they pay for an input  if few firms, both from
within and outside the industry, use that input. At the extreme, a single monopsonist firm
may purchase the entire supply  of the input and hence set the price of the input.
       Economists often evaluate the distribution of market share in an industry by
estimating four-firm concentration ratios (CR4s), eight-firm concentration ratios (CR8s), and
Herfindahl-Hirschman indexes  (HHIs). The CR4 and CR8 ratios measure the share of total
industry sales accounted for by  the top four and eight firms, respectively. The HHI is the sum
                                         4-1

-------
 Table 4-1. Market Concentration Measures for SIC 3229, Pressed and Blown Glass,
 N.E.C.:  1992

                       Index                                     Value
 Four-Firm Concentration Ratio (CR4)                                48
 Eight-Firm Concentration Ratio (CR8)                                72
 Herfindahl-Hirschman Index (HHI)                                 814
Source: U.S. Department of Commerce, Bureau of the Census.  1993. Concentration Ratios in Manufacturing.
       Washington, DC: Government Printing Office.
of squared market shares of all firms in the industry.  Table 4-1 presents the various measures
of market concentration at the four digit SIC level for 1992, which is the most disaggregated
level and the most recent year for which data are available.

       The CR4 ratio for this industry in 1992 was 48, meaning that the top four firms
accounted for 48 percent of the industry's total sales.  The CR8 ratio indicates that top eight
firms accounted for 72 percent of the industry's total sales. Although there is no widely
accepted criterion for determining market structure based on the values of concentration ratios,
the 1992 Department of Justice's Horizontal Merger Guidelines provide criteria for doing so
based on HHIs (U.S. Department of Justice and the Federal Trade Commission, 1992).
According to these criteria, industries with HHIs below 1,000 are considered unconcentrated
(i.e., more competitive), those with HHIs between 1,000 and 1,800 are considered moderately
concentrated (i.e., moderately competitive), and those with HHIs above 1,800  are considered
highly concentrated (i.e., less competitive). The HHI for SIC 3229 is 814, indicating that the
pressed and blown glass industry is fairly competitive. In general, firms in competitive
industries are more likely to be price takers and less likely to pass on the cost burden of a
regulation onto consumers.

       It would be appropriate to reiterate one important caveat at this stage. The proposed
NESHAP is  targeted at SIC 3229325, which is only a subset of SIC  3229. Table 4-2 puts the
relative size  of these industries in perspective.  While 411 firms manufacture pressed and
blown glass products (SIC 3229), only eight firms make fiberglass mats (SIC 3229325). Low
concentration measures at the four-digit SIC level may mask the fact that a particular
subcategory  at the seven-digit SIC level could be more concentrated.
                                         4-2

-------
Table 4-2. Relative Size of SIC 3229 versus SIC 3229325:  1992

SIC 3229
SIC 3229325
Industry
Pressed and Blown Glass, N.E.C.
Fiberglass Mats
Number of
Companies
411
8
Value of Product
Shipments ($106)
4,094.00
231.00
Source: U.S. Department of Commerce, Bureau of the Census. 1995a. 1992 Census of Manufactures, Industry
       Series—Glass Products. MC92-I-32A. Washington, DC: Government Printing Office.

4.2    Manufacturing Facilities

       Facilities comprise a site of land with plant and equipment that combine inputs (raw
materials, fuel, energy, and labor) to produce outputs.  As of 1997, there were 14 operating
wet-formed fiberglass production facilities in the United States (ERG, 1998).  Table 4-3
identifies these facilities and also provides information on their location, sales range, and
employment range. These facilities are mostly concentrated in the southern states of the
United States. Eight facilities have sales of less than $50 million, and seven facilities have
fewer than 250 employees.

4.3    Companies

       Companies that own individual facilities are  legal business entities that have the
capacity to conduct business transactions and make business decisions that affect the facility.
The terms company and firm  are synonymous and refer to the legal business entity that owns
one or more facilities. Based on the facility-level data presented in Table  4-3, eight companies
currently own and operate wet-formed fiberglass manufacturing facilities in the United States.

       Elcor Corporation is the parent company of Elk Corporation.  In the spring of 1996,
Elcor Corporation opened its  newest fiberglass roofing  mat manufacturing facility in Ennis,
TX. This facility has a capacity of 75 million squares, which tripled Elk's previous nonwoven
manufacturing capacity. In addition to captive production of fiberglass mats, this facility
provides mats to other roofing manufacturers (Elcor, 1998).
                                          4-3

-------
Table 4-3.  Sales and Employment of Wet-Formed Fiberglass Mat Facilities: 1997
Name
Celotex3
Elk Corporation*5
Ennis Acquisitions0
GAP Materials
GS Roofing
Herbert Malarkey Roofing
Corporation
Owens-Corning (1)
Owens-Corning (2)
Owens-Corning
Schuller(l)
Schuller (2)
Schuller
Tamko Roofing
Tamko Roofing
NA = not available
Facility Location
Birmingham
Ennis
Ennis
Chester
Charleston
Portland
Aiken
Aiken
Fort Smith
Waterville
Waterville
Etowah
Tuscaloosa
Joplin

AL
TX
TX
SC
SC
OR
SC
SC
AR
OH
OH
TN
AL
MO

Sales Range
($106)
$20-50
$20-50
NA
$10-20
$20-50
$10-20
$100-500
$100-500
$10-20
$100-500
$100-500
$20-50
$20-50
$50-100

Employment
Range
100-249
100-249
NA
50-99
100-249
50-99
1,000-4,999
1,000-4,999
50-99
500-999
500-999
250-499
100-249
250-499

a Celotex Corporation is owned by Jim Walter Corporation.
Elk Corporation is owned by Elcor Corporation.
Sources: American Business Information Inc. (ABI). 1997. The American Business Disk [computer file].
        Omaha, NE: American Business Information.

        Eastern Research Group (ERG).  Memorandum from Brian Palmer, ERG, to Juan Santiago, EPA.
        January 30, 1997.
                                           4-4

-------
       The Jim Walter Corporation owns the Celotex facility in Birmingham, AL. The
facility identified as Ennis Acquisitions used to be operated by Schuller. The facility recently
changed hands and is currently owned by Tamko Roofing Products Inc. (ERG, 1997).

4.3.1   Size Distribution

       Table 4-4 presents sales and employment data for these companies for 1995. As
shown, Owens-Corning is the largest company with respect to total sales ($3.6 billion). In
addition, Schuller and GAP Materials Corporation also report over $1 billion in sales.  Four
companies reported sales between $200 and $400 million, and one company reported sales
below $50 million.  Owen's Corning is the largest employer (17,300 employees). Six
companies reported more than 1,000 employees, and the remaining two reported fewer than
1,000 employees in 1995.

4.3.2   Vertical Integration

       Vertically integrated firms produce inputs they can use to produce other final outputs.
These firms can handle different stages of production at one facility, or they may own several
facilities or subsidiaries that directly or indirectly supply its inputs. Many of the companies
affected by the proposed rule are vertically integrated.  They manufacture asphalt roofing
products in addition to manufacturing fiberglass mats.  For example, Elk Corporation also
manufactures fiberglass-based shingles at its Ennis, TX, facility.  The GS Roofing facility in
Charleston, SC, also produces fiberglass shingles and built-up roofing.  The Herbert Malarkey
company also manufactures fiberglass shingles, built-up roofing, and modified bitumen
roofing. Although the Owens-Corning facilities in Aiken, SC, and Fort Smith, AR, specialize
in the production of fiberglass mats, the parent company operates several other facilities (e.g.,
in Atlanta, GA; Denver, CO; Memphis, TN) that manufacture fiberglass shingles. Similarly,
although the Schuller facilities in Waterville, OH, and Etowah, TN, only manufacture
fiberglass  mats, Schuller operates several other facilities that manufacture built-up roofing and
modified bitumen roofing (ARMA,  1997).
                                         4-5

-------
Table 4-4. Sales and Employment Data for Wet-Formed Fiberglass Mat Companies:
1995
Number of
Company Facilities
Elcor Corporation3
GAP Materials
GS Roofing
Herbert Malarkey Roofing Corporation
Jim Walter Corporation11
Owens-Corning
Schuller
Tamko Roofing Products Inc.d
Totals
1
1
1
1
1
3
3
3
14
Organization
Type
Public
Private
Private
Private
Private
Public
Public
Private

Sales
($106) Employment
$159.1
1,030.0
275.4
27.9
350.0C
3,612.0
1,391.5
240.0
7,085.9
704
4,300
1,200
100
2,700
17,300
7,500
1,300
35,104
a Elcor Corporation owns Elk Corporation.
c  Estimated by Information Access Corporation.
d  Tamko Roofing Products Inc. owns Ennis Acquisitions.

Sources:  Information Access Corporation.  1997. Business Index [computer file].  Foster City, CA:
         Information Access Corporation.

         Elcor Corporation. September 26, 1995. 1 OK for year ended June 30, 1995.  EDGAR Database:
         . Bethesda, MD:
         Lexis/Nexis.

         Owens-Corning. February 23, 1996.  1 OK for year ended December 31, 1995. EDGAR Database:
         . Bethesda, MD:
         Lexis/Nexis.

         Schuller Corporation. 10K for year ended December 31, 1995. World Wide Web site:
         . As obtained March 23, 1998a.
                                              4-6

-------
                                     SECTION 5

                                     MARKETS
       This section examines domestic and foreign market statistics for the fiberglass mat
industry. Historical data for this industry are provided for value of product shipments and
foreign trade.

5.1    Production and Shipment Data

       Table 5-1 presents value of product shipment data for SIC 3229 (Pressed and Blown
Glass, N.E.C.), SIC 32293 (glass fiber, textile-type), and SIC 3229325 for the period 1992
through 19951 (U.S. Department of Commerce, 1997).  During this period, the value of
product shipments grew significantly by 25.6 percent for SIC 3229 and by 36.6 percent for
SIC 32293.  For 1992, fiberglass mats (SIC 3229325) represented  18.2 percent of total textile
glass fiber product shipments. Although disaggregated data for fiberglass mats are not
available for 1993 to 1995, the Agency estimated shipments for fiberglass mats based on
1992 share data. As Table 5-1 shows, product shipments for fiberglass mats are estimated at
$248.0 million for 1993, $278.2 million for 1994, and $315.8 million for 1995.

5.2    Foreign Trade

       Table 5-2 presents U.S. foreign trade data for fiberglass mats for the period 1992
through 1996 (U.S. Department of Commerce, 1996a and b). Data for both quantity and
value of exports and imports are provided for the Harmonized Tariff System Classification
(HS) Grouping 701931, which includes mats, nonwoven of glass fiber. Based on dollar
values, the United States was a net importer of fiberglass mats from 1993 through 1995.
Imports of fiberglass mats increased from $24.88 million in 1993 to $36.78 million in 1996.
This represents an increase of approximately 48 percent in terms of value of imports and
26 percent in terms of volume of imports over this period.
'Value of product shipments represents the total value of all primary products shipped and excludes products
   classified secondary to the industry and miscellaneous receipts (i.e., repair work, sale of scrap, research and
   development, installation receipts, and resales).

                                         5-1

-------
Table 5-1.  Value of Product Shipments:  1992-1995 ($106)
Year
1992
1993
1994
1995
SIC 3229
4,088.2
4,339.7
4,746.2
5,133.3
SIC 32293
1,264.4
1,357.5
1,522.8
1,728.3
SIC 3229325"
231.0
248.0
278.2
315.8
Note: Value of product shipments represents the total value of all products classified as primary to an industry.

a  The Agency computed estimates for 1993 to 1995.

Sources:        U.S. Department of Commerce. Bureau of Census. 1995a. 7992 Census of Manufactures,
               Industry Series-Glass Products. MC92-I-32. Washington, DC: Government Printing Office.

               U.S. Department of Commerce. Bureau of Census. 1997. 7995 Census of Manufactures,
               Value of Product Shipments. M95(AS)-2. Washington, DC: Government Printing Office.
                                             5-2

-------
Table 5-2. U.S. Foreign Trade for Fiberglass Mats: 1992-1996
Exports
Year
1992
1993
1994
1995
1996
Quantity (pounds)
NA
NA
NA
19,763,040
20,050,736
Value
($)
$23,209,607
24,386,841
29,350,960
29,019,369
37,302,424
Imports
Quantity (pounds)
NA
30,372,207
40,413,518
37,778,604
38,320,258
Value
($)
$16,974,565
24,884,162
34,469,079
36,273,948
36,784,043
Note:  Data are for the Harmonized Tariff System Commodity Classification Group 701931 (mats, nonwoven
       of glass fiber).

Sources:   U.S. Department of Commerce, Bureau of the Census.  1996a.  "U.S. Exports History: Historical
          Summary 1991-1995." [CD-ROM]. Washington, DC: Bureau of the Census.
          . As obtained on March 20, 1998.

          U.S. Department of Commerce, Bureau of the Census.  1996b.  "U.S. Imports History: Historical
          Summary 1991-1995." [CD-ROM]. Washington, DC: Bureau of the Census.
          . As obtained on March 20, 1998.
                                               5-3

-------
                                   SECTION 6

            ECONOMIC IMPACTS AND SMALL BUSINESS ANALYSIS
       The Regulatory Flexibility Act (RFA) of 1980, as amended in 1996 by the Small
Business Regulatory Enforcement Fairness Act (SBREFA), requires that special
consideration be given to small entities affected by federal regulations. This section focuses
on identifying the small businesses affected by the proposed NESHAP, presenting
incremental compliance cost data, and assessing the economic impact of these incremental
costs.
6.1     Identifying Small Businesses
       The Small Business Administration (SBA) defines a small business in SIC 3229
Pressed and Blown Glass, N.E.C. (under which fiberglass mats are classified) as an
independent company with fewer than 750 total employees (SBA, 1998). The Agency has
used this definition to identify one small business that will potentially be affected by the
proposed NESHAP. Table 6-1 presents data on the number of employees and the small
business status for all potentially affected companies. Although Elcor Corporation employed
only 704 people in  1995, it had expanded to 800 employees by 1997; therefore, the Agency
did not count it as a small business. Based on the most recent data available, the Herbert
Malarkey Roofing Corporation employs only 100 workers, thereby qualifying as a small
business. This small company owns and operates one facility, or 7.1 percent of all facilities
currently manufacturing wet-formed fiberglass mats in the United States. The annual
capacity of this facility was 2,932 tons, or 7.2 percent of the industry's total capacity of
approximately 403,058 tons in 1995 (ERG, 1998).

6.2     Estimated Incremental Emission Control Costs
       Some facilities might need to modify or improve their current operations and control
devices in order to achieve one of the following control  options for formaldehyde emissions
from each drying and curing oven: (1) 96 percent reduction or (2) 0.05 Ib/ton (0.03 kg/Mg)
mass emission limit. Table 6-2 presents the Agency's estimates of existing control costs and
the incremental annual costs that each affected facility would have to incur to comply with
                                        6-1

-------
Table 6-1.  Employment Data for Affected Companies: 1995


Company Name
Elcor Corporation3
GAP Materials
GS Roofing
Herbert Malarkey Roofing Corporation
Jim Walter Corporation0
Owens-Corning
Schuller
Tamko Roofmgd
Total
Number of
Affected
Facilities
1
1
1
1
1
3
3
3
14

Number of
Employees
704
4,300
1,200
100
2,700
17,300
7,500
1,300
35,104

Small
Business
Nob
No
No
Yes
No
No
No
No
1
a Elcor Corporation owns Elk Corporation.
b By 1997, Elcor Corporation had 800 employees, which brings it over the small business threshold.
c The Jim Walter Corporation owns Celotex Corporation.
d Tamko Roofing owns Ennis Acquisitions.
the proposed rule. The incremental annual cost includes the cost of additional operation,
maintenance labor and materials, supervisory labor, utilities, administrative charges, taxes,
insurance, and the annualized cost of the new capital. Nine out of the 14 affected facilities
already have existing thermal oxidizers on their drying and curing oven exhaust and have
little or no new compliance costs. The total incremental cost burden on the affected facilities
is estimated to be $1.564 million per year. The average incremental cost per facility, based
on all facilities is $0.112 million per year. The average incremental cost for the seven
facilities that might have to modify their operations to achieve the emissions standard is
$0.223 million per year.
                                           6-2

-------
Table 6-2. Estimated Facility-Level Compliance Costs (1995$)
Facility
Celotex
Elk Corporation
Ennis Acquisitions
GAP Materials
GS Roofing
Herbert Malarkey
Roofing
Corporation
Owens-Corning (1)
Owens-Corning (2)
Owens-Corning
Schuller(l)
Schuller (2)
Schuller
Tamko Roofing
Tamko Roofing
Total
Location
AL
TX
TX
sc
sc
OR
SC
SC
AR
OH
OH
TN
AL
MO

Annual Cost
Currently for Existing
Controlled Control Device
Emissions ($106)
Drying oven
Drying oven
Drying oven
None
Drying oven
None
Drying oven
Drying oven
Drying oven
Drying oven and
binder application
Drying oven
None
None
None

$0.291
$0.143
$0.387
$0.000
$0.133
$0.000
$0.299
$0.278
$0.317
$0.409
$0.541
$0.000
$0.000
$0.000
$2.798
Incremental
Annual Cost
due to
Proposed
Rule
($106/yr)
$0.000
$0.000
$0.000
$0.486
$0.013
$0.096
$0.000
$0.000
$0.009
$0.000
$0.000
$0.528
$0.146
$0.286
$1.564
Source: Eastern Research Group (ERG). Memorandum from Danny Greene, ERG, to Juan Santiago,
       Environmental Protection Agency. February 17, 1998. Regulatory alternative analysis for the
       wet-formed fiberglass mat industry.
                                            6-3

-------
6.3    Economic Impact
       A "sales tesf'or screening test is conducted to assess the economic impact of this rule,
as recommended in EPA (1997). This test involves estimating the ratio of incremental
annual emission control costs to baseline sales revenues (cost to sales ratios) for each affected
company and for the industry on average. The cost to sales ratio refers to the increase in
control cost divided by the sales revenues of a particular good or goods being produced in the
process for which additional pollution control is required.  This ratio can be estimated for
either individual firms or as an average for some set of firms such as affected small firms.
While the ratio has different significance for different market situations, it is a good rough
gage of potential impact. If costs for the individual (or group) of firms are completely passed
on to the purchasers of the good(s) being produced, the ratio  is an estimate of the price
change (in percentage form after multiplying the ratio by 100). If costs are completely
absorbed by the producer with no price increase, it is an estimate of the percentage of existing
firm revenues necessary to cover the increases costs (in percentage  form after multiplying the
ratio by 100).  The distribution of cost to sales ratios across the whole market, the
competitiveness of the market, and profit to  sales ratios are among the obvious factors that
may influence the determination of significance of any particular cost to sales ratio for an
individual facility. As shown in Table 6-3, the incremental annual compliance cost is less
than 1 percent for all the companies in this industry. The mean cost to sales ratio  is
0.076 percent for all companies, and 0.102 percent for the six companies that might incur
additional costs to meet the emissions standard. The Agency's screening analysis does not
indicate a significant negative impact for any of the affected  companies.  Furthermore, the
single small business affected, Herbert Malarkey Roofing Corporation, is also expected to
incur incremental  compliance costs of less than 1 percent of sales. Thus, the EPA finds that
the proposed regulation to limit emissions from the production of wet-formed fiberglass mat
will not significantly impact a substantial number of business entities. The impact to the
industry as a whole and firms within the industry is anticipated to be minimal.
                                         6-4

-------
Table 6-3. Ratio of Annual Compliance Cost to Baseline Sales for Wet-Formed
Fiberglass Manufacturing Companies: 1995



Company
Elcor Corporation3
GAP Materials
GS Roofing
Herbert Malarkey
Roofing Corporation
Jim Walter Corporation11
Owens-Corning
Schuller
Tamko Roofing0
Total/average

Number
of
Facilities
1
1
1
1

1
3
3
3
14


Small
Business
No
No
No
Yes

No
No
No
No
1


Sales
($106/yr)
$230.0
$1,030.0
$275.4
$27.9

$350.0
$3,832.0
$1,552.1
$240.0
$7,537.4
Incremental
Annual
Compliance
Cost
(SlOVyr)
$0.000
$0.486
$0.013
$0.096

$0.000
$0.009
$0.528
$0.432
$1.564


Cost-to-
Sales Ratio
0.0000%
0.0472%
0.0048%
0.3441%

0.0000%
0.0002%
0.0340%
0.1800%
0.0763%
a Elcor Corporation owns Elk Corporation.
b The Jim Walter Corporation owns Celotex Corporation.
c Tamko Roofing owns Ennis Acquisitions.
                                         6-5

-------
                                 REFERENCES
American Business Information Inc. (ABI).  1997.  The American Business Disk [computer
       file].  Omaha, NE: American Business Information.

Asphalt Roofing Manufacturers Association (ARMA). September 24, 1997. Facsimile on
       member company plant listings. Calverton, MD.

Certainteed.  1998. "New Dimensions in Asphalt Shingles." World Wide Web Site:
       .

Eastern Research Group (ERG). Memorandum from Brian Palmer, ERG, to Juan Santiago,
       U.S. Environmental Protection Agency. January 30, 1997. Maximum achievable
       control technology (MACT) analysis for the wet-formed fiberglass mat industry.

Eastern Research Group (ERG). Memorandum from Danny Greene, ERG, to Juan Santiago,
       U.S. Environmental Protection Agency. February  17, 1998. Regulatory alternative
       analysis for the wet-formed fiberglass mat industry.

Elcor Corporation. September 26, 1995. 1 OK for year ended June 30, 1995. EDGAR
       Database: .  Bethesda, MD:  Lexis/Nexis.

Elcor Corporation. World Wide Web Site: .  As obtained March 13, 1998.

Friedman, Dan. "Roof Shingle Failures: Asphalt and Fiberglass-based Asphalt Warranty
       Claims, Data and Assistance." American Home Service Company. Last updated
       March 26, 1998.  .

Information Access Corporation.  1997. Business Index [computer file]. Foster City, CA:
       Information Access Corporation.

Owens-Corning. February 23, 1996.  1 OK for year ended December 31, 1995.  EDGAR
       Database: .  Bethesda, MD:  Lexis/Nexis.
                                       R-l

-------
Richards, David E.  1979.  "Fibrous Glass Felt." In Roofing Systems: Materials and
      Applications. J. Watson, ed. Reston, VA: Prentice-Hall Company.

Schuller Corporation. 10K for year ended December 31, 1995.  World Wide Web site:
      . As obtained March 23, 1998a.

Schuller Corporation. World Wide Web Site:  . As obtained March 13,
      1998b.

Starr, Trevor. 1997. Glass Fiber Directory and Databook.  2nd ed. London:  Chapman &
      Hall.

U.S. Department of Commerce, Bureau of the Census. 1993. Concentration Ratios in
      Manufacturing. Washington, DC: Government Printing Office.

U.S. Department of Commerce, Bureau of the Census. 1995a. 1992 Census of
      Manufactures, Industry Series—Glass Products. MC92-I-32A. Washington, DC:
      Government Printing Office.

U.S. Department of Commerce, Bureau of the Census. 1995b. 1992 Census of
      Manufactures, Industry Series—Petroleum and Coal Products. MC92-I-29A.
      Washington, DC:  Government Printing Office.

U.S. Department of Commerce, Bureau of the Census. 1996a. "U.S. Exports History:
      Historical Summary 1991-1995." [CD-ROM]. Washington, DC:  Bureau of the
      Census.  . As obtained on
      March 20, 1998.

U.S. Department of Commerce, Bureau of the Census. 1996b. "U.S. Imports History:
      Historical Summary 1991-1995." [CD-ROM]. Washington, DC:  Bureau of the
      Census.  . As obtained on
      March 20, 1998.

U.S. Department of Commerce, Bureau of the Census. 1997. 1995 Annual Survey of
      Manufactures—Value of Product Shipments.  M95 (AS-2).  Washington, DC:
      Government Printing Office.

U.S. Department of Justice and the Federal Trade Commission. April 2, 1992. Horizontal
      Merger Guidelines. Washington, DC:  Government Printing Office.
                                       R-2

-------
U.S. Environmental Protection Agency. February 5, 1997.  EPA Interim Guidance for
       Implementing the Small Business Regulatory Enforcement Fairness Act and Related
       Provisions of the Regulatory Flexibility Act. Washington, DC: Government Printing
       Office.

U.S. Small Business Administration. "Size Standards by SIC Industry." . Last updated January 1998.
                                        R-3

-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO. 2.
EPA-452/R99-001
4. TITLE AND SUBTITLE
Economic Impact Analysis for the Proposed N
Emission Standard for Hazardous Air Pollutan
Production of Wet-Formed Fiberglass Mat
ational
ts from the
7. AUTHOR(S)
Linda M. Chappell, Innovative Strategies and Economics Group
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
12. SPONSORING AGENCY NAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
January 1999
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report evaluates the impacts of the proposed rule for controls of hazardous air pollutants (HAPs) in the
wet formed fiberglass mat industry. The impacts of the rule are measured in terms of cost to sales ratios.
Impacts on small entities are also examined.
1 7. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
economic impacts
small entity impacts
1 8. DISTRIBUTION STATEMENT
Release Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS c COSATI Field/Group
Air Pollution control
Economic Impact Analysis
Regulatory Flexibility Analysis
1 9. SECURITY CLASS (Report) 2 1 . NO. OF PAGES
Unclassified 152
20. SECURITY CLASS (Page) 22. PRICE
Unclassified
EPA Form 2220-1 (Rev. 4-77)    PREVIOUS EDITION IS OBSOLETE
                                                       U.S. Environment*! Protection Agencv

-------