vvEPA
United States Office of Air Quality
Environmental Protection Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-79-0053
June 1979
Air
Glass Manufacturing
Plants
Background Information:
Proposed Standards
of Performance
Draft
EIS
-------
-------
Background Information
and Draft
Environmental Impact Statement
for Glass Manufacturing Plants
Volume 1
Type of Action: Administrative
Prepared by:
t
Director, Ennssiqn Standards and Engineering Division
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
(Date)
Approved by:
Assistant Administrator for Air, Noise and Radiation
Environmental Protection Agency
Washington, D.C. 20460
Draft Statement Submitted to EPA's
Office of Federal Activities Review on
June, 1979
(Date)
Additional copies may be obtained at:
Environmental Protection Agency Library (MD-35)
Research Triangle Park, North Carolina 27711
This document may be reviewed at:
Central Docket Section
Room 2903B, Waterside Mall
Environmental Protection Agency
401 M Street, S.W.
.Washington, D.C. 20460
-------
-------
EPA-450/3-79-005a
Glass Manufacturing Plants
Background Information:
Proposed Standards of Performance
Emission Standards and Engineering Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1979
-------
This report is issued by the Environmental Protection Agency to report technical data of interest to a
limited number of readers. Copies are available - in limited quantities - from the Library Services
Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service, 5285 Port Royal Road,
Springfield, Virginia 22161.
Publication No. EPA-450/3-79-005a
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES .' vi
LIST OF TABLES viii
CHAPTER 1. SUMMARY j.j
1.1 Proposed Standards 1-1
1.2 Environmental Impact 1-3
1.3 Economic Impact .1-3
CHAPTER 2. INTRODUCTION 2-1
2.1 Authority for Standards 2-1
2.2 Selection of Categories of Stationary
Sources 2-6
2.3 Procedure for Development of Standards'of'
Performance .... ,..' 2-8
2.4 Consideration of Costs 2-11
2.5 Consideration of Environmental Impacts ... 2-12
2,6 Impact on Existing Sources 2-14
2.7 Revision of Standards of Performance 2-15
CHAPTER 3. THE GLASS MANUFACTURING INDUSTRY, 3-1
3.1 General , 3-1
3.2 Glass Manufacturing Processes and Their
Emissions 3-2
3.3 References' for Chapter 3 3-25
CHAPTER 4. EMISSION CONTROL TECHNIQUES ... 4-1
4.1 Introduction 4-1'
4.2 Process Modifications 4-3
4.3 All-Electric Mel ters 4-6
. 4.4 Conventional Fabric Filter Systems 4-9
4.5 Venturi Scrubber System .. 4-14
4.6 Electrostatic Precipitators 4-19
4.7 Additional and Developing Control
Techniques 4-27
4.8 Summary of Particulate Control Techniques 4-28
4.9 Control of Sulfur Oxides, Fluoride,
Arsenic, and Lead Emissions from Glass
Manufacturing 4-31
4.10 References for Chapter 4 4-33
-------
TABLE OF CONTENTS (Continued)
Page
CHAPTER 5. MODIFICATION AND RECONSTRUCTION 5-1
5.1 General 5"1
5.2 Modification of Glass-Producing Plants ... 5-1
5.3 Reconstruction of Glass-Producing Plants . 5-2
5.4 References for Chapter 5 5-5
CHAPTER 6. ALTERNATIVE REGULATORY OPTIONS 6-1
6.1 Basis for Regulatory Options 6-1
6.2 Alternative Regulatory Options for
Container Glass Manufacturing 6-2
6.3 Alternative Regulatory Options for Pressed
and Blown Manufacturing Soda-Lime Glass
Formulations 6-3
6.4 Alternative Regulatory Options for Pressed
and Blown Manufacturing Other Than
Soda-Lime Formulations 6-4
6.5 Alternative Regulatory Options for Wool
Fiberglass Manufacturing 6-4
6.6 Alternative Regulatory Options for Flat
Glass Manufacturing 6-5
6.7 Summary of Numerical Emission Limits 6-6
6.8 Numerical Emission Limits for Fuel
Oil-Fired Glass Melting Furnaces 6-7
6.9 Model Plant Parameters 6-8
6.10 Comparison of Alternative Regulatory
Options With State Compliance Limits for
Existing Glass Facilities 6-8
CHAPTER 7. ENVIRONMENTAL IMPACT - 7'1
7.1 Air Quality Impact 7-1
7.2 Water Pollution Impact /-19
7.3 Solid Waste Impact 7-ZO
7.4 Energy Impact '~^t
7.5 References for Chapter 7 I-*-'
CHAPTER 8. ECONOMIC IMPACT 8"1
8.1 Industry Economic Profile 8-2
8.2 Cost Analysis of Alternative Control
Systems : 8-50
8.3 Other Environmental Cost Considerations .. 8-89
8,4 Economic Impact Assessment .., * 8-92
-------
TABLE OF CONTENTS (Concluded)
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D
APPENDIX E
Page
A-l
B-l
C-l
D-l
E-l
-------
LIST OF FIGURES
Figure
3-1 Typical Flow Diagram for the Manufacture of Soda-Lime
Glass 3-5
3-2 Typical Side-Port Furnace and End-Port Furnace 3-9
3-3 Example of One Type of Forming Used in Container
Glass Production 3~lj
4-1 A Simple Two-Cell Inside Out Baghouse Equipped for
Shake Cleaning 4~10
4-2 Typical Scrubber System 4-15
4-3 Conventional and NAFCO Electrostatic Precipitators .. 4-20
8-1 New Jersey SIP for Glass Manufacturing Furnace, the
Baseline Case 8'54
8-2 Reported Installed Costs of Fabric Filter Control
Systems Compared With Estimated Cost Curve Used in
This Study
8-57
8-3 Reported Installed Costs of Electrostatic Precipitator
Control Systems Compared With Estimated Cost Curve
Used in This Study 8'58
8-4 Reported Installed Costs of Scrubber Control Systems
Compared With Estimated Cost Curve Used in This Study 8-61
8-5 Cost Effectiveness of Control Options for the
Container Segment
8-80
8-6 Cost Effectiveness of Control Options for the Pressed
and Blown (Borosilicate, Opal, and Lead) Segment 8-81
8-7 Cost Effectiveness of Control Options for the Pressed
and Blown (Soda-Lime) Segment 8'82
8-8 .Cost Effectiveness of Control Options for the Wool *
Fi berglass Segment
8-83
VI
-------
LIST OF FIGURES (Concluded)
Figure
8-9
Cost Effectiveness of Model Plant Control Alterna
tives Using Fabric Filters ............ .
8-10 Cost Effectiveness of Model Plant Control Alterna
tives Using Electrostatic Precipitators
Page
8-85
8-86
-------
LIST OF TABLES
Table
1-1 Matrix of Environmental and Economic Impacts of the
Proposed Participate Emission Limits ................. 1-5
3-1 1976 Production Rates and Values of Shipments ........ 3-3
3-2 Projected 1985 Production Rates ...................... 3-3
3-3 Raw Material Batch Recipes ........................... 3-15
3-4 Emissions From Uncontrolled Glass Melting Furnaces for
Each Industry Category ..................... ......... 3-20
3-5 State Parti cul ate Regulations for Existing Stationary
Sources .............................................. 3-24
4-1 All -Electric Glass Melting Furnace Parti cul ate
Emissions Tests ...................................... 4"8
4-2 Parti cul ate Emission Test Results for Glass Melting
Furnaces Equipped With Fabric Filters ---- . ........... 4-12
4-3 Parti cul ate Emission Test Results for Glass Melting
Furnaces Equipped With Venturi Scrubbers ............. 4-17
4-4 Parti cul ate Emission Test Results for Glass Melting
Furnaces Equipped With Electrostatic Precipitators ... 4-23
4-5 Representative Parti cul ate Emissions From Glass
Melting Furnaces ..................................... 4~29
5-1 GPI List of Glass Melting Furnace Maintenance and
Alterations .......................................... 5~3
6-1 Summary of Alternative Regulatory Options ............ 6-7
6-2 Model Plant Parameters ............................... 6~10
6-3 -Comparison of Alternative Regulatory Options With New
Jersey State Compliance Regulations .................. 6-14
-------
LIST OF TABLES (Continued)
Table
Page
7-1 Glass Industry Growth in the Period From
1978 to 1983 7-2
7-2 Container Glass Category Magnitudes of, and Distances
to Maximum 24-Hour and Annual Average Particulate
Concentrations 7-4
7-3 Pressed and Blown: Soda-Lime 50 TPD Magnitudes of,
and Distances to Maximum 24-Hour and Annual Average
Particulate Concentrations 7-7
7-4 Pressed and Blown: Soda-Lime 100 TPD Magnitudes of,
and Distances to Maximum 24-Hour and Annual Average
Particulate Concentrations i 7-8
7-5 Pressed and Blown: Other Than Soda-l|ime 50 TPD
Magnitudes of, and Distances to Maxi'mum 24-Hour and
Annual Average Particulate Concentrations 7-11
7-6 Pressed and Blown: Other Than Soda-Lime 100 TPA
Magnitudes of, and Distances to Maximum 24-Hour and
Annual Average Particulate Concentrations 7-12
7-7 Wool Fiberglass Magnitudes of, and Distances to
Maximum 24-Hour and Annual Average Particulate
Concentrations 7-14
7-8 " Flat Glass Magnitudes of, and Distances to Maximum
24-Hour and Annual Average Particulate
Concentrations 7-17
7-9 Energy Requirements for Control Systems 7-23
7-10 Control Equipment Combinations and Energy
Requi rements 7-24
8-1 Flat Glass Plants , 8-9
8-2 Share of Total Packagi ng Market 8-13
8-3 ' Shipments of Wool Fiberglass 8-23
8-4 Shipments of Textile Fiberglass 8-30
IX
-------
LIST OF TABLES (Continued)
Table
8-5
8-6
8-7
8-8
8-9
8-10
8-11
8-12
8-13
8-14
8-15
8-16
8-17
8-18
8-19
Estimated New Sources for Pressed and Blown Glass
(1977-1982)
Page
8-47
Control Combinations 8-51
Uncontrolled Exhaust Parameters for Model Plant 8-52
Regulatory Options for Particulate Emissions 8-53
Component Capital Costs Estimated Separately by
Module 8-56
Calculation of Annualized Costs of Air Pollution
Control Systems 8-64
Control Device Parameters for Regulatory Options and
Baseline SIP - 8'66-
Capital Costs of Particulate Control for New Glass
Furnaces - Container Segment 8-68
Capital Costs of Particulate Control for New Glass
Furnaces Flat Segment -
8-69
Capital Costs of Particulate Control for new Glass
Furnaces Pressed and Blown (Borosilicate, Opal,
and Lead) Segment 8-70
Capital Costs of Particulate Control for new Glass
Furnaces Pressed and Blown (Soda-Lime) Segment ... 8-71
Capital Costs of Particulate Control for new Glass
Furnaces Wool Fiberglass Segment 8-72
Incremental Annualized Costs of Particulate Controls
Allocable to the Regulatory Options for a New Glass
Furnace Container Segment
8-75
Incremental Annualized Costs of Particulate Controls
Allocable to the ReguTatroy Options for a New Glass
Furnace Flat Segment
8-76
Incremental Annualized Costs of Particulate Controls
Allocable to the Regulatory Options for a New Glass
Furnace Pressed and Blown (Borosilicate, Opal, and
Lead) Segment 8"77
x
-------
LIST OF TABLES (Concluded)
Table
8-20 Incremental Annualized Costs of Particulate Controls
Allocable to the Regulatory Options for a New Glass
Furnace Pressed and .Blown (Soda-Lime) Segment
Paqe
8-78
8-21 Incremental Annualized Costs of Particulate Controls
Allocable to the Regulatory Options for a New Glass
Furnace Wool Fiberglass Segment 8-79
8-22 Glass Manufacturing Water Pollution Control Costs
(Increment of NSPS Over BPT) 8-91
8-23 Data Sources for Model Plant Characteristics 8-94
8-24 Selected Control Cost Estimates for New Grassroots
Glass Plants 8-96
8-25 Discriminant Analysis Results 8-98
8-26 Scaled Control Costs and Recalculated Price
Increases 8-100
8-27 Changes in Rate of Return on Assets After the
Imposition of NSPS Controls 8-101
8-28 Discounted Cash Flow Methodology . 8-103
8-29 Handmade Consumerware Discounted Cash Flow 8-104
8-30 Container Glass Mid-Range Estimates Discounted Cash
Flow (500 TPD) 8-105
8-31 Container Glass Mid-Range Estimates Discounted Cash
Flow (500 TPD) 8-106
8-32 New Sources: Fifth Year Annualized Costs 8-110
-------
-------
1. SUMMARY
1.1 PROPOSED STANDARDS
This Background Information Document (BID.) (formerly Standards Support
and Environmental Imoact Statement... or SSEIS) supports proposed standards
for particulate emissions from glass melting furnaces within glass ;
manufacturing plants. Atmospheric emissions from glass manufacturing
plants, 98 percent of which originate from glass melting furnaces,
include particulates, nitrogen oxides, and sulfur oxides. Demonstrated
control technology exists only for particulates;. therefore, the proposed
standards of performance anply to particulates only. .Additional information
and regulatory rationale may be found in the preamble.and regulation for
Subpart CC in the Federal Register.
The glass manufacturing industry is. divided into, production categories
based on industry definitions embodied in the Standard Industrial Classification
(SIC) system.. The division into categories based on the SIC system provided
a basis for the economic analysis and also provided a basis for other analyses
based on technical differences within the glass industry. There are
four glass manufacturing industry SIC codes:
SIC 3211 - Flat glass ... . .
. SIC 3221. - Container glass
SIC 3229 - Pressed and blown glass, not elsewhere
. ... . classified (n.e.c.)
SIC 3296 - Wool fiberglass
1-1
-------
In addition, SIC 3229 is sufficiently diversified to warrant a technical
division into two subcategories based on glass formulation: (a) glass
produced using soda-lime formulation, and (b) glass produced using other
than soda-lime formulation.
The proposed emission standards would restrict particulate emissions
from natural gas-fired glass melting furnaces to:
0.1 g/kg (0.2 Ib/ton) of glass pulled from glass melting
furnaces used for container glass production;
0.1 g/kg (0.2 Ib/ton) of glass pulled from furnaces used for
pressed and blown, n.e.c., glass production of soda-lime formulation;
0.25 g/kg (0.5 Ib/ton) of glass pulled from furnaces used for
Dressed and blown, n.e.c., glass production of other than soda-lime
formulation;
0.2 g/kg (0.4 Ib/ton) of glass pulled from furnaces used for
wool fiberglass production; and
0.15 g/kg (0.3 Ib/ton) of glass pulled from furnaces used for
flat glass production.
Although natural gas is the fuel traditionally used to fire glass
melting furnaces, there is a growing trend toward use of fuel oil, which
does not burn as cleanly as natural gas. Therefore, an increment of 15
percent over these emission limits is proposed for fuel oil-fired glass
melting furnaces.
Control of particulate emissions from glass manufacturing plants is
achieved by installation of an emission control system to remove particulate
matter from the exhaust gas stream. Electrostatic orecipitators (ESP)
and fabric filters have been adequately demonstrated to be the best
technological systems of continuous emission reduction for glass melting
1-2
-------
furnaces producing container glass; pressed and blown glass, n.e.c., of
soda-lime formulation; pressed and blown glass, n.e.c., of other than
soda-lime formulation; and wool fiberglass. No installed control systems
have been evaluated for oarticulate emission reduction from flat glass
installations; however, ESP are considered adequately demonstrated
systems of emission reduction because of the similarity of processes
used for flat glass and container glass production and because of a
performance guarantee underwritten for a flat glass facility by an ESP
manufacturer. These emission reduction methods are not equally applicable
to glass melting furnaces in all sectors of the industry, hence the
ft <>
proposal of separate standards for each of the five nroduction categories.
1.2 ENVIRONMENTAL IMPACT
The oroposed emission limits would reduce oarticulate emissions
from glass melting furnaces placed on line in glass manufacturing plants
between 1978 and 1983 by 90-96 percent without adversely affecting
water quality, solid waste disoosal, energy conservation, or noise level.
The environmental imoacts of the orooosed emission limits are summarized
in Table 1-1.
1.3 ECONOMIC IMPACT
An economic impact assessment of the proposed emission limits
has been prepared as required under Section 317 of the Clean Air Act
(as amended in 1977). The proposed limits would have negligible imoact
on compliance costs, inflation or recession, competition with respect to
small business, consumer costs, and energy use. The standards would
reduce profitability (as measured by rate of return on assets) by
1-3
-------
approximately 0.2-0.8 percent. To offset this loss of return, glass
manufacturers may increase prices by similar amounts.
The Agency's guideline for determining the necessity for developing
an Inflationary Impact Statement is increased operating costs in the
fifth year of operation of more than $100 million. The increase in
operating costs in the fifth year associated with the proposed limits is
about $10.4 million oer year.
The economic impacts of the proposed emission limits are summarized
in Table 1-1. .
-------
oo
oo
=£
_J
O
ii
§
n.
LU
OO
O
D-
O
o;
Q.
O
00
O
i tO
O to
oo 3
o
S- -i-
0) 4->
4-> 3
O
D_
cu
_a
ns
X
X
X
X
X
X
X
X
CO
i i T3 to
o c:
O- S- 03
o o_ +->
x
x
X
X
x
x
X
X
X
X
C\J
E
O
ex
o
w
o
fO
-o
in
-a
ea
o
c
to o
4J -0
oo a;
to
o
O
4J re
O -M Q.
n3 o E
Q. (C > i
E Q-
E
4-3 I >
CD <1>
S-
S-
O
oo _i i <
XXX
X X
X
O) ra
4-> r Q.
O JD E
CO T- ti
Q- CD
E T- f
u
(O
Q-4->
E U
t-< 03
a.
I-H ro
o.
fO >i
O QJ
I- l/l
M- i.
cu cu
c. >
.cu -a
JCQ =t
1+ i
1-5
-------
-------
2. INTRODUCTION
Standards of performance are proposed following a detailed investi-
gation of air pollution control methods available to the affected
industry and the impact of their costs on the industry. This document
summarizes the information obtained from such a study. Its purpose is
to explain in detail the background and basis of the proposed standards
and to facilitate analysis of the proposed standards by interested
persons, including those who may not be familiar with the many technical
aspects of the industry. To obtain additional copies of this document
or the Federal Register notice of proposed standards, write to EPA
Library (MD-35), Research Triangle Park, North Carolina 27711. Specify
"Glass Manufacturing Plants, Background Information: Proposed Standards
of Performance," report number EPA-450/3-79-005a when ordering.
2.1 AUTHORITY FOR THE STANDARDS.
Standards of performance for new stationary sources are established
under section 111 of the Clean Air Act (42 U.S.C. 7411), as amended,
hereafter referred to as the Act. Section 111 directs the Administrator
to establish standards of performance for any category of new stationary
source of air pollution which "... causes or contributes significantly
to, air pollution which may reasonably be anticipated to endanger public
health or welfare."
2-1
-------
The Act requires that standards of performance for stationary
sources reflect, ". . . the degree of emission limitation achievable
through the application of the best technological system of continuous
emission reduction . . . the Administrator determines has been adequately
demonstrated." In addition, for stationary sources whose emissions
result from fossil fuel combustion, the standard must also include a
percentage reduction in emissions. The Act also provides that the cost
of achieving the necessary emission reduction, the nonair quality health
and environmental impacts and the energy requirements all be taken into
account in establishing standards of performance. The standards apply
only to stationary sources, the construction or modification of which
commences after regulations are proposed by publication in the Federal
Register.
The 1977 amendments to the Act altered or added numerous provisions
which apply to the process of establishing standards of performance.
1. EPA is required to list the categories of major stationary
sources which have not'already been listed and regulated under standards
of performance. Regulations must be promulgated for these new categories
on the following schedule:
25 percent of the listed categories by August 7, 1980 j;
75 percent of the listed categories by August 7, 1981 ,-F
TOO percent of the listed categories by August 7, 1982
A governor of a State may apply to the Administrator to add a category
which is not on the list or to revise a standard of performance.
2. EPA is required to review the standards of performance every
four years, and if appropriate, revise them.
2-2
-------
3. EPA is authorized to promulgate a design, equipment, work
practice, or operational standard when an emission standard is not
feasible.
4. The term "standards of performance" is redefined and a new term
"technological system of continuous emission reduction" is defined. The
new definitions clarify that the control system must be continuous and
may include a low-polluting or non-polluting process or operation.
5. The time between the proposal and promulgation of a standard
under section 111 of the Act is extended to six months.
Standards of performance, by themselves, do not guarantee protection
of health or welfare because they are not designed to reflect the degree
of emission limitation achievable through application of the best adequately
demonstrated technological system of continuous emission reduction,
taking into consideration the cost of achieving such emission reduction,
any nonair quality health and environmental impact and energy requirements.
Congress had several reasons for including these requirements.
First, standards with a degree of uniformity are needed to avoid situations
where some States may attract industries by relaxing standards relative
to other States. Second, stringent standards enhance the potential for
long-term growth. Third, stringent standards may help achieve long-term
cost savings by avoiding the need for more expensive retrofitting when
pollution ceilings may be reduced in the future. Fourth, certain types
of standards for coal burning sources can adversely affect the coal
market by driving up the price of low-sulfur coal or effectively excluding
certain coals from the reserve base because their untreated pollution
2-3
-------
potentials are high. Congress does not intend that new source performance
standards contribute to these problems. Fifth, the standard-setting
process should create incentives for improved technology.
Promulgation of standards of performance does not prevent State or
local agencies from adopting more stringent emission limitations for the
same sources. States are free under section 116 of the Act to establish
even more stringent emission limits than those established under section
111 or those necessary to attain or maintain the national ambient air
quality standards (NAAQS) under section 110. Thus, new sources may in
some cases be subject to limitations more stringent than standards of
performance under section 111, and prospective owners and operators of
new sources should be aware of this possibility in planning for such
faci1i ti es.
A similar situation may arise when a major emitting facility is to
be constructed in a geographic area which falls under the prevention of
significant deterioration of air quality provisions of Part C of the
Act. These provisions require, among other things, that major emitting
facilities to be constructed in such areas are to be subject to best
available control technology. The term "best available control tech-
nology" (BACT), as defined in the Act, means ". . . an emission limitation
based on the maximum degree of reduction of each pollutant subject to
regulation under this Act emitted from or which results from any major
emitting facility, which the permitting authority, on a case-by-case
basis, taking into account energy, environmental, and economic impacts
and other costs, determines is achievable for such facility through
application of production processes and available methods, systems, and
2-4
-------
techniques, including fuel cleaning or treatment or innovative fuel
combustion techniques for control of each such pollutant. In no event
shall application of 'best available control technology' result in
emissions of any pollutants which will exceed the emissions allowed by
any applicable standard established pursuant to section 111 or 112 of
this Act."
Although standards of performance are normally structured in terms
of numerical emission limits where feasible, alternative approaches are
sometimes necessary. In some cases physical measurement of emissions
from a new source may be impractical or exorbitantly expensive. Section
lll(h) provides that the Administrator may promulgate a design or equipment
standard in those cases where it is not feasible to prescribe or enforce
a standard of performance. For example, emissions of hydrocarbons from
storage vessels for petroleum liquids are greatest during tank filling.
The nature of the emissions, high concentrations for short periods
during filling, and ,low concentrations for longer periods during .storage,
and the configuration of storage tanks make direct emission measurement - -
impractical. Therefore, a more practical, approach to standards of
performance for storage vessels has been equipment specification.
In addition, section lll(h) authorizes the Administrator to grant
waivers of compliance to permit a source to use innovative continuous
emission control technology. In order to grant the waiver, the Administrator
must find: (1) a substantial likelihood that the technology will produce
greater emission reductions than the standards require, or an equivalent
reduction at lower economic energy or environmental cost; (2) the-proposed
system has not been adequately demonstrated; (3) the technology will not
2-5
-------
cause or contribute to an unreasonable risk to the public health,
welfare or safety; (4) the governor of the State where the source is
located consents; and that, (5) the waiver will not prevent the attainment
or maintenance of any ambient standard. A waiver may have conditions
attached to assure the source will not prevent attainment of any NAAQS.
Any such condition will have the force of a performance standard.
Finally, waivers have definite end dates and may be terminated earlier
if the conditions are not met or if the system fails to perform as
expected. In such a case, the source may be given up to three years to
meet the standards, with a mandatory progress schedule.
2.2 SELECTION OF CATEGORIES OF STATIONARY SOURCES
Section 111 of the Act directs the Adminstrator to list categories
of stationary sources which have not been listed before. The Adminstrator,
". . . shall include a category of sources in such list if in his judgement
it causes, or contributes significantly to, air pollution which may
reasonably be anticipated to endanger public health or welfare." Proposal
and promulgation of standards of performance are to follow while adhering
to the schedule referred to earlier.
Since passage of the Clean Air Amendments of 1970, considerable
attention has been given to the development of a system for assigning
priorities to various source categories. The approach specifies areas
of interest by considering the broad strategy of the Agency for implementing
the Clean Air Act. Often, these "areas" are actually pollutants which
are emitted by stationary sources. Source categories which emit these
pollutants were then evaluated and ranked by a process involving such
2-6
-------
factors as (.1) the level of emission control (if any) already required
by State regulations; (2) estimated levels of control that might be
required from standards of performance for the source category; (3)
projections of growth and replacement of existing facilities for the
source category; and (4) the estimated incremental amount of air pollution
that could be prevented, in a pre-selected future year, by standards of
performance for the source category. Sources for which new source
performance standards were promulgated or are under development during
1977 or earlier, were selected on these criteria.
The Act amendments of August, 1977, establish specific criteria to
be used in determining priorties for all source categories not yet
listed by EPA. These are:
1) the quantity of air pollutant emissions which each such category
will emit, or will be designed to emit;
2) the extent to which each such pollutant may reasonably be anticipated
to endanger public health or welfare; and
3) the mobility and competitive nature of each such category ..of
sources and the consequent need for nationally applicable new source
standards of performance. :
In some cases, it may not be feasible to immediately develop a
standard for a source category with a high priority. This might happen
when a program of research is needed to develop control techniques or
because techniques for sampling and measuring emissions may require
refinement. In the developing of standards, differences in the time
required to complete the necessary investigation for different source
categories must also be considered. For example, substantially more
time may be necessary if numerous pollutants must be investigated from a
2-7
-------
single source category. Further, even late in the development process
the schedule for completion of a standard may change. For example,
inablility to obtain emission data from well-controlled sources in time
to pursue the development process in a systematic fashion may force a
change in,scheduling.. Nevertheless,,priority ranking .is, and will
continue to be, used to establish the order in which projects are initiated
and resources assigned.
After the source category has been chosen, determining the types of
facilities within the source category to which the standard will apply
must be decided. A source category may have several facilities that
cause air pollution and emissions from some of these facilities may be
insignificant or very expensive to control. Economic studies of the
source category and of applicable control technology may show that air
pollution control is better served, by applying standards to the more
severe pollution sources. For this reason, and because there be no
adequately demonstrated system for controlling emissions from certain
facilities, standards often do not apply to all facilities at a source.
For the same reasons, the standards may not apply to all air pollutants
emitted. Thus, although a source category may be selected to be covered
by a standard of performance, not all pollutants or facilities within
that source category may be covered by the standards.
2.3 PROCEDURE FOR DEVELOPMENT OF STANDARDS OF PERFORMANCE
Standards of performance must (1) realistically reflect best demon-
strated control practice; (2) adequately consider the cost, and the
nonair quality health and environmental impacts and energy requirements
of such control; (3) be applicable to existing sources that are modified
2-8
-------
or reconstructed as well as new installations; and (4) meet these conditions
for all variations of operating conditions being considered anywhere in
the country.
The objective of a program for development of standards is to
identify the best technological system of continuous emission reduction
which has been adequately demonstrated. "The legislative history of
section 111 and various court decisions make clear that the Administrator's
judgement of what is adequately demonstrated is not limited to systems
that are in actual routine use. The search may include a technical
assessment of control systems which have been adequately demonstrated
but for which there is limited operational experience. In most cases,
determination of the ". . . degree of emission reduction achievable ."
is based on results of tests of emissions from well-controlled existing
sources. At times, this has required the investigation and measurement
of emissions from control systems found in other industrialized countries
that have developed more effective systems of control than those available
in the United States.
Since the best demonstrated systems of emission reduction may not
be in widespread use, the data base upon which standards are developed may
be somewhat limited. Test data on existing well-controlled sources are
obvious starting points in developing emission limits for new sources.
However, since the control of existing sources generally represents
retrofit technology or was originally designed to meet an existing State
or local regulation, new sources may be able to meet more stringent
emission standards. Accordingly, other information must be considered
before a judgement can be made as to the level at which the emission
standard should be set.
2-9
-------
A process for the development of a standard has evolved which takes
into account the following considerations:
1. Emissions from existing well-controlled sources as measured.
2. Data on emissions from such sources are assessed with considera-
tion of such factors as: (a) how representative the tested source is in
regard to feedstock, operation, size, age, etc.; (b) age and maintenance
of control equipment tested; (c) design uncertainties of control equipment
being considered; and (d) the degree of uncertainty that new sources
will be able to achieve similar levels of control.
3. Information from pilot and prototype installations, guarantees
by vendors of control equipment, unconstructed but contracted projects,
foreign technology, and published literature are also considered during
the standard development process. This is especially important for
sources where "emerging" technology appears to be a significant alternative.
4. Where possible, standards are developed which permit the use of
more than one control technique or licensed process.
5. Where possible, standards are developed to encourage or permit
the use of process modifications or new processes as a method of control
rather than "add-on" systems of air pollution control.
6. In appropriate cases, standards are developed to permit the use
of systems capable of controlling more than one pollutant. As an example,
a scrubber can remove both gaseous and particulate emissions, but an
electrostatic precipitator is specific to particulate matter.
7. Where appropriate, standards for visible emissions are developed
in conjunction with concentration/mass emission standards. The opacity
standard is established at a level that will require proper operation
and maintenance of the emission control system installed to meet the
2-10
-------
concentration/mass standard on a day-to-day basis. In some cases,
however, it is not possible to develop concentration/mass standards,
such as with fugitive sources of emissions. In these cases, only opacity
standards may be developed to limit emissions.
2.4 CONSIDERATION OF COSTS
Section 317 of the Act requires, among other things, an economic
impact assessment with respect to any standard of performance established
under section 111 of the Act. The assessment is required to contain an
analysis of:
(1) the costs of compliance with the regulation and standard including
the extent to which the cost of compliance varies depending on the
effective date of the standard or regulation and the development of less
expensive or more efficient methods of compliance;
(2) the potential inflationary recessionary effects of the standard
or regulation;
(3) the effects on competition of the standard or regulation with
respect to small business;
(4) the effects of the standard or regulation on consumer cost;
and,
(5) the effects of the standard or regulation on energy use.
Section 317 requires that the economic impact assessment be as
extensive as practicable, taking into account the time and resources
available to EPA. , .
The economic impact of a proposed standard upon an industry is
usually addressed both in absolute terms and by comparison with the
control costs that would be incurred as a result of compliance with
typical existing State control regulations. An incremental approach is
2-11
-------
taken since both new and existing plants would be required to comply
with State regulations in the absence of a Federal standard of performance.
This approach requires a detailed analysis of the impact upon the industry
resulting from the cost differential that exists between a standard of
performance and the typical State standard.
The costs for control of air pollutants are not the only costs
considered. Total environmental costs for control of water pollutants
as well as air pollutants are analyzed wherever possible.
A thorough study of the profitability and price-setting mechanisms
of the industry is essential to the analysis so that an accurate estimate
of potential adverse economic impacts can be made. It is also essential
to know the capital requirements placed on plants in the absence of
Federal standards of performance so that the additional capital requirements
necessitated by these standards can be placed in the proper perspective.
Finally, it is necessary to recognize any constraints on capital availability
within an industry, as this factor also influences the ability of new
plants to generate the capital required for installation of additional
control equipment needed to meet the standards of performance.
2.5 CONSIDERATION OF ENVIRONMENTAL IMPACTS
Section 102(2)(C) of the National Environmental Policy Act (NEPA)
of 1969 requires Federal Agencies to prepare detailed environmental
impact statements on proposals for legislation and other major Federal
actions significantly affecting the quality of the human environment.
The objective of NEPA is to build into the decision-making process of
Federal agencies a careful consideration of all enviornmental aspects of
proposed actions.
2-12
-------
In a number of legal challenges to standards of performance for
various industries, Federal Courts' of Appeals have held that environmental
impact statements need not be prepared by the Agency for proposed actions
under section 111 of the Clean Air Act. Essentially, Federal Courts of
Appeals have determined that ". . . the best system of emission reduction,
. . . require(s) the Administrator to take into account counter-productive
environmental effects of a proposed standard, as well as economic costs
to the industry. . ." On this basis, therefore, the Courts ". .
established a narrow exemption from NEPA for EPA determination under
section 111."
In addition to these judicial determinations, the Energy Supply and
Environmental Coordination Act (ESECA) of 1974 (PL-93-319) specifically
exempted proposed actions under the Clean Air Act from NEPA requirements.
According to section 7(c)(l), "No action taken under the Clean Air Act
shall be deemed a major Federal action significantly affecting the quality
of the human environment within the meaning of the National Environmental
Policy Act of 1969."
The Agency has concluded, however, that the preparation of environmental
impact statements could have beneficial effects on certain regulatory
actions. Consequently, while not legally required to do so by section
102(2)(C) of NEPA, environmental impact statements will be prepared for
various regulatory actions, inlcuding standards of performance developed
under section 111 of the Act. This voluntary preparation of environmental
impact statements, however, in no way legally subjects the Agency to
NEPA requirements.
To implement this policy, a separate section is included in this
document which is devoted solely to an analysis of the potential environmental
2-13
-------
impacts associated with the proposed standards. Both adverse and bene-
ficial impacts in such areas as air and water pollution, increased solid
waste disposal, and increased energy consumption are identified and
discussed.
2.6 IMPACT ON EXISTING SOURCES
Section 111 of the Act defines a new sources as ". . . any stationary
source, the construction or modification of which is commenced ..."
after the proposed standards are published. An existing source becomes
a new source if the source is modified or is reconstructed. Both modification
and reconstruction are defined in amendments to the general provisions
of Subpart A of 40 CFR Part 60 which were promulgated in the Federal
Register on December 16, 1975 (40 FR 58416). Any physical or operational
change to an existing facility which results in an increase in the
emission rate of any pollutant for which a standard applies is considered
a modification. Reconstruction, on the other hand, means the replacement
of components of an existing facility to the extent that the fixed
capital cost exceeds 50 percent of the cost of constructing a comparable
entirely new source and that it be technically and economically feasible
to meet the applicable standards. In such cases, reconstruction is
equivalent to a new construction.
Promulgation of a standard of performance requires States to establish
standards of performance for existing sources in the same industry under
section m(.d) of the Act if the standard for new sources limits emissions
of a designated pollutant (i.e., a p.ollutant for which air quality
criteria have not been issued under section 108 or which has not been
listed as a hazardous pollutant under section 112). If a State does not
2-14
-------
act, EPA must establish such standards. General provisions outlining
procedures for control of existing sources under section lll(d) were
promulgated on November 17, 1975, as Subpart B of 40 CFR Part 60 (40 FR
53340). :
2.7 REVISION OF STANDARDS OF PERFORMANCE
Congress was aware that the level of air pollution control achievable
by any industry may improve with technological advances. Accordingly,
section 111 of the Act provides that the Administrator ". . . shall, at
least every four years, review and, if appropriate, revise ..." the
standards. Revisions are made to assure that the standards continue to
reflect the best systems that become available in the future. Such
revisions will not be retroactive but will apply to stationary sources
constructed or modified after the proposal of the revised standards.
2-15
-------
-------
3.0 THE GLASS MANUFACTURING INDUSTRY
3.1 GENERAL
3.1.1 INTRODUCTION
This chapter presents prominent features of the glass manufac-
turing industry including a description of the glass producing
processes and a discussion of atmospheric emissions. This chapter
focuses primarily on particulate matter emitted from the glass
melting furnace. Other emissions are discussed briefly.
3.1.2 GLASS MANUFACTURING INDUSTRY STATISTICS
The glass manufacturing industry is classified in accordance
with the industry definitions embodied in the Standard Industrial
Classification (SIC) system. Under this system of classification,
an industry is generally defined as a group of establishments pro-
ducing a single product or a more or less closely related group of
products. Accordingly, for the glass industry there are four SIC
codes:
SIC 3211 Flat glass
SIC 3221 Container glass
SIC 3229 Pressed and blown glass, not eleswhere
classified (N.E.C.)
SIC 3296 Wool fiberglass
Glass manufacturing facilities are located throughout the United
States and are usually situated in areas that ensure the availa-
bility of raw materials. These plants are found in 34 states with
almost three-quarters of these plants in the following 10 states:
California, Illinois, Indiana, New Jersey, New York, Ohio, Oklahoma,
Pennsylvania, Texas, and West Virginia. In early 1978 there are
129 primary glass-producing companies which together operate 338
individual plants. Placing these plants in SIC categories, there
are 32'flat glass plants (SIC 3211); 117 container glass plants
3-1
-------
(SIC 3221); 165 pressed and blown, N.E.C., plants including 13
textile fiberglass plants (SIC 3229); and 24 wool fiberglass plants
(SIC 3296). Appendix D lists, by state, individual glass plants
and the SIC designations for each plant.1'2'3'4'5
Recent production rates and dollar values of shipments for each
segment of the industry are summarized in Table s-i.6*7'8'9*10'11'12
A significant result of these statistics shows that, assuming 77
percent of glass produced in the pressed and blown is soda-lime
glass as it was in 1973, over 90 percent of the total glass pro-
duced in 1976 is soda-lime glass. Additionally, the figures on
this table form the bases of the industry growth predictions
derived in Section 3.1.3.
3.1.3 INDUSTRY GROWTH PROJECTIONS
In this section, the anticipated growth of the glass industry
is estimated by a straight line formulation used in previous glass
manufacturing projections. Table 3-2 presents the projected
weight of glass produced in 1985 by each segment of the industry.
The annual growth rates used in these calculations were determined
in a previous study and reflected the growth, from 1958 to 1972,
15
of the dollar value of glass products corrected for inflation.
Although these corrected dollar values do not strictly apply to
the weight of the glass produced, it is assumed that, within the
accuracy of the projection scheme, these annual growth rate dollar
values can be used to predict glass weight production rates for the
period from 1976 to 1985.
3.2 GLASS MANUFACTURING PROCESSES AND THEIR EMISSIONS
Although there are numerous unit operations used in the manu-
facturing of glass, most key processing steps, which generate the
largest amounts of atmospheric emissions, are common throughout
the industry. In this subsection, the basic operations and the
3-2
-------
Table 3-1. 1976 PRODUCTION RATES AND VALUES OF SHIPMENTS
Segment
Flat Glass
Container
Glass
Pressed and
Blown
(N.E.C.)
Wool
Fiberglass
SIC
Code
3211
3221
3229,
3296
Production Rate
in 1976a
2.56 Tg (2.91 MM Tons)6
11.8 Tg (13.0 MM Tons)8
£
1.73 Tg (1.95 MM Tons)0
0.896 Tg (0.986 MM Tons)11
Dollar Value of
Shipments in 1976
(In Millions of Dollars)
6457
3,2519
1,59810
i f\
81712
a 12
Tg is an abbreviation for 10 grams. MM tons
represents one million tons.
Table 3-2. PROJECTED 1985 PRODUCTION RATES
Segment
Flat
Container
Pressed and Blown
(N.E.C.)
Wool Fiberglass
SIC
Code
3211
3221
3229
3296
Annual
Growth Rate
(Percent)
1.816
3.116
3.516
7. 117
1985 Production Rate
Tg
3.1
15.0
2.3
1.5
(MM Tons)
( 3.4)
(17.0)
( 2.5)
( 1.6)
3-3
-------
major pollutants they emit are identified and assessed. Special
types of unit operations or additional pollutants which signifi-
cantly affect the total emissions of the primary glass industry
are discussed.
3.2.1 BASIC PROCESS
Glass is manufactured in a high termperature conversion of
raw materials into a homogeneous melt capable of fabrication into
useful articles. This process can be broken down into three sub-
processes: raw material handling and mixing; melting; and forming
and finishing. Emission points within each subprocess are dis-
cussed in Section 3.2.2. Figure 3-1 gives a typical flow diagram
for the manufacture of soda-lime glass; however, it has general
application to other commercial glass formulations.
3.2.1.1 Raw Material Handling and Mixing
The raw materials are received in packages or in bulk and are
unloaded by hand, vibrator-gravity, drag shovels, or vacuum systems.
Cullet, crushed recycled glass, must be segregated and transferred
to storage bins by various processes including the utilization of
bucket elevators, belt conveyors, or screw conveyors. In addition to
bulk raw materials, certain minor constituents are packaged and
stored in their original containers until mixed with the batch.
Prior to being fed into the melting unit, the raw materials
are mixed according to the desired product recipe. Weighing and
mixing operations may be automated or carried out by hand depend-
ing on the size or specialty of the operation. The melters
themselves are charged manually or automatically, usually through
screw or reciprocating-type feeders.
Emissions generated in raw material and handling processes
are discussed in Section 3.2.2.1.
3-4
-------
Class tand
SiOj g 99S
to yield Si02
crushed. washed
and screened
to 20-100
TneSh
^_
Soda ash
Na2C03
to yield NarO
~ 20-120 mesh
or granular
limestone
or burnt lime
to yield CaO.
Usually some
MgO also results
20-120 mesh
Feldspar
RjO.A:?03 eS.Oj
to yield
A!j03.S.O.
N»jO and K;0
pulverized or
granular
Other additions
for K?0. MgO.
2nO, BaO, PbO,
etc and those for
fining, oxidizing.
coloring and
decolorizing
Side-port
continuous tank,
looking down
through top
Submerged
throat in
brtdgewall
Crushed cutlet
of same
composition
as that to
be melted
Cooling
Temperature 1300 °C
Distributing
Cutlet
crushing
Temperature = 800-1100°C
depending on article-
*nd process
Forming hot. viscous glass
shaped by pressing.
blowing drawing or rolling
Inspection and
product testing
60-90 minutes in
contmuous-bel: tunne'.
lehr; hot zone500 "C
Packing, warehousing.
and shipping
Figure 3-1. Typical Flow Diagram for the Manufacture of Soda-Lime Glass
Reproduced from Encyclopedia of Chemical Technology, 2nd Edition,
Volume 10, pg. 549, 1966 . .
3-5
-------
3.2.1.2 The Melting Process
From the handling area, the weighed raw materials are deliv-
ered into the furnace where they are transformed through a sequence
of chemical reactions into glass. In operation, the raw materials
float on the bed of molten glass until they dissolve. Mixing in
the molten glass bed"is effected by gases evolved in chemical
reactions and by natural convection currents in the molten glass
bed. In addition-, some furnaces have air injected in the bottom
of the bed to augment ebulient mixing.
Within the temperature range of the furnace (nominally
1,500°C to 1,700°C), the glass exists as a liquid free of crystal-
line matter with a viscosity of 10 Newton-seconds per square meter
(N-s/m2) which is equivalent to 100 poise. Because the viscosity
of the glass exiting the furnace must be compatible with the form-
ing operations, the temperature of the molten bed is decreased
gradually to a point until the viscosity of the glass is about 100
to 1,000 N-s/m2 (103-104 poise). In addition to cooling as it
flows through the furnace, the glass is retained in the furnace
long enough so that gaseous inclusions can be removed. From the
introduction of raw materials to the extraction of a homogeneous
melt suitably ready for forming, a furnace accomplishes three
functions in glassmaking: to bring raw materials together to
react; to hold the molten glass until it is free of bubbles and
inclusions; and to condition the glass for forming.
Energy required for melting glass is supplied by burning
either natural gas or fuel oil and sometimes by augmenting the
energy produced from these fossil fuels with electricity which is
converted to heat within the liquid glass bed. A recent study
shows that these three energy sources provide over 99 percent of
the energy consumed in the industry. Although this study inven-
tories all energy utilized and therefore includes energy expended
on processes other than melting, consumption of energy in the
3-6
-------
furnace predominates consumption in all other areas. Data for
1976 show that natural gas provides 74 percent of the energy, fuel
oil 14 percent, and electricity 11 percent. Natural gas has been
the preferred primary fuel with fuel oil used when natural gas
supplies were curtailed. This pattern is expected to change soon
with fue.l oil becoming the predominant fossil fuel used in furnace
firing. Coal is not presently burned or gasified to provide heat
to the glass industry. Although a successful pilot study has been
20
made, coal is not expected to supply a significant amount of
energy to the industry in the future.
There are three types of fossil fuel-fired melting units used
in the glass industry: day pots, day tanks, and continuous tanks.
Typically, day pots are used where other larger tanks are not
economically justified because of limited production of special
compositions of glass. The range of the capacities of day pots
varies from 9 kilograms (20 pounds) to 1,800 kilograms (2 tons)
with these quantities melted in 24-hour batches. Although the
capacity of a typical day tank is slightly larger than that of a
typical day pot, the primary distinction between the two units is
the material of construction of the vessel walls.
Although day pots and day tanks are used to produce glass,
most glass tonnage is melted in larger capacity, continuously
operating regenerative or recuperative furnaces. Of these two,
there are more regenerative units producing glass than recuperative
units. Generally, regenerative furnaces maintain a larger produc-
tion rate than recuperative furnaces. These types of furnaces
differ in the types and modes of operation of the heat exchangers
used to recover heat from the furnace exhaust gases.
Regenerative furnaces utilize two chambers of refractory
called checkerworks in the following manner: at any one time,
while combustion flue gases heat the refractory in one checkerwork
chamber, the other checkerwork preheats combustion air; then after
3-7
-------
intervals ranging from 10 to 30 minutes, this gas flow is diverted
so that combustion air is drawn through the chamber previously
heated by flue gases, and flue gases heat the refractory in the
other chamber previously used to preheat combustion air. Regener-
ative furnaces, themselves, are divided into two functional
categories side-port and end-port furnaces depending on the
furnace flame firing pattern. Figure 3-2 illustrates a typical
side-port furnace (called "side-fired" in the figure) and an end-
port furnace (called "end-fired" in the figure).
Recuperative furnaces employ one continuously operating shell
and tube type heat exchanger to preheat combustion air instead of
the checkerwork heat exchangers used in regenerative furnaces.
In existing furnaces where fossil fuel provides the bulk of
energy consumed, electricity is often used to supply energy needed
to increase production. Additionally, in specially designed fur-
naces called "all-electric melters," electricity has also been
used as the sole energy source for glass production after the
liquefaction of a glass bed.
The significant emission point in glass melting furnace opera-
tions is the stack where combustion gases are released to the
atmosphere. Emissions generated in melting furnaces are discussed
in Section 3.2.2.2.
3.2.1.3 Forming and Finishing
In the forming and finishing step, the molten glass is
extracted from the furnace, shaped to the desired form, and then
annealed at high temperature. The final product is then either
inspected and shipped or sent for further finishing such as temper-
ing or decorating. . .. ,
In practice, the molten glass, while at a yellow-orange temper-
ature, is drawn quickly from the furnace and worked in forming
3-8
-------
END-FIRED BOX TYPE REGENERATOR GLASS
.FURNACE
SIDE-FIRED BOX TYPE REGENER/
TOR GLASS FURNACE
Figure 3-2. Typical Side-Port Furnace and End-Port Furnace
Courtesy of G.P.I.
3-9
-------
machines by a variety of methods: pressing; blowing in molds; and
drawing, rolling, and casting. Immediately, this formed glass is
conveyed to continuous annealing ovens to remove internal stresses
in the glass by controlled cooling. Figure 3-3, illustrates a
typical forming operation in the glass container industry.
Emissions generated in forming and finishing operations are
discussed in Section 3.2.2.3.
3.2.2 PROCESSES AND THEIR EMISSIONS
The material in this section describes the nature of air
pollutants generated in the uncontrolled manufacture of glass pro-
ducts and quantifies the levels of these pollutants released to
the atmosphere. As substantiated in these sections, emissions
from furnace operations are greater by two orders of magnitude
than those from raw material handling and from forming and finish-
ing operations. Examples of regulations of several states limiting
particulate emissions are also presented.
3.2.2.1 Emissions From Raw Material Handling Operations
Emissions from this glassmaking subprocess are limited to
solid particles becoming airborne in the movement or storage of
bulk raw materials. Since no chemical reactions occur between
these materials at ambient conditions, the chemical composition of
the fugitive dusts is the same as the raw materials from which
they were entrained.
The glass industry uniformly contains fugitive dust emissions
by enclosing the unloading and conveying areas and controls them
by venting storage areas through fabric filters. Because the
emission factors for controlled raw material handling operations
are, on the average,22'23'24 two orders of magnitude less than
those'for glass melting, these fugitive emissions will not be
discussed further.
3-10
-------
XMwtry
Settle Mow
Counter
Transit' from blank mold to blow maid
Final
Figure 3-3. Example of One Type of Forming Used in Container Glass
Production
Reproduced from Encyclopedia of Chemical Technology, 2nd Edition,
Volume 10, pg. 559, 1966
3-11
-------
3.2.2.2 Emissions From Uncontrolled Glass Furnace Operations
By weight, the major pollutants emitted from fossil fuel-fired
furnaces producing soda-lime glass are oxides of nitrogen, oxides
of sulfur, and submicron-sized particulates. Carbon monoxide and
hydrocarbons have been detected in furnace exhaust but in much
lesser quantities. Pollutants such as arsenic, borates, fluorides,
and lead, which are emitted in the exhaust of pressed and blown
glass, make significant contributions to the emissions of the glass
industry.
Pollutants generated in the melting of glass arise from two
sources: from the combustion of fuel and from the vaporization
of raw materials. The pollutants formed in the combustion of fuel
are NO , SO , CO, and hydrocarbon. Nitrogen oxides constitute the
largest mass emission from glass melting furnaces. This gas phase
pollutant forms from the reaction between nitrogen and oxygen at
high temperatures. Two sources can provide the nitrogen for the
reaction: nitrogen gas present in combustion air and nitrogen
contained chemically in the fuel burned. Values taken from the
National Emission Data System (NEDS)25 show emissions varying from
less than 1 up to 10 grams of NOX (calculated at N02) emitted per
kilogram of glass produced (2 to 20 Ib/ton). The emissions for
container glass and flat glass averaged 3.94*> and 3.82 g/kg N02
(7.88 and 7.64 Ib/ton) with standard deviations close to 3 g/kg
(6 Ib/ton); the emissions for the pressed and blown sector averaged
to a value of 4.24 g/kg28 (8.50 Ib/ton) with a standard deviation
of about 2 g/kg (4 Ib/ton). Standard deviation is an index of the
dispersion of data about the average value; two-thirds of the data
cluster within one standard deviation on both sides of the average
value. The few NOX emissions reported for fiberglass fall within
the above range of values (1 to 10 grams per kilogram). The amount
of nitric oxide being formed in a glass melting furnace may vary
among glass formulations; the formulations requiring higher
3-12
-------
firebox temperatures generate more nitric oxide than those formu-
lations- needing lower temperatures.
Some of the sulfur oxides emitted from the manufacture of
glass are caused by the oxidation of sulfur compounds in the fossil
fuels. All of the source assessment authors for the different
categories of the glass industry state that sulfur present in fuel
?Q 30 "31
oil leaves the glass melting furnace as SCL. Figures
reported in the literature estimate that a fuel oil containing 1
percent sulfur by weight will yield approximately 600 ppm of S00
32
in the flue gas. The authors also note that since the NEDS data
were measured when natural gas was universally used as a glass
melting fuel, the effect of shifting fuels from natural gas (with
no sulfur) to fuel oil will be increased sulfur dioxide emissions.
The remainder of the SO^ emissions form from the volatization and
reaction of raw materials and will be discussed later with other
pollutants from that source.
Both carbon monoxide and hydrocarbon emissions from glass pro-
ducing furnaces are due primarily to the incomplete combustion of
fossil fuel and, additionally, to the decomposition of powdered
coal added to the batch materials. With highly efficient furnace
operating practices uncontrolled emission rates of these pollutants
have been kept at low levels. Established emission factors for CO
for flat, container, and pressed and blown glass production are:
0.02 g/kg (0.04 lb/ton),33 0.06 g/kg (0.12 lb/ton),34 and 1.10 g/kg
OC
(2.20 Ib/ton), respectively. Emission factors for hydrocarbons
are reported to be 0.04 g/kg (0.08 Ib/ton),36 0.08 g/kg (0.16 lb/
ton),37 and 0.15 g/kg (0.30 Ib/ton),38 respectively, for these
types of glass production. It is reported that both CO and hydro-
carbon emission factors should be independent of the type of glass
produced.
The other source of air pollutants from the glass melting
furnace is the vaporization of raw materials from the glass melt.
3-13
-------
The major raw materials used in all types of glass manufacturing
are:glass, sand, soda, ash, limestone, and culled (Gullet con-
sists of recycled, crushed glass.) Typical raw material batch
recipes for several types of product glasses are given in Table
3-3.39
The common oxides listed in the table can be cateqorized as
formers, fluxes, and stablizers. By themselves, formers account
for the random three-dimensional atomic structure characteristic
of glass. Fluxes are added to lower the melting points and the
working temperatures which must be maintained in the furnace.
Stabilizers improve the chemical durability of the glass product
by lowering the coefficient of expansion and preventing glass
crystallization. Of the raw materials listed in Table 3-3, the
borates increase the thermal durability of the glass product by
lowering the coefficient of expansion; lead increases the refrac-
tive index and density; aluminum increases glass strength; feld-
spar, reportedly, lowers the mixture melting point and prevents
devitrification; sodium accelerates the melting process; and
arsenic compounds aid in fining (removing bubbles from the melt).
In addition to these compounds, trace amounts of various metal
oxides are added to the batch to change the color of the glass by
either imparting a color or neutralizing the tints caused by batch
contaminants.
These raw materials react in the molten bed of soda-lime glass
in the furnace releasing carbon dioxide in the following proposed
40
reaction sequence:
Na2C03
CaCO
/3Si02
ySiO,
-> CaO |3SiO(
Na20*XSi02
co2
co2
CO +
SO,
3-14
-------
Table 3-3. RAW MATERIAL BATCH RECIPES
Raw Material
Sand (S102)
Limestone (CaC03)
Dolomite (CaCQ3«MgC03)
Soda Ash (Na2C03)
Potassium Carbonate (K2C03)
Red Lead (Pb304)
Nepheline Syenite (25% A1907; 15% Na~0;
60% Si02) * J *
Felspar (K20»Al20»6Si02)
Anhydrous Borax (Na20«2B203)
Boric Acid (B203»3H20)
Sodium Sulfate (Na2 SO^)
Sodium Nitrate (NaN03)
Sodium Chloride (Nad)
Arsenious Oxide (As^O,)
Carbon (C)
Total Weight
Weight
in
Soda-
Lime
Container
Glass
2,000
490
650
200
15
15
3,370
Weight
in
Soda-
Lime
Sheet
Glass
2,000
50
480
680
150
60
3
3,423
Weight
in
Boro-
silicate
Glass
2,000
190
350
230
0.5
1
1
2,772.5
Weight
in
Lead
Crystal
2,000
290
1,310
3,600
Note: Weight of materials is measured in pounds based on 2,000 pounds of dry sand.
3-15
-------
As mentioned previously, chemical side reactions occurring in the
furnace lead to the formation of gaseous sulfur oxides from sodium
41
sulfate. The proposed pathway is:
Na2S04 1000°C»Na20 + S03
The major type of air pollutant released in glass melting is
parti oil ate which vaporizes from the molten glass surface and con-
denses at lower temperatures in the checkerwork or in the stack.
Since testing of this particulate shows no silica compounds, no
contribution to the particulate emissions by raw materials being
entrained in the flue gases and then being swept from the furnace
is made by batch carryover. Testing does show that the largest
percentage of the particulate from soda-lime glass is sodium
sulfate which is predicted in the following reaction sequence
42
Na20
H20-
SO, + 1/2 0
-2NaOH (occurring in a vapor state over
the melt)
(occurring in a vapor state over
the melt)
2NaOH + SO, ^Na9SO/, + H?0 (occurring in a checker or
d * exhaust system with the
product in a liquid state)
The submicron size distribution of the particulate reinforces the
condensation mechanism. In general, testing shows that over 75
percent of the1 particualte catch has a characteristic size of
43
less than one micron.
As with soda-lime glass, the chemical composition of the par-
ticulate emitted from the manufacturing of other formulations of
glass depends on the raw materials processed through the. furnace.
The particualte emitted in borosilicate glass manufacture consists
of boric acid and alkali borates. In the production of opal glass
B203, NaF, and Na2SiF6 appear in the particulate catch. For lead
glass production in a natural gas-fired furnace, the chemical
3-16
-------
composition of the participate is lead oxide and lead sulfate.44
Although the use of arsenic as a fining and decolorizing agent in
flat and container glass has been reduced, arsenic is still used
where it is required in the product specifications for glasses
classified as pressed and blown products. Arsenic has also been
assayed in small levels in the particulate emitted from these
45
furnaces. Boron and fluoride compounds are also found in the
melting furnace emissions from wool and textile fiberglass manu-
46
facturing. Of the fluorine fed to the furnace in the raw mater-
ials, some may leave as a gaseous compound, HF, or as a particulate.
The fluoride species which have been detected are: HF, NaF, PbF0,
47
BFg, SiF^, and Na2SiFg. In recent testings of uncontrolled glass
furnaces, up to 10 percent of the fluorine added in the feed was
4R
measured in the furnace exhaust.
In addition to the nature of particulate emissions, depending
on the kinds and amounts of raw materials fed to the furnace,
operating parameter values affect the levels of pollutants emitted
from the glass furnace. Key operating parameters are: the furnace
(or bridgewall) temperature, the amount of cullet in the raw batch,
the use of electric boosting, the surface area of the molten glass
bed, the production (or pull) rate of glass exiting the furnace,
and the type of fuel being burned. Of these operational variables
for a fixed glass composition, temperature is the signal parameter.
Increasing the temperature over the melt vaporizes more of the
volatile materials than at lower temperatures. Maintaining high
temperature requires more fuel to be consumed and should, therefore,
increase the levels of pollutants derived from fossil fuels NO
X
and, if sulfur containing fuel oil is burned, SO . Other parameters
A
previously listed influence pollutant emission levels by changing
the temperature required to maintain production. For example,
increasing the cullet proportions in the raw batch lowers bridge-
wall temperature, thereby, lowering emissions. In the same way,
3-17
-------
electric boosting lowers the furnace temperature required to main-
tain glass production thereby resulting in less emissions than at
an identical production rate without electric boosting.
The amount of surface area of molten glass exposed to combus-
tion gases has been shown to affect particulate emissions. With
all other parameters constant, a larger exposed area generates
49
more particulate than a smaller area.
For a furnace producing a single type of glass, increasing
the pull rate requires more energy, which if supplied by the com-
bustion of fossil fuels', causes an increase in furnace temperature'
with a concomitant increase in emissions. The dependence of emis-
sion rates on furnace throughput is well established. This depend-
ence is incorporated within the compliance regulations of several
states. These compliance regulations show an exponential dependence
of less than one for the weight rate emissions versus weight through-
put rate, indicating that particulate emissions per kilogram of glass
produced decrease as production rate increases. In the limiting
case of no pull rate, data have been published which show that
particulates are still emitted from the molten glass bed. For
this case, the emission levels at zero pull rate were roughly 20
percent of those at the normal pull rate with both measurements
being taken at the same temperature.
The glass manufacturing industry is divided into natural
categories which basically follow the Standard Industrial Classifi-
cations. This classification differentiates the following glass
manufacturing types: Flat Glass, Container Glass, Pressed and
Blown Glass (including glass products not classified elsewhere),
and the Wool Fiberglass portion of SIC 3296. Examples of end-
products from each classification are listed in Appendix E.
Of these SIC designations only the Pressed and Blown classi-
fication needs to be modified to adequately describe the industry.
Based on amounts of particulate emitted from melting furnaces, two
3-18
-------
glass formulation types are distinguishable a type consisting
of soda-lime formulations and a type comprising the remaining
glass formulations in this SIC (borosilicate, opal, lead, and
other glass formulations). Based on information from the glass
industry, two Pressed and Blown model plant production rates
characterize the facilities in this SIC. The production rates
correspond to small and moderate size glass melting furnaces. The
glass formulation types are utilized in this subsection to quantify
the uncontrolled melting furnace emissions and in Chapter 6.0 to
set regulatory options. The production rate characterizations are
used in Chapter 7.0 to determine environmental impacts.
Gaseous and particulate emissions from uncontrolled glass
melting furnaces are depicted in Table 3-4 for each industry
category. Values of gaseous emissions are taken from source
51 52 5"3 54 55
assessment documents. '»'* Particulate emissions are
based:on the results of emission tests performed for the EPA; on
the results of emission tests provided by the glass industry in
response to questionnaires I and on the emissions reported in source
assessments of the screening of study documents..
From emission test results of Flat Glass furnaces, particu-
late emissions range from roughly 0.9 grams per kilogram of glass
produced to roughly 1.1 g/kg (1.8 to 2.2 Ib/ton). The table entry
of 1.5 g/kg (3.0 Ib/ton) represents a conservative numerical value
of emissions from this industry category. For the Container Glass
category, the tabled value of 1.25 g/kg (2.5 Ib/ton) reflects the
average value of 1.15 g/kg (2.3 Ib.ton) measured in eight uncon-
57
trolled Container Glass melting furnaces. Based on the similar-
ity of the glass formulation of soda-lime glass manufactured in
Pressed and Blown furnaces as compared with that of soda-lime
glass manufactured in Container furnaces and based on the equiva-
lence 'of furnace configurations and sizes in these two glass
manufacturing categories, the particulate emissions for the Pressed
3-19
-------
o:
o
CD
CJ
DC
to
o:
o
to
LU
to
CO
a
Ixl
ce
u.
CO
o
CO
CO
I
to
IE
V-
cu
in
tu
o
c
o
k.
re
u
o
I
S
x
vt
fi»
o
3
w
u
K
§
[Industry
Category
5
,0
**
en
.x
en
e*
o
XI
O
cn
en
"c?
o
xT
en
en
_
o
^
£
JC
en
I
-O
s
en
Q
*»
XI
e»
en
c
o
^
c.
e
en
§
0
_
o
e
*"*
CM
o
O
3-
"""""
in
»-
o
n
«-
o
CD
**
O
"
re
i/i
in
re
i"
a
s
§
o
^,
CO
to
o-
"""
in
§
S
Cvl
"**
^.
«-
in
*
CM
in
CM
«-
1C
^
*""
n
CM
U
a>
C A
» in
re in
«j re
§ o,
-^
o
in
o
CM
o
10
«n
in
S
cu
^^
*r>
CM
in
CM
m
GO
*^
in
CM
"
1 i
re -
o c"7
a; 5 re
m o -o
V* *~- O
0 CD m
Q.
O
£!
o
cc
z
a:
§
in
**
en
en
e> c k. i
oxo re
in o x: ^3
m ^ *^ o
O O O *A
u
o.
«M
«^
§
tx
z
in
CM
in
CM
o
e
in
_^
o
-*
in
_
O
in
in
CM
L.
0
XI O
-»- v»
v- m
re
"o "en
S
re
n.
o
re
o
o
u
in
e
o
u
2
X
en
£
cu
CM
in
^ S «
in e m
10
.a
cu i. -o -o
u e> u c
£
01
01 in
*^
c cu
in
in
ci
u
O O O u
«J <-> U «^
XXX
o o
f> tf> V)
S
cu x x
*> o o o
U Z Z Z
C. l^ H_ ««-
01 O C O
n »n *n
^ c e c
o o o o
z re Xi u
o
ink*
O re
O
*** re
e
o
n
"E
3-20
-------
and Blown: soda-lime category match that of the Container Glass
category. This matching is substantiated in an emission test.58
For the Pressed and Blown: other than soda-lime category, EPA
emission test results and emissions reported in the source assess-
ment vary considerably. The emission test results from the EPA
program range from 3.2 to 3.5 g/kg (6.4 to 7.0 lb/ton)959 while
the source assessment values are an order of magnitude higher.60
The table entry of 5 g/kg (10 Ib/ton) reflects the EPA value.
Lastly, for the Wool Fiberglass category, the particulate emissions
correspond to the value reported for regenerative furnaces in the
screening study.
As is the case with gaseous pollutants, switching combustion
sources from natural gas to fuel oil affects the amounts and
characteristics of particulate species emitted from the raw mater-
ials in a glass melting furnace. To estimate the effect on
particulate emissions from switching to fuel oil, the New Jersey
Bureau of Air Pollution tested different types of industrial direct-
/"p
fired combustion sources other-than glass furnaces. The emission
factor for incremental particulate emissions generated by the com-
bustion of number 2 fuel oil, as determined by this study, corre-
sponds to 10-percent of the particulate emissions arising from the
manufacture of soda-lime glass in uncontrolled, natural gas-fired
furnaces, as shown in Table 3-4 (furnace systems without add-on
control equipment are termed uncontrolled). The incremental par-
ticulate emissions from numbers 4, 5, and 6 fuel oil correspond to
15-percent of the soda-lime particualte emissions listed in Table
3-4. Recent emission tests on uncontrolled, soda-lime glass melt-
ing furnaces substantiate the roughly 10-percent increase in
particulate levels from fuel oil-fired furnaces over the emissions
CO
from natural gas-fired furnaces. For borosilicate glasses,
although only one of these emission tests compares fuel oil and
natural gas particulate emission levels, the emissions from fuel
3-21
-------
oil-fired furnaces are 10-percent more than the natural gas emis-
sions.64 In addition, these recent tests show that the character-
istic size distribution of the particles evolved in glass melting
furnaces remains independent of the type of fuel combusted.
The chemical nature of some pollutants does change if sulfur
is present in the fuel. In the manufacture of lead glass when
natural gas is burned, lead oxide is formed; when fuel oil contain-
ing sulfur is burned, this particulate is converted stoichiometric-
ally to lead sulfate.65
3.2.2.3 Emissions From Forming and Finishing Operations
The final steps of glassmaking are termed "forming and finish-
ing" and, depending on the final product, may include combinations
of the following operations: forming by pressing, by blowing, or
by both; surface treatment by metal chlorides to improve glass
strength; annealing to remove internal stresses; decorating,
dipping, or spraying a phenolic resin onto glass fibers. The
nature of the pollutants released to the atmosphere from these steps
. . 66 «67.68
invariably exist as gaseous emissions.
Emissions have been tabulated in documents covering every
glass industry segment, but reflect fewer emission tests than do
emission factors for glass melting. Emissions have been determined
for hydrocarbons, oxides of nitrogen, oxides of sulfur, metal °*1jies,
metal chlorides, hydrogen fluoride, ammonia, and particualtes. ' '
Despite the imprecision of these values, the emission factors are
less by two orders of magnitude than those of glass melting furnace
emissions. As is the case with emissions from raw material handling,
these forming and finishing emissions will not be discussed further
in this document.
3-22
-------
3.2.2.4 State Emission Compliance Regulations
Table 3-5 lists participate compliance limits for various
glass production rates as allowed by states in which most of the
glass manufacturing facilities are located. The table entries are
calculated for existing Container Glass furnaces assuming that 85
percent of process weight rate is transformed into glass production,
which corresponds to the normal 15 to 20 percent cullet usage.
The limits of Illinois,69 Indiana,70 New York,71 Ohio,72
Oklahoma, and Texas are formulated on a mass basis with an
exponential dependence on process weight. California limits are
represented by the allowable particulate emissions of the South
Coast Air Quality Management District. The New Jersey regulation
for glass plants is also determined on a mass basis but with a
linear dependence on process weight and with allowances1 made 'fo^*
increased cullet utilization. The New Jersey Department of
Environmental Protection makes an exception to the process weight
limits for the case of furnaces producing lead glass where a com-
pliance schedule of 0.02 Gr/SCF applies. West Virginia compliance
regulations are interpolated from a table based on process weight.
Pennsylvania maintains a concentration basis regulation for Flat
and Container Glass and mass basis regulation for Pressed, Blown,
78
or Spun Glass.
3-23
-------
o
cz.
Ci
t/}
o
I
co
CD
*-H
I
CO
t1
X
LU
OC.
O
LI-
CO
LU
g
O
«I
s
CC
O-
UJ
B
CO
in
i
CO
O)
ID
«^-
"i1
;»
4->
S
P
&
s?
)
l_
'a
1
1
(D 3
5
£
51
eg""
*1
« I! «^-
c o»-
IB JZr-
f n >
^i
o
V)
.*
.
.c
X>
i.
j:
Ol
^c
S-
JD
^^
L.
x:
s-
^.
J=
J3
L.
.£
Oi
^:
^^
JE
!e
t~
j:
"o>
^
.-.
X)
*^*
i
en
.X
4-> 1.
g =
-^- «n
*l
L.
O)X
,
re
» =
-1
«
CM
__
*~
CM
«
«n
-
in
in
CM
0
«
GO
f-
»O
1*.
»~
«o
^r
,_
CM
CM
^
^~
W
CM
O>
w
CM
CM
CO
«0
^
CO
te
CO
CM
in
in
in
CM
«
in
t^
«
CO
in
r*.
«r
«n
in
CM
^.
CM
S
«
^
<
ro
O>
CO
^
w
u>
0
CO
»~.
to
__
CO
CM
»
__
in
CO
o>
in
«
^.
co
,_
CO
in
r»
rr>
CO
CO
CO
CO
3
er>
«
^
CO
CO
PS-
CO
CO
CO
et
3-24
-------
3.3 REFERENCES FOR CHAPTER 3.0
1. Schorr, J.R., D.T. Hooie, M.C. Brockway, P.R. Sticksel, and
D.E. Niesz, Source Assessment: Pressed and Blown Glass Manu-
facturing Plants, prepared for Environmental Protection Agency,
Industrial and Environmental Research Laboratory, Research
Triangle Park, North Carolina, Contract No. 68-02-1323,
NTIS 600/2-77-005, Task 37, January 1977, Appendix A.
2. Reznik, R.B., Source Assessment: Flat Glass, Manufacturing
Plants, prepared for Environmental Protection Agency, Indus-
trial and Environmental Research Laboratory, Research Triangle
Park, North Carolina, Contract No. 68-02-1874, NTIS 600/2-76-0325,
March 1976, Appendix B.
3. Schorr, J.R., D.T. Hooie, P.R. Sticksel, and C. Brockway,
Source Assessment: Glass Container Manufacturing Plants,
prepared for Environmental Protection Agency, Industrial and
Environmental Research Laboratory Research Triangle Park;
North Carolina, Contract No. 68-02-1323, NTIS 600/2-76-269,
Task 37, October 1976, Appendix A.
4. The Glass Industry Directory Issue 1976-1977, The Glass Industry
Volume 57, No. 10, 1976-1977. ^~~
5. Final Report of Screening Study to Determine Meed for Standards
of Performance for New Sources in the Fiberglass Manufacturing
Industry, prepared for Environmental Protection Agency, Indus-
trial Studies Branch, Research Triangle Park, North Carolina,
Contract No. 68-02-1332, Task 23, December 1976, Tables 4 and 5,
6. U. S. Department of Energy, "Voluntary Industrial Energy Conserva-
tion," Progress Report No. 5, July 1977, page 81,
7. U. S. Bureau of Census, "Current Industrial Reports, Flat Glass
Fourth Quarter 1976, "MQ-32A(76)-4, February 1977.
8. .Glass. Packaging Institute, "Glass Packaging Institute 1977
Annual Report."
9. U. S. Bureau of Census, "Current Industrial Reports, Glass
Containers Summary for 1976," M 326(76)-13, May 1977.
10. U. S. Bureau of Census, "Current Industrial Reports, Consumer,
Scientific, Technical, and Industrial Glassware," 1976, MA-
32E(76)-1.
11. Reference 5, Table 1.
3-25
-------
12. U. S. Bureau of Census, "Current Industrial Reports, Fibrous
Glass 1976," MA-32J(76)-1» June 1977.
13. Reference 1, page 4.
14. Hopper, T.6., and W.A. Marrone, Impact of New Source Per-
formance Standards on 1985 National Emissions From Stationary
Sources, prepared for Environmental Protection Agency, Emis- ,
sion Standards and Engineering Division, Research Triangle
Park, North Carolina, Contract No. 68-02-1382, NTIS 450/3-76-017,
Task 3, October 1975, Volume I, Tables 6-1, 6-3, 6-6, and 6-16.
15. A Screening Study to Develop Background'Information to Deter-
mine the Significance of Glass Manufacturing, prepared for
the Environmental Protection Agency, Research Triangle Park,
North Carolina, Contract No. 68-02-0607, Task 3, December 1972,
Table 1-4.
16. Reference 15.
17. Reference 5, Table 1.
18. Hutchings, J.R., and R.V. Harrington, "Glass" In: Kirk and
Othmer: Encyclopedia of Chemical Technology, Second Edition,
John Wiley and Sons, New York, New York, 1966, page 549.
19. Reference 6, pages 81, 85, and 87.
20. Hanks, G.F., "A Trial on 100% Coal Firing" The Glass Industry,
April 1977, page 10.
21. Reference 18, page 559.
22. Reference T, page 35.
23. Reference 2, page 39.
24. Reference 3, page 36.
25. Point Source Listing for Glass, SCC 3-05-014, National Emis-
sion Data System, Environmental Protection Agency, Research
Triangle Park, North Carolina, June 22, 1977, (Confidential).
26. Reference 3, page 40.
27. Reference 2, page 41.
28. Reference 1, page 39.
3-26
-------
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
Reference 1, page 42.
Reference 2, page 43.
Reference 3, page 40.
Reed, R.J., "Combustion Pollution in the Glass Industry,"
The Glass Industry 54 (4) 24-26, 36, 1973.
Reference 2, page 45.
Reference 3, page 44.
Reference 1, page 44.
Reference 2, page 46.
Reference 3, page 45.
Reference 1, page 45.
Van Thoor, T.J.W., Editor, Materials and Technology, Barnes
and Noble, 1971, page 358.
Shrieve, R.N., Chemical Process Industries, McGraw-Hill, New
York, New York, 1967, page 196.
Reference 2, page 42.
Santy, M.J., "Particulate Control Through Process Modification
Chemical System Screening," prepared for Glass Container
Manufacturers' Institute, New York, New York, December 1971,
Phase II, page 2.
Stockham, J.D., "The Composition of Glass Furnace Emissions,"
Journal of the Air Pollution Control Association. 21:713-175,
November 1971.
Reference 1, page 45.
Preliminary Data from Emission Tests on Uncontrolled Glass
Furnaces, prepared for Environmental Protection Agency,
Industrial and Environmental Research Laboratory, Cincinnati,
Ohio, April 1978.
Reference 5, Tables 9 and 11.
Reference 1, page 44.
3-27
-------
48. Reference 45.
49. Daniel son, J.A., Editor, Air Pollution Engineering Manual,
Public Health Service Publication No. 999-AP-40, Cincinnati,
Ohio, 1967, page 730.
50. Ryder, R.J. and J.J. McMackin, "Some Factors Affecting Stack
Emissions From a Glass Container Furnace," The Glass Industry,
50, 307-311, June 1969.
51. 'Reference 1, Table 7.
52. Reference 2, Table 8.
53. Reference 3, Table 8.
54. Reference 5a Table 9.
55. Reference 5, Table 11.
56. Reference 45.
57. Reference 45. .
58. Reference 45.
59. Reference 45.
60. Reference 44.
61. Reference 5, Tables 9 and 11.
62. O1Sullivan, W., Krauss, C., and Londres, E., "Fuel Oil Particu-
late Emissions From Direct Fired Combustion Sources," New Jersey
Bureau of Air Pollution Control, January 26, 1976.
63. Reference 45.
64. Reference 45.
65. Communication between Dave Powell of PES and John Cherill,
Engineer, Corning Glass Company, Corning, New York, August 1977.
66. Reference 1, Table 8.
67. Reference 2, Table 9.
68. Reference 3, page 80.
3-28
-------
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
Environmental Reporter, State Air Laws, State Index, Illinois,
-TQ7Q - .
ly/O* ........ ,._ , . _ _
Environmental Reporter, State Air Laws, State Index, Indiana,
1978.
Environmental Reporter, State Air Laws, State Index, New York,
1978.
Communication between Dave Powell of PES and Charles M. Taylor,
Chief, Air Programs Development and Review, Office of Air
Pollution Control, Ohio Environmental Protection Agency, October
20,1977. -..:-.-.....
Environmental Reporter, State Air Laws, State Index, Oklahoma,
1978.
Texas Air Control Board, Regulation I, Control of Air Pollu-
tion from Visible Emissions and Particulate Matter, October, 1977.
Rule 405, South Coast Air Quality Management District, Rules
and Regulations, May 1976.
Bardin, D.J., New Jersey Department of Environmental Protec-^
tion. Report of Public Hearing - Control and Prohibition of ..
."Particles from ManufacturIng Processes, September 29, and 30»
1976.
West Virginia Administrative Regulation, 16-20, Series VII,
Sections 1.16, 1.17, 3.01, and 3.02, 1978.
Communication between Dave Powell of PES and Morris Maiin of
the Pennsylvania Bureau of Air Quality and Noise Control,
Department of Environmental Resources, June 29, 1977.
3-29
-------
-------
4.0 EMISSION CONTROL TECHNIQUES
4.1 INTRODUCTION
As identified in Chapter 3.0 of this document, emissions of
nitric oxides, particulatess and sulfur oxides comprise the largest
weight of pollutants released to the atmosphere by the uncontrolled
manufacture of glass products. Examination of the emissions of
these key pollutants for the three major operations of glass manu-
facturing, namely, raw material handling, glass melting, and
forming and finishing, shows that essentially 100 percent of the
oxides of nitrogen, 98 percent of the particulate, and essentially
all of the oxides of sulfur are generated in the melting of glass.
Because emissions are centered in the glass melting operations,
the emission control techniques described in this chapter deal with
the reduction of airborne emissions in the furnace exhaust. In
addition to the previously mentioned major pollutants, which are
emitted from all fossil-fuel fired glass melting furnaces, other
pollutants emitted only from the production of special glass formu-
lations pose potential health problems. These pollutants are:
fluorine, lead, and arsenic.
As broadly applied in the glass industry, manufacturing methods
termed "process modifications" lower glass melting furnace emissions
either by altering raw material recipes or by modifying furnace
equipment. In contrast to this definition, add-on control equip-
ment refers, to devices which treat only the glass melting furnace
gaseous exhaust. In the next section of this chapter process modi-
fications are discussed; all-electric melters are described in
Section 4.3; in the following three sections add-on control tech-
niques are described; in the last sections the control techniques
are summarized and the reduction of arsenic, lead, fluorine, and
sulfur oxide emissions are discussed. For each glass furnace test,
the values of pertinent manufacturing rates and control system
4-1
-------
parameters are listed in this chapter. No additional discussions
of the tests are made elsewhere in this document.
In general, two stack sampling methods have been used to
measure particulate levels in the stack gases from glass melting
furnaces. Both methods ensure that the sample withdrawn from the
stack accurately represents the stack exhaust. Both methods use
the same sampling equipment a stack probe, a filter, and a set
of impingers maintained at a temperature of 0°C (32°F). The basic
difference between the two methods is the configuration of the
sampling equipment. In one method, called the EPA Method 5, the
filter is maintained at about 120°C (250°F) and is placed upstream
of the impingers. In the other method, developed by the Los
p
Angeles Air Pollution Control District, the impingers are placed
upstream of the filter. The calculation of particulate emissions
in the EPA Method 5 involves determining the dry weight of particu-
lates captured in the probe and in the filter. In the Los Angeles
method, the increase in weight of the impingers is measured by
evaporating the impinger solutions, and this dry weight is included
with the dry weights of particles captured in the probe and filter
to determine the particulate emissions.
The EPA Method 5 has become the standard method for analyzing
particulate emissions and is used as the basis of emissions in
this document. Although no study has compared results of these
methods on the same furnace exhaust, knowing the chemical composi-
tion of glass particulate emissions, comparisons can be projected.
It is expected that the Los Angeles sampling configuration should
not affect the particulate catch to any extreme. Additionally,
the Los Angeles testing method should calculate slightly higher
particulate emission levels than the EPA Method 5.
4-2
-------
4.2 PROCESS MODIFICATIONS
4.2.1 BATCH FORMULATION ALTERATIONS
Process modifications employed in the manufacturing of glass
in order to lower emissions include: reducing the amounts of
materials in the feed which vaporize at furnace temperatures;
increasing the fraction of recycled glass in the furnace feed;
installing sensing and controlling equipment on the furnace; modi-
fying the burner design and firing pattern; and, utilizing electric
boosting. The applicability of electric boosting for lower glass
'melting furnace emissions is discussed in the following subsection.
Some process modifications offer the double benefits of lowering
pollutant emission rates as well as lowering fossil fuel consump-
tion rates.
Because emission tests are not available to document the
lowering of particulate emissions by using process modifications,
the evidence substantiating the efficacy of these methods is not
as quantitative as is that for the other control strategies
discussed later in this chapter. Nevertheless, these control
methods and the approach to particulate emission control warrant
consideration.
One of the principles used by glass manufacturers to lower
emissions is straightforward. It involves the alteration of raw
material recipes to lower or eliminate volatile constituents in
the feed to the furnaces. Significant among compounds which have
been removed from the feed in container glass manufacture is
4
arsenic. Feed rates of soda, fluorides, and selenium have been
minimized. Since glass formulations fall in the area of propri-
etary information, no emission tests were obtained that show the
decrease of emissions concomitant with the decreases of volatile
compounds. The amounts of volatile raw batch materials may be
decreased until one of two general types of lower constraints is
4-3
-------
reached. One lower limit is prescribed by the glassmaking process
itself. An example of this type is soda, whose batch levels may
be reduced until the glass product quality falls below production
criteria. The other limitation on some batch constituents may be
glass product specifications. Two examples of this sort are: the
governmental regulations requiring minimum levels of lead and
arsenic in television tubes;5 and, the military specifications for
textile fiberglass.6
Another alteration of raw material recipes which affects
pollutant emissions involves increasing the levels of recycled
glass in the raw batch mix. Since this recycled glass does not
require heat to react, the furnace may be maintained at a lower
temperature than that needed for a smaller cullet mix fraction.
The lower temperature reduces the amounts of pollutants generated
in the combustion of fossil fuel and the compounds vaporized from
the glass bed. Once again, no emission test data are available
to substantiate these results quantitatively. Normal cullet frac-
tions in container glass range from 15 to 20 percent. For some
specialty glasses, the mass fraction of cullet in the feed may
increase to about 70 percent. Glass manufacturers claim that
cullet may be used only up to the level at which impurities in the
cullet deleteriously affect glass product quality.
4.2.2 ELECTRIC BOOSTING
Electric boosting is the term applied to the technique of
dissipating electrical current through molten glass. Electrical
energy is converted to heat because of the high electrical resist-
ance of the molten glass. For a fixed furnace throughput, utilizing
electric boosting decreases the required bridgewall temperature,
decreasing the fuel consumption rate, and thereby decreasing both
particulate and gaseous pollutant levels. Boosting has normally
been used to increase production rate since it does not require
4-4
-------
substantial modifications of the glass furnace. Boosting is com-
monly employed in container glass plants and is less commonly
found in other types of glass plants.
In general, documentation of the lowering of emissions by
electric boosting has not been available in the format of EPA
Method 5 emission testing. For one natural gas-fired container
glass furnace using electric boosting the particulate emissions
per kilogram of glass produced dropped 55 percent from the uncon-
trolled level (the boosting electrodes supplied 18 percent of the
total energy consumed in the furnace, despite a 12 percent increase
in glass production rate). Emissions of S09 did not decrease when
O L-
boosting was used. Although information on the percentage of the
total energy supplied by electric boosting was not available, it
was determined that, on a rough basis, electric boosting provided
less energy for this furnace than for the first furnace.
4-2-3 SUMMARY OF PARTICULATE EMISSIONS WITH PROCESS MODIFICATIONS
To assess the levels of particulate emissions from glass melt-
ing furnaces using process modifications including electric boosting,
the particulate emissions prorated on the basis of glass production
were determined for the furnace discussed in Section 4.2.2 and
for furnaces identified by the Glass Packaging Institute as practic-
ing most types of furnace and process control techniques.10 For all
these electrically boosted furnaces the particulate emission values
range from 0.34 to 0.88 g/kg (0.68 to 1.76 Ib/ton).11'12 Although
some of these emission tests do not match rigorous EPA Method 5
-procedures,, they are adequate enough to indicate rough levels of
emissions.
Because of the narrow extent of these data and because of the
lack of ample supporting emission tests, these values only exemp-
lify the range to which particulate emission levels can be reduced
4-5
-------
by process modifications, including electric boosting. As such,
they overlap with the emission levels indicated in Tables 3-5 and
3-6 and sh'ow that, in general, levels of particulate emissions
from glass melting furnaces using process modifications are indis-
tinguishable from the-uncontrolled cases discussed in Chapter 3.0.
4.3 ALL-ELECTRIC MELTERS
In contrast to conventional fuel-fired furnaces, the surface
of the melter in a cold top electric furnace is maintained at ambient
temperature, and fresh raw batch materials are fed continuously
over the entire surface. As molten glass is withdrawn from the
melter, raw batch drops in the melter gradually heating and finally
reacting in the liquid phase. This processing minimizes losses
from vaporization. The gases discharged through the batch crust
consist of carbon dioxide and water vapor.
Design objectives for all-electric melters have not been based
primarily on emission control, but rather on efficient melting and
product control. Construction is less expensive than that for
fossil fuel furnaces since there are no regenerator chambers, port
necks, checkers, flues, or reversing valves, and in most cases,
stacks can be eliminated. Additionally, there is no need for duct-
work, combustion blowers, fans, extra piping, burners, or special
refractory shapes.
Accomplishment of design objectives resulted in a low surface
temperature and a finer control on the glass melt formulation and
therefore, small levels of emissions. The exact level of emission
control capability is not soundly documented since some electric
melting units employ no exhaust stacks and are vented openly inside
the plant building. However, from the nature of the melting pro-
cess, potential emissions can be deduced and possible relative
amounts of emissions can be estimated. Since there is no combus-
tion taking place, fuel-derived pollutants are eliminated. The
4-6
-------
only air emissions emitted are from the decomposition of carbonates,
sulfates, and nitrates, with the majority of the exhausts being C02.
Finer control of the glass melting process has meant lower emissions
since electric melters retain more borates, phosphates, and fluorides
than fossil fuel burning furnaces.13 In addition, there is no solid
disposal problem as with fabric filters or with electrostatic pre-
cipitators and no water disposal problem as with scrubber systems.
, The development of all-electric melters has occurred relatively
recently. All-electric melting technology has several key limita-
tions which, at present, hinder the. application of this technique
throughout the glass industry. Not all glasses possess the elec-
trical properties required for successful all-electric melter
operation; other glass, formulations attack the electrodes presently
used in all-electric melters.14 Additionally, the all-electric ,
technology may not be far enough advanced to satisfactorily produce
glass in large capacities.
Actual emission test results from all-electric furnaces are
presented in Table 4-1. Little operational information was avail-
able on the melters except that they were maintained at normal
operating conditions during the emission tests. The borosilicate
glass melters were tested in accordance with the EPA Method 5 pro-
cedure; the soda-lime melter, although not using EPA Method 5,
used an EPA approved sampling procedure with results including
both condensed and filtered particulate. The particulate emissions
from both glass formulations were equal in magnitude and ranged
from 0.05 to 0.12 g/kg (0.1 to 0.24 Ib/ton) based on the type of
glass produced. These tests only partly indicate the potentially
achievable emission control since only some all-electric melters
have installed the exhaust stacks required for emission sampling.
In visits to glass manufacturing facilities, all-electric melters
were observed to discharge into the plant building. At these
installations no visible emissions were detected and neither
4-7
-------
CO
CO
UJ
CO
CO
CO
cc:
i- C
4-> OJ
re u
£_ S-
4J QJ-0
CO. C
QJ Q)
o CM en
o *~~ x
o o o
+J
0) in
4-J-o in
ro QJ Q)
t 4-> O
3 U X
O QJ UJ
i- S-
t^ t-
S- 0
re c_>
o.
c
o
in
in
r"
j£=
in
in
E,
QJ
«b
in
in
re
CD
^i '
y |
in -4-> i
in (/>
re s i
i "O -1
CD C I
i C
QJ
U
C
4J OJ
> i-
QJ QJ
J- l4»
QJ
CC
' O -.--
uT
C^
to
f^
^^^
S-
CD
^MV'
JD "
*5f
I
O
X
CO
£
C
^^
en
Cbb
O
^-^
*i«^
J3
^~
<**^*
JJ)
^
Ol
>K,
*
too
D
En
V
_5
TJ
re
S-
QJ
E
3
z:
_
o"
^
«
o
^1^^
in
o
O
QJ
^>
re
-JT-
1 in
re o
o s-
o o
CO CO
in
in
re
"en
Q)
C3 «^3
O *^
SU.
VD
f^
. , ,
a-
^-»
*
^^
|^^
o
*
0)
^J
re
i '_
i in
ro o
T3 i-
0 0
CO CO
in
in
re
^~
en
i QJ
O J=>
o -
a: u-
t^
!"
"
CO
r^
=-
^^^
CD
O
'
QJ
4^
re
w u
_1 -r-
i in
re o
o i-
o o
CO CO
in
in
re
^
en
i QJ
0 JO
0 -i-
3 lo-
co
r~
^-^
Cft
PO
o
^^^-
l£>
Cft
*.
-
Oil
*
^**"
CVJ
1
_ 1
1
re
o
o
co
J-
OJ
c
in re
in *J
re c
r O
CD C-5
CTi
^"
re
OJ
o
OJ
a.
OJ
o ,
CM
S-
o
CD
c
P"
o
QJ
3
in
re
i.
*t-
o
o
Q)
re
OJ
4->
re
in
QJ
o
QJ
I
QJ
C£
re
, ^->
Q) I- X
4-) QJ O II
Q.
in
in c
in QJ
o QJ CT>
4J X X
ro QJ o
i.
8
re
re
c c
QJ QJ
o o
J_ S-
QJ QJ
Q. O.
CM CO
4J
re
re
in
ro .u. u- u.
S_ O CJ> O
en oo co oo
o o o o
£ tr tr tT
i^ CD CD CD
O O T5
4-8
-------
fluoride nor sulfur odors were detected.15 Based on these assump-
tions, the results from the emission tests represent relatively
high levels of particulate emitted from all-electric melters.
In summary, all-electric melting has demonstrated that parti-
culate emission levels equivalent to or less than 0.1 g/kg (0.2
Ib/ton) can be maintained in the production of soda-lime and boro-
silicate glasses. Comparison of all-electric melters with other
control techniques is made in Section 4.8.
4.4 CONVENTIONAL FABRIC FILTER SYSTEMS
Several glass manufacturing facilities utilize fabric filter
systems to collect particulates in the glass melting furnace
exhaust. In these systems, the furnace exhaust is first cooled
and then passed through a fabric filter which retains particulate,
and allows the gases to vent to the atmosphere. The physical
characteristics of the filtering fabrics and the agglomerating
tendency of submicron particles have made the fabric filter systems
viable control techniques for the collection of glass melting
furnace particulates.
Figure 4-1 illustrates a typical baghouse system. In opera-
tion, a fan pulls the furnace gases through devices which cool the
gases to a temperature compatible with the filter material. Cool-
ing is accomplished by duct cooling, dilution air addition, or
water injection. The gases are then forced through the filter
bags. Periodic cleaning of the bags is necessary to maintain high
collection efficiencies. Filter bags are cleaned through shaking
or reverse air pulsations. Conveyors transfer the collected dusts
to hoppers for disposal.
Fabric filter systems are claimed to have the advantages of:
high collection efficiency (99 percent); low pressure drop across
pi
the system; and, low energy requirements. Collection efficien-
cies are not affected by the electrical resistivity of the particles,
4-9
-------
CLEAN AIR
OUTLET
DIRTT XIK
INLET
CELL PLATE
Figure 4-1.
A Simple Two Cell Inside Out Baghouse Equipped
for Shake Cleaning23
4-10
-------
In addition, bag life is about two years depending on the bag con-
struction material.22 There are certain disadvantages to the
application of fabric filters to glass melting furnace gases: the
temperature of gases entering the fabric filter must be below a
maximum value to inhibit attack on the filtering media as well as
above a minimum value to prevent condensation of sulfur trioxides;
and, too high a moisture content of the gases can form an irre-
movable plug within a filter bag.
Table 4-2 lists emission test results for furnaces using bag-
house systems. The following summarizes testing parameters and
irregularities encountered for each test.
Test 24 results are from a natural gas-fired soda-lime glass
melting regenerative furnace. The Los Angeles approved particulate
sampling configuration emission test was used. The fabric filter
system consists of six modules entailing a total bag surface area
of 1,204 m2 (12,960 ft2). The design a/c ratio is 1:1 but during
testing the a/c ratio was about 0.65:1. The pressure drop across
the system is normally 1,250 to 1,500 Pascals (Pa), which is equi-
valent to 5 to 6 inches of water.
Central to the interpretation of this test data is the design
basis of this fabric filter system. The unit.was designed only to
meet local opacity regulations. Since the unit met the regulations
after startup, no improvement of particulate collection was attempted.
In operation the system incurred mechanical failures in the
first year of operation but slight modifications to the fabric
filter internals eliminated the difficulties. Also, the original
onstream cleaning method used reverse air blown between the bags
to collapse the inner bag, cleaning the bag without taking a
section offstream. This original method was modified to the pre-
sent arrangement of a reverse air cleaning cycle where a baghouse
section is taken offstream, with the double bag construction being
retained.
4-11
-------
CO
to
u.
I
CO
UJtO
o: a:
LU
co .
*-iee
tn co
MO.
O. UJ
CM
I
0>
!o
re
1/1
o
«/1
«n
I
QJ
>
ID tn
k tn
"c u
Q? X
o«
tJCJ
«J*~
*>,0
*s ^D
W fl)
7 u
t- 0)
£ k
5
«n
«n
in
"5
£
in
tn
1-
f a
3 >
i-ne
^
*«!*
o>
o"*>
O> Cl
u
U k
t- O)
UI
ci
V)
4A O
.*
>> X
> *J p
> i/» e?>
> s e
^3 Q) ^3 1O ^^ O) O*
gl §5 fc fc |
i XI XI XI
^i -o k »-. j-
ai « a> o u. «*- -
in-o m-e
in o «n ** » i
£tn U) O o o o
k O O O
o. c. at * *
4^ in o '** c&
5 &1 CU BNJ CM
ai
J?
o
4J
C
o>
- s
£ x,
-------
The results for emission test 25 were measured on a glass
melting.regenerative furnace burning low sulfur number 5 fuel oil
and producing soda-lead borosilicate glass, a specialty glass
classified in the Pressed and Blown category of manufacturing.
Emission tests using EPA Method 5 were made on the furnace exhaust
before ?nd past the fabric filter allowing the calculation of the
particulate removal efficiency. The design value of the air-to-
cloth ratio (a/c) is 0.6:1 with all four modules exposed to furnace
exhaust and is 0.8:1 with three modules exposed to the furnace
gases and one module being cleaned. In addition, no operational
difficulties with this fabric filter system were reported.
Data listed for emission test 26 are the preliminary results
of an EPA Method 5 test recently performed on a natural gas-fired
glass melting furnace producing wool fiberglass. This fabric
filter is considered undersized by the glass manufacturer.
Emission tests 27 and 28 report particulate emissions as
calculated by the front-half and back-half catches for the EPA
Method 5 sampling configuration. Therefore, these results are
higher than the Method 5 particulate determinations. The glass
formulation melted in these furnaces is soda-lime borosilicate
producing an end-product classified in the Wool Fiberglass industry
category. The furnace of test 27 is a regenerative type. The
fabric filter in test 28 controls emissions from a small recupera-
tive-type furnace, a raw material batch house, and an electric-
melt-gas-boosted furnace. Particulate concentrations in the fabric
filter discharge are not corrected to 12 percent oxygen as the
oxygen concentrations during the tests were not available.
Particulate emissions for the tests listed in Table 4-2 range
from 0.12 g/kg (.24 Ib/ton) to 0.55 g/kg (1.1 Ib/ton). The'high
collection efficiency claimed for fabric filters is substantiated
in the' soda-lead borosilicate glass test. As mentioned before, the
particulate collection efficiency of the fabric filter treating the
4-13
-------
soda-lime furnace exhaust may be lower than the efficiency which
is technically feasible because particulate collection was never
maximized in this system. In conclusion, fabric filters have
demonstrated reductions of particulate emissions to levels equiva-
lent to less than 0.2 g/kg (0.4 Ib/ton) for glass^ formulations in
two glass industry categories: Wool Fiberglass and Pressed and
Blown: other than soda-lime. Additionally, based on the assess-
ment of test 24, appropriately sized and optimized fabric filter
systems can be expected to reduce particulate emissions from soda-
lime melting furnaces to levels of 0.1 g/kg (0.2 Ib/ton).
4.5 VENTURI SCRUBBER SYSTEMS
Although scrubber systems have been built to control particu-
late emissions in the glass industry, presently only a few devices
are in use controlling container glass emissions. The most common
system in operation is the venturi scrubber. A typical venturi
scrubber is depicted in Figure 4-2. In a venturi scrubber,
particle-laden gases are accelerated through a restriction in the
ducting where water is injected into the gas stream. The velocity
of the gas stream provides the dual function of atomizing the
scrubbing fluid while at the same time providing a differential
velocity between particles and the resulting liquid droplets. By
utilizing high power fans to accelerate the gas stream, it is
possible to generate high gas velocities at the throat of the
venturi. Since the particulates are mostly water soluble, the
scrubber provides a means of removing these emissions. Addition-
ally, some gases are absorbed as condensables.
The scrubber liquor is acidic due to the absorbed acid gases,
and, before being recycled to the venturi, is pH controlled by
caustic solution injection. A bleed stream and makeup water
addition insure that the scrubber liquor is not saturated. Typi-
cally, a bleed rate of 1.3 x 10"4 m3/s (2 GPM) is discharged for a
4-14
-------
c
-S
4J
o
ex
o
u
s-
O)
JD
JD
3
J-
o
CO
u
o.
tM
I
U.
to
-------
2.1 kg/s (200 TPD) container glass plant. Even for a larger
furnace, the bleed rate would be expected to be less than 3.2 x 10
tn3/s (5 6PM).29
The pressure drop to obtain high velocities in the throat of
a scrubber is directly proportional to the gas velocity squared
and the liquid to gas ratio; therefore, high velocities are poss-
ible only at substantial pressure drops which result in high fan
energy expenditures. Typical pressure drops are approximately
on
7,500 Pa (30 inches of water).
Table 4-3 lists emission test results for furnaces using
scrubber systems. Due to the limited number of such systems used
in the glass industry, limited data were available. The following
summarizes testing parameters and irregularities for the available
data.
Test 31 results are for a dual throat venturi scrubber
installed on a container glass furnace burning 0.5 percent sulfur
fuel oil. The liquid water-to-gas ratio is 3.9 x 10" [m /s]/[m /s]
(0.0029 GPM/SCFM) for this system with an 8,212 Pa (33 inches of
water) pressure drop. There is an estimated 0.0053 kg/s (42 Ib/hr)
of Na2S04 dissolved in the water discharge which is diluted by
plant cooling water and discharged without further treatment.
Although the system was not designed for S02 control, approximately
90 percent of the S02 was removed from the furnace exhaust. This
system has experienced startup problems and after startup, two
major maintenance problems: the replacement of a fan due to lining
failure;, and, the rebuilding of a hydraulic reservoir tank due to
collapse. The testing method is that of EPA Method 5.
Test results 32 and 34 are from a packed-bed preconditioning
chamber, variable throat scrubber installed on a natural gas-fired
glass melting furnace. Test 32 is an emission test using the Los
Angeles testing method and Test 34 uses EPA Method 5. The design
liquid to gas ratio is 2.3 x 10-3[m3/s]/[m3/s] (0.0017 GPM/SCFM)
4-16
-------
oo
0:
LU
CD
CO
o:
o
oo
i
ii
a:
UJ
a.
Qu
o-
UJ
OO
UJ
o
oo
oo
o
lo-
co
ID
00
00
oo
00
UJ
UJ
o
C£
O-
co
I
CD
J3
rc
in
C
o
in
in
ff
E
UJ
QJ
ro
^
o
r-
S_
ro
o.
4J
QJ
f»
-[ ^
3
S-
QJ
JO
3
S_
LJ
IO
i
r
S-
3
J ^
C
QJ
C
1 QJ
J- cr>
O >>
«_> X
O
in
c in
o 10
r- OJ
in o
r- UJ
UJ ^8
CM
QJ i
ro O
i »->
3
O T3
r- QJ
4-^ 4-^
S- 0
ro QJ
a. s-
in
c
O
.p*
I/)
in
r-
UJ
in
to
ro
. i^~*
U-
t ^
oo
o
^x,^
S-
cr.
* -
QJ QJ
O O O
r- E 0 i.
4-S QJ !- OJ
i- o: i*- o.
ro M-
0- UJ
to
in QJ
ro o.
yu - ^.
O 1
>} ^->
J_ ^_
in 4-> O
in in cn
ro 3 QJ
i -O 4->
o c ro
c- 1 O
QJ
c o ro
0 C S-
i- 4-> QJ QJ
to in i- _Q
in QJ QJ E
i- t <4- 3
E QJ Z
i i i rv
0
^ ^. * ^. ^ *» ^"^
CM ȣ> vo ro
^a- i i CM
o o o o
*
0
^_" <^V ^«^ ^-^
OO O *i5
i co «s3- un
«
,1 ^ «JI ^ ^^"^ ' * »
«d" «* OO O
f^ CM CM "53-
*
o
^.^ ^Htf' ^_^ *»-^
r^ CM «* o
CO I i CM
o
tn VD
CM CTl
oo r«-
E E § E
r> r~ r*" ^~"
1111
ro f& fO ^3
O "O "O "O
o o o o
oo oo oo i/)
s_ s- s_ s-
QJ QJ QJ QJ
C C C C
r- -I -r- -r-
n3 03 to n3
4-^ 4-^ 4-^ 4-^
C C C C
o o o o
0000
JD JH
i CM CO «d"
CO CO CO CO
S-
OJ
-)->
n.
ro
j^
o
QJ
O
-o
QJ
QJ
QJ
in
QJ
to
in
QJ
O
C.
QJ
QJ
QJ
c:
QJ
Cn
X
o
in
in
QJ
u
X
QJ
4-5
C
QJ
QJ
O.
o
QJ
4->
o
QJ
S-
O
o
QJ -t- O
IQ JO
o
4-17
-------
with a 7,500 Pa (30-inch water) pressure drop. The liquid effluent
is released directly to the sewer. Also, a weak alkaline solution,
which is recirculated through the packed tower and venturi scrubber,
is used to scrub S02 and particulates from the gases. The design
calls for 0.0011 m3/s (16.8 GPM) of makeup water, 9.64 x 10"7 m3/s
(22 6PD) of 50 percent caustic and produces a waste liquid stream,
of 8.2 x 10"4 m3/s (1.3 GPM), containing 1 to 2 percent dissolved
solids and 1 to 2 kg/m3 (1,000 to 2,000 ppm) suspended solids.
There has been no major equipment failure to date; no plugging has
been experienced; and, no problems with corrosion have arisen. A
number of minor operating difficulties have been encountered,
almost all relating to the instrumentation system. In addition to
the particulate reduction, there was a 75 percent reduction in
sulfur oxides with a 7 ppm S02 system discharge.
Test 33 results are from a packed-bed preconditioning chamber,
dual throat scrubber using the EPA Method 5 testing procedure for
an oil-fired container glass furnace. The liquid-to-qas ratio is
9.4 x 10"3[m3/s]/[m3/s] (0.07 GPM/SCFM) estimated from design
conditions with an 8,500 Pa (34-inch water) pressure drop. There
is a 6.3 x 10~4m3/s to 9.5 x 10"V/s (10 to 15 GPM) bleed rate
which is discharged directly to the sanitary sewer. This system'
has experienced a problem with the scrubber exhaust fan which
caused the system to be shut down. During the three months of
operation, it was necessary to twice clean the impeller blades and
fan housing in order to eliminate an imbalance. Also, there were
problems with the pH control system, the soda ash solution mixing
apparatus, and other minor items. The system has been operating
continuously for three months. Operational and maintenance prob-
lems are still being analyzed. In addition to particulate
reduction, sulfur oxides were reduced 86.3 percent with a 100 ppm
discharge concentration.
4-18
-------
Table 4-3 lists particualte emission tests for venturi
scrubbers installed on container glass melting furnaces. Test
number 33 reports results for an oil-fired furnace. Although the
pull rate for this test was only 57 percent of the maximum furnace
capacity, this test data was included as it substantiates the
particulate control efficiencies achievable by venturi scrubbers.
As discussed previously, tests 32 and 34 are from the same furnace
but represent different sampling methods. The emissions per kilo-
gram of glass produced for these tests range from 0.12 to 0.20 g/kg
(0.24 to 0.4 Ib/ton). These tests demonstrate that venturi scrub-
bers can lower the particulate emissions from uncontrolled
container glass melting furnaces to a level equivalent to or less
than 0.20 g/kg (0.4 Ib/ton). A comparison of scrubber systems
with other control techniques is discussed in Section 4.8.
4.6 ELECTROSTATIC PRECIPITATORS
Presently, more than 19 electrostatic precipitators are
installed on glass furnace exhaust systems throughout the country,
(more than any other control technique).
The fundamental steps of. electrostatic precipitation are
particle charging, collection, and removal and disposal of the
collected material. Particulate charging is accomplished by
generating charge carriers which are driven to the particulates
by an electric field. Collection occurs as the charged particu-
lates migrate to electrodes to which the charged particles adhere.
Applying a mechanical force to the collection electrodes dislodges
the collected material which then falls into hoppers. Effective
transfer of dust to the hopper depends on the formation of chunks
or agglomerations of dust which fall with a minimum of reentrainment.
There are two types of electrostatic presipitators used in
the glass industry. Both types are shown in Figure 4-3. One
type consists of a large rectangular chamber divided by a number
4-19
-------
y- Collecting Electrodes
Discharge Electrodes
Charging Electrode Weights
:tive
Negative Charging
Electrodes (-)
Positive Grounded
Collector Plates (+)
Conventional Electrostatic Precipitator
Positiv
Charging
Needles
Positive
Electrode
Plates
Non-Uniform
Electric
Section
Uniform
Electric
Negative Grounded Section
Collecting Plate
;ative Collecting
'.ate Electrodes
_ ..
Ioni±ing
Section
Positive Plate
Electrodes
Collec1:ing
Section
NAFCO Electrostatic Precipitator
Figure 4-3. Conventional and NAFCO Electrostatic Precipitators
35
4-20
-------
of parallel rows of collection plates that form gas flow ducts.
Between these plates are hung a number of small diameter wires
which are connected to a high voltage direct current potential
forming a corona discharge around the wire. This corona generates
electrons which migrate into the incoming gas stream to form gas
ions which attach these charged particles. The charged particles,
in turn, are collected by the grounded collection plates.
The other type of ESP has a multitude of stainless steel
needles fastened to the leading and trailing edge of the discharge
plates. This design configuration requires a low voltage which
allows close spacing between the two collecting surfaces in each
field: the positively charged discharge plates, which have the
attached needles; and, the grounded collector plates. This close
plate spacing permits short collecting sections and relatively
high flow-through velocities.35 Additionally, the regions
between the needles exhibit a uniform electric'field which aids
particle agglomeration. Dust is retained on both the collector
plates and discharge plates.
Electrostatic precipitators can be designed and guaranteed to
collect 99 percent of the particulate in the glass melting furnace
exhaust. Resistivity of the particulate is a determining design
parameter. If the particulate cannot conduct the ionic current
from the corona discharge, it will be entrained and will be released
to the atmosphere. Resistivities are highly dependent upon tempera-
ture with a decrease in resistivity occurring with an increase in
temperature. Some typical resistance figures for various types of
^ *
glass are:00
Borosilicate glass
Lead glass
Soda-lime glass
12
10 ohm - cm
10 ohm - cm
107 to 1010 ohm - cm
(Depending on temperature and
moisture content)37
4-21
-------
Table 4-4 lists emission test results for electrostatic
precipitator-controlled glass melting furnace exhaust. In some
plant configurations one or more electrostatic precipitators
collect particulates from several furnaces. In these cases the
table entries list the total pull rates from all furnaces whose
exhausts are controlled during testing and the sum of the particu-
late emissions of all electrostatic precipitators in the plant.
The following are summaries of the testing parameters and irregu-
larities experienced.
Test 38 results are from a test employing the Los Angeles
testing procedure on a furnace producing soda-lime glass. No data
were available for ESP operational parameters. The unit has been
running successfully since startup.
Test 39 results are from a test employing the Los Angeles
testing procedure on a soda-lime melting furnace. The design
specific collection area of the unit is 138 m2/[Nm3/s] (0.65 ft2/
SCFM), and during testing the unit was operating at about 83 per-
cent of design SCGM. Natural gas was fired during testing. There
have been generally satisfactory results with the operation of
this unit.
Test 40 results, also on a soda-lime furnace, are from a test
employing the EPA Method 5 procedure. The design specific collec-
tion area of the unit is 237 m2/[Nm3/s] (1.12 ft2/SCFM), and during
testing the unit was operating at about 116 percent of design
conditions.
Tests 41 through 44 report particulate emissions from borosili-
cate glass formulations melted in furnaces classified in the Pressed
and Blown: other than soda-lime category.
Test 41 results are from a test employing the EPA Method 5
procedure. The design specific collection area of the unit is
225 m2/[Nm3/sl (1.06 ft2/SCFM), and during testing the unit was
4-22
-------
Table 4-4. PARTICULATE EMISSION TEST RESULTS FOR GLASS MELTING
FURNACES EQUIPPED WITH ELECTROSTATIC PRECIPITATORS
(Mission
Test
Reference
Hunter*
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
S3
54
55
Glass Industry
Category
Container
Container
Container
Pressed and Blown:
Other than Soda-line
Pressed and Blown:
Other than Soda-lln
Pressed and Blown:
Other than Soda-live
Pressed and Blown:
Other than Soda-llw
Pressed and Blown:
Other than Soda- lint
Pressed and Blown:
Other than Soda-llM
Pressed and Blown:
Other than Soda-llMj
Pressed and Blown:
Other than Soda-lime
Pressed and Blown:
Other than Soda-lln
Pressed and Blown:
Other than Soda- 1 In
Pressed and Blown:
Other than Soda-Hm
Pressed and Blown:
Other than Soda-line
Wool Fiberglass
Wool fiberglass
Wool Mberglasi
Glass Typ*
Soda-1 In
Soda-tin
Soda-1 In
Boroslllcate
Boroslllcate
BoraslKcitt
BaroslHcat*
Fluoride/
Opal
lead
Lead
lead
Lead
Lead
lead
Potash-
Soda-lead
Boroslllcate
Boroslllcat*
Boroslllcate
Specific
Collection Area
*/tllm3/i] (FtJ/SCFH)
b
133
237
225
138
290
179
379
233
337
183
195
237
220
222
216
b
(0.65)
(1.12)
0.06)
(0.65)
(1.37)
( .85)
(1.79)
(1.09)
(1.59)
( .86)
< .92)
(1.12)
(1.04)
(1.05)
(1.02)
Percent
of
Design
SCFN
During
Ttst
83
116
100
89
43
84
75
117
91
80
122
Paniculate
Removal
Efficiency
91
98
97
PrectplUtor Outlet Particular Enlssloni
Mass Emissions
«/tg
0.06
.07
.06
.57
.48
.10
.17
.06
.08
.08
.07
.18
.27
.03
.36
.09
.09
(Ib/ton)
(0.12)
( .14)
( .12)
(1.14)
( .96)
( .20
( .34)
( .12)
( .16)
( .1«)
( .14)
( .36)
( -54)
( .06)
( .72)
( .19)
( .17)
Corrected to
12f Excess Oxygen
kg/H*3 x 10'
0.14
.36
.25
1.37
.18
.32
.10
.12
.15
.06
.42
.30
. .'4
.18
.12
(Gr/OSCF)
(0.006)
( .015)
( .010)
( .056)
( .007)
( .002)°
( .012)
( .004)
( .OOS)C
( .006)
( .002)
( .004)c
( -OH)
( .012)
( .026)
( -007)
( .005)
a. References are listed at the end of the chapter
b. Claimed proprietary
c. Not corrected to 12S excess oxygen
4-23
-------
operating at about design conditions. Natural gas was fired during
the testing period. There have been no major problems encountered
with this unit.
Test 42 results are from a test employing the EPA Method 5
procedure. The design specific collection area of the unit is
138 m2/[Nm3/s] (0.65 ft2/SCFM), and during testing the system was
operating at about 89 percent of design SCFM. Natural gas was
fired during testing. There are no available comments regarding
operating problems.
Test 43 results are calculated using EPA Method 5. This
electrostatic precipitator is sized for two glass melting furnaces,
but only one furnace was operating during the test. The glass pull
rate is calculated as 85 percent of the process rate. The manufac-
turer has encountered dust build-up on the blades of the fan used
with this electrostatic precipitator.
Particulate emissions from test 44 are evaluated from the EPA
Method 5 technique. Number 5 fuel oil was fired for this test.
There are no other available comments regarding the operation of
this precipitator or regarding difficulties encountered in its use.
Results listed for test 45 report particulate emissions for
an electrostatic precipitator installed on a glass melting furnace
producing fluoride-opal glass. Pull rate is assumed to be 85 per-
cent of process weight rate. Natural gas was combusted during this
test.
Tests 46 through 51 report particulate emissions from electro-
static precipitators installed on Pressed .and Blown: other than
soda-lime furnaces melting lead glass formulations.
EPA Method 5 was used to determine the particulate emissions
for test 46. The design value of specific collection area is
233 ifjlHn?/s] (1.09 ft2/SCFM); and during the test, the flow rate
through the unit was 75 percent of the design value. The glass
4-24
-------
pull rate is calculated as 85 percent of the process weight rate.
Problems arising in the application of this control device were:
dust build-up on the blades of the exhaust fan, broken insulators,
and arcing.
Test 47 results are from a test employing the EPA Method 5
procedure. The design specific collection area of the unit is
337 m /[Nm3/s] (1.59 ft2/SCFM), and during testing the unit was
operating at about 117 percent of design flow rate. Natural gas
was fired during testing. There have been no major operational
problems encountered.
In test 48, the particulate emissions are reported from an
EPA Method 5 test. Natural gas was used during this test. Again,
pull rate is assumed to equal 85 percent of process weight rate.
Test 49 lists particulate emissions for a natural gas-fired
furnace using EPA Method 5. Pull rate is calculated as being 85
percent of the process weight rate. No additional comments are
available about the unit.
Test 50 results are from a test employing the EPA Method 5
procedure. The design specific collection area of the unit is
195.07 m2/[Nm3/s] (0.92 ft2/SCFM), and during testing the unit
was operating at about 80 percent design SCFM. Natural gas was
used in the furnace during testing and there was no available
information as to operating problems encountered with the device.
No other information is available on test 51 other than: the
testing procedure followed, (EPA Method 5); the furnace fired
natural gas; and, the data listed in Table 4-4.
Test 52 results are,from a natural gas-fired furnace produc-
ing potash-soda-lead glass with emissions determined by a sampling
train similar to the EPA Method 5 train with two exceptions: A
Whatman filter was used and the filter temperature was not main-
tained at 250°C. The design specific collection area is 237 m2/
4-25
-------
[Nm3/s] (1.12 ft2/SCFM), and during testing the unit was operating
at about 120 percent design flow rate. No data were available as
to collection efficiency; the capture dust was analyzed as follows:
79.4 percent PbO, 1.66 percent As^, 5.33 percent As205. There
have been no major operational problems with this unit.
Tests 53 through 55 report particulate emissions from wool
fiberglass plants equipped with electrostatic precipitators. No
other information or comments from the glass manufacturer are
available other than those listed in Table 4-4.
The particulate emissions for soda-lime formulations produced
in the Container Glass category and for the lead, fluoride-opal,
and potash-soda-lead formulations produced in the Pressed and Blown:
other than soda-lime category which are listed in Table 4-4 range
from 0.3 g/kg to .27 g/kg (0.6 Ib/ton to .54 Ib/ton). These results
include tests on both precipitator configurations illustrated in
Figure 4-3. For borosilicate glass formulations manufactured in
the Pressed and Blown: other than soda-lime category and in the
Wool Fiberglass category, the particulate emission test results
range from .09 g/kg to .57 g/kg (.17 to 1.14 Ib/ton). Two factors
could explain the higher emissions for borosilicate emissions
despite the larger special collection area: the higher electrical
resistivity of borosilicate dusts; arid, the tendency for the
collected dusts to bridge in the precipitator. Since the resis-
tivity of the lead dusts is nearly equal to the resistivity of
borosilicate dusts and since the lead particulate is collectable,
the second factor may control the collection of borosilicate glass
melting furnace emissions.
In conclusion, electrostatic precipitators have demonstrated
particulate emission control levels of 0.06 g/kg (.12 Ib/ton) for
soda-lime, lead, and potash-soda-lead glass formulations, and.
levels of about 0.2 g/kg (0.4 Ib/ton) for borosilicate glass formu-
lations. A comparison of electrostatic pre'ci pi tators with other
control techniques is discussed in Section 4.8.
4-26
-------
4.7 ADDITIONAL AND DEVELOPING CONTROL TECHNIQUES
4.7.1 FABRIC FILTER WITH ADDITIVE INJECTION
This control technique utilizes the continuous injection of
chromatographic solids to agglomerate submicron particulate and
to absorb gaseous pollutants. These chromatographic solids are
separated from the gas stream by a conventional fabric filter.
The solids can either by recycled or disposed of in landfill. This
dry system consists of the following equipment: a gas quench-
humidification system, a metering additive injector, and a fabric
filter. The typical pressure drop across the system is about 2,000
Pa (9 inches of water). The additive injection and fabric filter
system has been tested on emissions from a furnace producing float
glass (the most common type of flat glass), on fiberglass furnace
emissions, and on container glass melting furnace emissions.
Although emission testing methods are not indicated for the
float glass or fiberglass tests, particulate removal efficiencies
are reported to be over 95 percent.57 In emission tests of a
container glass melting furnace using this system the particulate
removal efficiency averaged 85 percent with a zero opacity visible
CO
outlet emission. For all types of glass, the grain loadings are
less than 0.12 x 10"4 kg/NM3 (0.005 Gr/DSCF).
4.7.2 MIST ELIMINATORS
Mist eliminators, developed primarily for removing liquid
mist emissions in the sulfuric acid industry, have been pilot
tested on a slipstream of a natural gas-fired container glass
melting furnace. The mist eliminators utilize impaction, intercep-
tion, and Brownian movement to collect on irrigated fibers. Gases
containing mist and spray particles pass through a'fiber bed. 'The
particles are then collected on the fibers in the bed and coalesce
into liquid films. These films fall from the fiber bed by gravity
and the liquid drains out through drain legs. The mist eliminator
4-27
-------
element consists of a cylindrical fiber bed with gas flow passing
through the annular bed and out the center of the element. Gases
emerge from the bed and rise to the system exit.
The results of particulate sampling with an Andersen particle
fractionating sampler show a 96.4 percent collection efficiency of
particulate smaller than 3 microns across the high efficiency ele-
ment, but, due to condensation of nonsulfate compounds and reentrain-
ment from the prefilter, the total system collection efficiency was
found to be 93.6 percent.59 The measured concentration of S02 and
S03 vapor did not decrease through the system. Total pressure drop
through the system was about 2,600 Pa (10.5 inches of water).
Because of the sampling method used and because of the prelim-
inary state of the pilot application of the mist eliminator to
glass melting furnace exhaust, no firm conclusion about the parti-
culate removal efficiency can be made in this document.
4.8 SUMMARY OF PARTICULATE CONTROL TECHNIQUES
Table 4-5 assesses the levels of particulate emissions emitted
from the control systems discussed in this chapter for each indus-
trial glass category except flat glass manufacturing. The emission
levels listed in Table 4-5 represent particulate control technically
achievable as substantiated by test reports, and therefore, reflect
the lower values from the previous tables.
All-electric melting of glass has been shown to be effective
in greatly reducing the particulate emissions from glass melting
furnaces without the addition of add-on control equipment. This
technique is not applicable to the entire glass industry as, at
present, only formulations of appropriate resistivity and furnaces
of relatively moderate production rates can utilize all-electric
melting.
4-28
-------
o
UJ
CO
oo
u_
C/)
z:
o
»~H
oo
in
IH
UJ
c:
CL.
UJ
oo
LU
BL
O.
tn
i
O)
J3
to
U t-
rr O
ID <0
VI «i-
U O>
HI i.
t O.
UJ
i.
01
XI
Xl
3
U
CO
s.
3
C
o>
i.
co
^
£
U
1C
« 1
U
«»
J-
4J Cn
o c
r 4J
LU i
JT
OJ
ex
H-
V)
10
S3
>>
3 >>
0 L.
C O
~- cn
OJ
'VI ID
10 U
^s
C
s
XI
^>
en
cn
*c"
o
**J
"^^
XI
^^
cr
cn
c
o
*^
XI
en
.*:
cn
C
o
XI
cn
:J5
cn
^
-T_-
O
o"
»J.
o
CVJ
^^
CM
*~*
CO
"-;
g
^
*Y
(0
1
L.
O)
c
m
c
0
^y
CNJ
*"
CJ
1
""
*Y
o
o
fcO
1
CD E
iT
10 IO
o
o o
O> CO
CO
o>
0.
in
~_ -
§
1C
OJ
_J
___i
O ro
i X>
CD O
in
C c
c
"^ 'io
gcS"
LT
-.1
13
CO O
CO
o
c c
10 to
x:
c *J
CD
VI t-
VI QJ
0.0
0
~
o
^.
CO
»
O)
re
vt t.
VI Ol
Ol X
0.0
o"
CNJ
- -
o
o
cn
-~
in
CM
^*
**^
o
OJ
to
U
*
"Ln
£
0
CO
VI
TO
EV
(D
,^.
o
O
4-29
-------
Fabric filters have been installed on existing furnaces classi-
fied in both the Pressed and Blown categories and in the Wool
Fiberglass category. As mentioned previously, in the fabric filter
system installed on the soda-lime formulation, particulate collec-
tion was never maximized., implying that the emissions could be
lowered for this melted glass type.
Venturi scrubbers have been installed on existing container
glass furnaces. Scrubbers have not been used to control borosili-
cate emissions because the chemicals discharged in the liquid
effluent present more of a disposal problem than those from soda-
lime glasses.
Electrostatic precipitators have been installed widely in the
glass manufacturing industry. Significant amounts of emission tests
substantiate the values listed in the table.
Switching fuels from natural gas to fuel oil adds particulate
formed in combustion to the particulate formed in producing glass.
The add-on control devices discussed in this chapter would be
expected to be equally efficient in controlling particulate emis-
sions with either fuel. As demonstrated in Tables 4-3 and 4-4,
venturi scrubbers and electrostatic precipitators have previously
been used on fuel oil-fired glass melting furnaces.
Although, as of June 1978, no add-on control system continu-
ously controls particulate emissions from a flat glass manufacturing
furnace, there is no technical evidence to preclude their use. The
flat glass furnaces produce more soda-lime glass than container
furnaces, but the physical and chemical natures of the resulting
particulate are identical. Because of the greater glass production
in flat glass furnaces and concomitant larger exhaust volume than in
container glass furnaces, an electrostatic precipitator would probably
best control the particulate emissions. For this reason, and -because
one flat glass manufacturer is presently installing an electrostatic
4-30
-------
precipitator, these devices are listed as the regulatory options
for the flat glass industrial category in Chapters 6.0 and 7.0.
4-9 CONTROL OF SULFUR OXIDES. FLUORIDE, ARSENIC, AND LEAD
EMISSIONS FROM GLASS MANUFACTURING :
Because sulfur oxides are present in gaseous form in glass
melting furnace exhaust, their control requires a different approach
than that of the control of particulate. One control technique,
the wet scrubber, had demonstrated on commercial scale glass plants
good control of both sulfur oxides and particulates, simultaneously.
As documented in this chapter, 75 and 85 percent reductions of
sulfur oxides were measured for two variable throat, venturi
scrubber systems. The concentrations of sulfur oxide emissions,
calculated as S02, from these-facilities were 7 ppm and 100 ppm
respectively.
Although one test on an electrostatic precipitator showed
some sulfur oxide removal, in general, the other add-on control
techniques discussed in this chapter do not reduce the levels of
sulfur oxides unless other equipment is installed. The one test
showed a 40 percent reduction of S03 and 15 percent reduction of
S02 across an electrostatic precipitator.61 This result has not
been documented in other tests. Treating the exhaust stream with
an alkaline spray has been claimed to convert the gaseous sulfur
oxides to solids which can then be collected by a fabric filter or
an electrostatic precipitator.
In addition, if sulfur oxides are not treated in the glass
melting furnace exhaust when certain fuel oils are burned, they
may lower the collection efficiencies of electrostatic precipitators.
If the fuel oil contains vanadium, the reaction of sulfur trioxide
to sulfuric acid will be catalyzed. This sulfur acid is not only
corrosive to the metal internals of the precipitator but also makes
the agglomerated particulate stick to the collector plates, lowering
CO
collection efficiencies.
4-31
-------
Fluorine used in several glass formulations classified in
pressed and blown glass manufacturing may be emitted in both par-
ti cul ate and gaseous forms in the melting furnace exhaust. Tests
on the uncontrolled glass melting furnace emissions show, on the
average, that one half of the fluorine is present in the particu-
late catch and the other half is present in the impinger and
Co
therefore, exists as a gas in the exhaust.
Not much analysis has been reported, but that which was avail-
able shows electric boosting reduces fluoride emissions by about 75
percent in particulate but increased the fluoride in gaseous form by
43 percent.64 When the exhaust from an opal glass manufacturing
furnace was treated with a lime slurry, 85 percent of the fluoride
emissions were captured in an electrostatic precipitator.
Little test data on arsenic emissions are available. One test
shows that about 80 percent of the arsenic captured in an emission
test was in particulate form.66 Electrostatic precipitators have
been shown to be 99.4 percent effective, and 42 percent effective
in the capture of this particulate form of arsenic.
Electrostatic precipitators have been shown, in two tests,
to collect 70 percent, and more than 90 percent of the lead parti-
68
culate in glass melting furnace exhaust.
4-32
-------
4.10 REFERENCES FOR CHAPTER 4.0
1,
2.
3.
4.
5.
6.
7.
8.
10.
11.
12.
13.
14.
15.
40 CFR Part 60 - Appendix A, December 1971.
Los Angeles County Air Pollution Control District, Air Pollu-
tion Source Testing Manual, Chapter 4, 1972.
Memo from D. Bivens, EMB, to W.O. Herring, ISB, October 28, 1977.
Schorr, J.R., et al, Source Assessment: Glass Container Manu-
facturing Plants, EPA Contract 68-02-1323, Task 37, October
1976, page 71.
Communication between Dave Powell of PES and J. Cherill,
Environmental Engineer, Corning, Glass, September 28, 1977.
Communication between Dave Powell of PES and L. Froberg, Director
of Support Science and Technology Laboratory, Owens/Corning
Fiberglass, September 29, 1977.
Reference 1, page 23.
Burre, D.C., Position Paper -Proposed Regulation Change for
Pressed, Blown, Spun Soda-Lime Glass Melting Furnaces, Carr-
Lowrey Glass Company, October 22, 1975.
Response to 114 Questionnaire by Brockway Glass Company,
Brockway, Pennsylvania, October 12, 1977.
Glass Packaging Institute, Issue Paper on Air Control Technol-
ogy in the Glass Container Industry, September 1, 1977, page 3.
Response to 114 Questionnaire by Glass Container Corporation,
Fullerton, California, October 14, 1977.
Stack Test conducted by Ball Corporation, Decemner 14, 1976.
Lose!, R.E., "Practical Data for Electric Melting," The Glass
-Industry, February; 1976. page 26. . .,;/ -.;--,,...-.--. -..,,**.,..'
Arrandale; R.S., "Pollution Control in Fuel Fired Tanks," Part
5, The Glass Industry, December 1974, page 17.
W.O. Herring, Trip Report to Corning Glass Works, Draft dated
October 7, 1977. . r
4-33
-------
16. Response to 114 Questionnaire by Owens/Corning Fiberglas,
Toledo,.Ohio, November 28, 1977.
17. Reference 16.
18. Reference 16.
19. Report Number C-2230 SCAQMD, Los Angeles, California, December
12, 1974, January 30, 1975, February 11, 1975.
20. Schorr, J.R., et al., Source Assessment: Pressed and Blown
:Glass Manufacturing Plants, EPA Contract 68-03-1326, NTIS
600/2-77-005, Task 37, page 103.
21. Teller, A.J., "Control of Furnace Emissions," The Glass
Industry, February 1976, page 22.
22. Response to 114 Questionnaire by Corning Glass, Corning, New
York, October 12, 1977.
23 Draft of SSEIS: Proposed Standards of Performance for
Electric Utility Steam Generating Units. Vol. I, Particulate
Matter, Figure 4-3, NTIS 450/2-78-006a, July 1978.
24. Report Number C-2027-B, SCAQMD, Los Angeles, California,
November 29, 1973.
25. Response to 114 Questionnaire from Owens-Illinois, Toledo,
" Ohio, October 24, 1977.
26. Preliminary data on emmission test done by the Environmental
Protection Agency at Owens/Corning Fiberglas.
27. Reference 16.
28. Reference 16. , - .
29 FMC booklet on Glass Furnace Emission Systems, FMC Corporation,
Environmental Equipment Division, Itasca, Illinois, July 1977.
30. Reference 11.
31. Reference 11.
32 Report Number C-2186, SCAQMD, Los Angeles, California,
January 24, 1975, January 29, 1975, January 31, 1975.
33. Reference 9.
4-34
-------
34. Preliminary data on emission test done by the Environmental
Protection Agency at Glass Containers Corporation.
35. Custer, W., "Electrostatic Cleaning of Emissions From Lead,
Borosilicate, and Soda-Lime Furnaces," in Collected Papers
of the 35th Annual Conference on Glass Problems. University
of Ohio, 1974.
36. Reference 4, pri.ge 79.
37. Reference 25.
38. Report Number C-2205, SCAQMD, Los Angeles, California, May 2,
1975.
39. Report Number C-2232, SCAQMD, Los Angeles, California, January
7, 1975, January 28, 1975, February 27, 1975.
40. Reference 25.
41. Response to 114 Questionnaire from General Electric, Richmond
Heights, Ohio, October 17, 1977.
42. Reference 25.
43. Reference 22.
44. Reference 41.
45. Reference 22.
46. Reference 22.
47. Reference 41.
48. Reference 22.
49. Reference 22.
50. Reference 25.
51. Data provided by GTE Sylvania.
52. Reference 25.
53. Data provided by J.N. Siegfried and J.J. McCarthy, Johns-
Man vi lie Company, March 7, 1978.
54. Reference 53.
4-35
-------
55. Reference 53.
56. Reference 35.
57. Reference 21.
58. Report Number C-2227 SCAQMD, Los Angeles, California, December
18, 1974, Janaury 3, 1975, January 9, 1975.
59. Monsanto Enviro-Chem, Brink Fact Guide (for mist eliminators),
St. Louis, Missouri (no date).
60. Communication at meeting of Owens/Corning Fiberglass repre-
sentatives and Dave Powell of PES, September 29, 1977.
61. Reference 25.
62. Communication between Dave Powell of PES and B. Gallagher,
Precipitaire, September 23, 1977.
63. Reference 22.
64. Reference 22.
65. Reference 22.
66. Reference 22.
67. Reference 22.
68. Reference 22.
4-36
-------
5.0 MODIFICATION AND RECONSTRUCTION
5.1 GENERAL
"Modification" and reconstruction" have special meanings when
applied to new source performance standards and are defined and
interpreted in Title 40 of Code of Federal Regulations, Parts 60.14
and 60.15 respectively. In general terms, a "modification" is any
physical or operational change to an existing facility which
increases emissions of certain pollutants to the atmosphere. If a
change does increase emissions, then the resulting incremental
emissions.must be controlled to a level such that the total emis-
sions from the facility to the atmosphere do not exceed those
levels which existed before the change. In contrast to a modifi-
cation, a "reconstruction" is a change which is so substantial so as
to reclassify the facility as a new source rather than as an altered
existing source. For this case the source may become subject to
the limits of the new source performance standard.
5.2 MODIFICATION OF GLASS-PRODUCING PLANTS
Paragraph 40 CFR 60.14(a-) reads as follows:
"Except as provided under paragraphs (d), (e), and (f)
of this section, any physical or operational change to an existing
facility which results in an increase in emission rate to the
atmosphere of any pollutant to which a standard applies shall be
considered a modification within the meaning of Section 111 of the
Act. Upon modification, an existing facility shall become an
affected facility for each pollutant to which a standard applies
and for which there is an increase in the emission rate to the
atmosphere."
By definition an "existing facility" is a piece of equipment
or -a component which was constructed prior to the date of proposal
of an applicable standard of performance. An "affected facility"
5-1
-------
is a piece of equipment or a component constructed or modified
after the date of proposal of an applicable standard of performance.
An"existing facility which undergoes a modification as defined in
the Act and 40 CFR 60.14 becomes an affected facility.
As stated in the regulation, an increase in the emission rate
(determined on a mass rate basis such as kg/hr) of any pollutant
for which a standard applies may constitute a modification and
thereby require the control of the incremental increase of emis-
sions. Paragraph 40 CFR 60.14 (e) lists exceptions to this general
rule of emission increase. The most relevant exceptions for glass
manufacturing exempt the control of incremental increases in
pollutants caused by the normal repair and maintenance of a glass
furnace, by an increase in production if the furnace were origin-
ally capable of such an increase, and by a designed change of fuel.
To identify changes of glass melting furnace systems which
should be assessed to determine if system.modifications occur,
the Glass Packaging Institute prepared a paper describing furnace
2
maintenance and alterations. In this paper, both cold and hot,
scheduled and'emergency repairs were listed. Table 5-1 shows the
repairs and maintenance named in this 6PI paper. Of the many
furnace system changes listed, none constitute a modification of
glass melting furnace systems as no emission rates of pollutants
would increase.
5.3 RECONSTRUCTION OF GLASS-PRODUCING PLANTS
Paragraph 60.15 (a) and (b) read as follows:
"(a) An existing facility, upon reconstruction, becomes
an affected facility, irrespective of any change
in emission rate.
(b) 'Reconstruction' means the replacement of compon-
ents of an existing facility to such an extent that:
5-2
-------
Table 5-1. .GPI LIST OF GLASS MELTING FURNACE
MAINTENANCE AND ALTERATIONS2
MAINTENANCE
Cold Repair, patch,
or rebrick
Sidewall
Doghouse
Throat refactory
Melter bottom
Ports
Checkers
Regenerator walls or crown
Refiner sidewall, breastwall or
bottom
Repair portions of superstructures
Clean or replace equipment in fuel, cooling, electrical
boosting, and combustion air systems
Hot
Operational
Non Operational
Fi11 cracks
Replace burner blocks
Add refractory
Repair worn wall sections
Repair or patch checker shafting
Repair electrical booster electrodes
Repair skimmer or mantle blocks
Patch hot checkers
Repair refactory at normal glass
bath level
Reset shadow wall brick
Repair doghouse
ALTERATIONS
Raise furnace
Deepen or expand melter
Modify regenerator
Modify combustion volume
Switch fuel
5-3
-------
(1) The fixed capital co'st of the new components
exceed 50 percent of the fixed capital cost
that would be required to construct a com-
parable entirely new facility, and
(2) It is technologically and economically feasible
to meet the applicable standards set forth in
this part."
The remainder of Section 60.15 specifies the information that
must be submitted to the Administrator and the basis upon which a
determination will be made, should an existing facility propose to
replace components to the extent stated above. The term "fixed
capital cost" is defined as "the capital needed to provide all the
depreciable components and it is intended to include such things
as costs of engineering, purchase and installation of major process
equipment, contractors' fees, instrumentation, auxiliary facilities,
buildings, and structures."
Of the repairs and alterations listed by the Glass Packaging
Institute and shown in Table 5-1, only those involving major rebriek-
ing of a glass melting furnace may be costly enough to reclassify a
furnace as a reconstructed facility. However, based on a communica-
tion from an industry representative, most repairs and alterations
involving rebrieking would not constitute a reconstruction. This
is so despite the fact that in special cases the cost of rebrieking
may exceed the 50 percent of replacement cost criterion of Paragraph
40 CFR 60.15(4).3
5-4
-------
5.4 REFERENCES FOR CHAPTER 5.0
1.- Subpart A, Part 60, Subchapter C, Chapter 1, Title 40, Code of
Federal Regulations, December 16, 1975.
2. Memorandum on Glass Container Furnace Maintenance and Alter-
ations, John Turk, GPI, to Stanley Cuffe, EPA, dated January
12, 1978.
3. Remarks of Mr. Siegfried, representative of Johns-ManviTle, at
NAPCTAC meeting, April 5, 1978.
5-5
-------
-------
6.0 ALTERNATIVE REGULATORY OPTIONS
6.1 BASIS FOR REGULATORY OPTIONS
The purpose of this chapter is to identify regulatory options
for limiting particulate emissions from glass melting furnaces
consistent with each category of glass manufacturing stated in
Chapter 3.0. Applications of control techniques, discussed in
Chapter 4.0, which meet the regulatory options are substantiated
by emission tests results and by transferring appropriate technol-
ogy if actual emission test results are not available.
Based on the technical discussions in Chapter 4.0, particu-
late emissions from glass melting furnaces can be reduced signifi-
cantly by the following emission control systems:
1. All-electric melters
2. Fabric filters
3. Venturi scrubbers
4. Electrostatic precipitators
These four stand out because of their efficiency in controlling
particulate emissions and because of their history of successful
operation on commercial-scale units. Although all-electric melt-
ing effectively reduces particulate emissions, its use is too
limited to serve as the sole basis of a regulatory standard; how-
ever, it may be utilized to meet a standard.
In the following subsections, two -alternate regulatory options
are listed for each glass manufacturing category and additionally
for the two subcategories of pressed and blown glass formulations.
In every case except for the Flat Glass category, Option I is
based on the lowest level of emissions attained by the control
technologies discussed in Chapter 4.0. Option II allows an emis-
sion level less stringent than Option I but, except for the Flat
Glass category, it is supported by a larger data base than Option I,
6-1
-------
The exission limits for flat glass are based on transferring the
control technology used on container glass melting furnaces.
6.2 ALTERNATIVE REGULATORY OPTIONS FOR CONTAINER GLASS MANUFACTURING
6.2.1 OPTION I
Under this option, a numerical emission limit of 0.1 grams of
particulate per kilogram of glass produced (0.2 pounds of particu- .
late per ton of glass pulled) would be selected. Emission test
results shown in Table 4-4 document that electrostatic precipitators
would be able to comply with a standard set on this option. As
shown in Table 4-2, .an emission test on a furnace manufacturing
products classified in the Pressed and Blown Glass category, but
melted in container glass melting furnaces, suggests that an appro-
priately sized fabric filter would control emissions to comply with
a 0.1 g/kg standard. Considering that the nature of particulate
emissions from both categories is similar due to similar glass
formulations and considering that the test method used for the
pressed and blown furnaces biases particulate results higher than
the EPA Method 5, a slightly smaller air-to-cloth ratio than that
designed for this pressed and blown furnace should meet the parti-
culate emission limit of this option. Additionally, uncontrolled
all-electric melting furnaces would be able to comply with this
option.
6.2.2 OPTION II
For this option, a numerical emission limit of 0.2 grams of
particulate per kilogram of glass produced (0.4 pounds per ton)
would be selected. All of the control techniques mentioned in
Section 6.2.1, that is, electrostatic precipitators, fabric filters.
and uncontrolled all-electric melters would meet this option.
Additionally, based on emission test results presented in Table 4-3,
venturi scrubber equipped glass container melting furnaces would
comply with this option.
6-2
-------
6.3 ALTERNATIVE REGULATORY OPTIONS FOR PRESSED AND BLOWN
MANUFACTURING SODA-LIME GLASS FORMULATIONS
6.3.1 OPTION I
Under this option, a numerical emission limit of 0.1 grams
of participate per kilogram of glass produced (corresponding to
0.2 pounds of particulate per ton) would be selected. Because the
production rates of furnaces in this glass manufacturing category
approximate those of container glass melting furnaces, because
these pressed and blown segment furnaces are built in the same
configuration as container glass melting furnaces, and because the
glass formulation melted in these pressed and blown furnaces is
essentially the same as that melted in container glass melting
furnaces, the quantities and chemical composition of the resulting
particulate approximately match those of container glass manufactur-
ing. Therefore, the control techniques which would meet container
glass Option I would be expected to comply with this option. These
control techniques are: electrostatic precipitators, fabric filters
with a slightly smaller air-to-cloth ratio than that listed in
Table 4-2, and uncontrolled all-electric melters.
6.3.2 OPTION II
A numerical emission limit of 0.2 grams of particulate per
kilogram of glass produced (equalling 0.4 pounds of particulate per
ton) would be selected for this option. Paralleling the rationale
of Pressed and Blown: Soda-Lime formulation OptionI, the control
techniques which would be expected to comply with this option are
those listed for Container Glass Option II, namely, electrostatic
precipitators, fabric filters, and uncontrolled all-electric melters.
6-3
-------
6.4 ALTERNATIVE REGULATORY OPTIONS FOR PRESSED AND BLOWN
MNUFACTURING OTHER THAN SODA-LIME FORMULATIONS
6/4.1 OPTION I
A numerical emission limit of 0.25 grams of particulate per
kilogram (0.5 pounds per ton) of glass produced would be selected
for the furnaces melting borosilicate, opal, lead, or other glass
formulations and manufacturing products classified in the Pressed
and Blown industrial category. As shown in Table 4-4, electro-
static precipitators with specific collection areas approximately
three times those used for container glass melting furnaces would
comply with this option for borosilicate, opal, and lead glass
formulations. Additionally, fabric filters would be expected to
comply based on data presented in Table 4-2.
6.4.2 OPTION II
This option consists of a numerical emission limit of 0.5
grams of particulate per kilogram (1.0 pounds per ton) of boro-
silicate, opal, lead, or other glass produced in melting furnaces
classified in the Pressed and Blown category. Emission test
results of electrostatic precipitator equipped furnace stacks show
that this level of control would be achievable with a smaller
specific area of collection than that required to comply with Option
I. Data in Table 4-4 shows most electrostatic precipitator install-
ations met this option. Additionally, fabric filters would be
expected to comply with this option.
6.5 ALTERNATIVE REGULATORY OPTIONS FOR WOOL FIBERGLASS MANUFACTURING
The following emission limitations are written specifically
for particulates emitted from glass melting furnaces and do not
include particulate generated in other processing steps such as,
the application of resin in the forming and finishing operations.
6-4
-------
6.5.1 OPTION I
_ Under this option, a numerical limit of 0.2 grams of particu-
late per kilogram of glass produced (0.4 pounds per ton) would be
selected. Because of test results presented in Table 4-4 that
show that two of three electrostatic precipitators met the emission
limit or this option, electrostatic precipitators would comply with
this option as illustrated by the data shown in Table 4-2. Addition-
ally, uncontrolled all-electric melters would comply with this
option as shown in Table 4-1.
6.5.2 OPTION II
An emission limit of 0.4 grams of particulate per kilogram of
glass produced (0.8 pounds per ton) would be set by this option.
The use of electrostatic precipitators, fabric filters, and uncon-
trolled all-electric melters would comply with this emission limit.
6-6 ALTERNATIVE REGULATORY OPTIONS FOR FLAT GLASS MANUFACTURING
A numerical emission limit of 0.15 grams of particulate per
kilogram of glass produced (0.3 pounds per ton) would be selected
for this option. With one exception, flat glass facilities have
not installed add-on control equipment to control particulate
emissions; and so, in the absence of add-on control emission test
results specifically for flat glass melting furnaces, this numerical
emission limit is based on the similarity of particulate emissions
generated from flat glass manufacture and container glass manufacture
and on the percentage of particulate control attained at container
glass melting furnaces.
The soda-lime formulations of flat glass and container glass
are essentially identical as are the chemical composition and
physical characteristics of the resulting particulate. The differ-
ence between the particulate emissions from flat glass manufacture
6-5
-------
and those from container glass manufacture is the larger quantity
of solids to be collected from flat glass manufacturing. The
satisfactory collection of this larger mass rate of particulate
appears to present no technical difficulties (a fact which is
corroborated by the guarantee underwritten by an electrostatic
precipitator manufacturer for a flat glass facility). Therefore,
an electrostatic precipitator would be expected to comply with
this standard.
The numerical emission limit reflects the roughly 90 percent
removal of particulate required in the Option I emission limits
for the Container Glass, Pressed and Blown, and Wool Fiberglass
industrial categories.
6.6.2 OPTION II
As mentioned at the beginning of this chapter, Option II is
based on a larger emission test data base than the lowest controlled
emissions on which Option I is based. Option II, therefore, allows
higher particulate emissions than Option I. The upshot of comparing
the numerical emission limits is that, for the previously discussed
categories, the Option II emission limit is twice that of Option I.
Applying this relationship to the Flat Glass category, the numerical
emission limit selected for Option II would be twice that of Option
I [0.3 g/kg (0.6 lb/ton)]. Again, electrostatic precipitators
would be expected to comply with this option.
6.7 SUMMARY OF NUMERICAL EMISSION LIMITS
A summary of the numerical emission limits for each industrial
glass manufacturing category for natural gas-fired melting furnaces
is presented below:
6-6
-------
Table 6-1. SUMMARY OF ALTERNATIVE REGULATORY OPTIONS
Glass
Manufact-
uring
Segment
Container
Pressed and
blown:
soda-lime
Pressed and
bl own :
other than
soda-lime
Wool fiber-
glass
Flat
Part icul ate Emissions
Uncontrolled
g/kg
1.25
1.25
5
5
1.5
(Ib/ton)
( 2.5)
( 2.5)
(10 )
(10 )
( 3 )
Option I
gAg
0.1
.1
.25
.2
.15
(Ib/ton)
(0.2)
( -2)
( .5)
.( -4)
( .3)
Option II
gAg
0.2
.2
.5
.4
.3
(Ib/ton)
(0.4)
( .4)
(1-0)
( .8)
( .6)
6.8 NUMERICAL EMISSION LIMITS FOR FUEL OIL-FIRED GLASS MELTING
FURNACES
As reported in Chapter 3.0, the combustion of fuel oil for
glass melting contributes particulate to the furnace exhaust. When
fuel oil is burned, the allowable numerical emission limits for the
regulatory options listed in this chapter will be increased 15
percent. For example, the Container Glass Option I emission limit
of 0.1 g/kg (0.2 Ib/ton) would be adjusted to 0.12 g/kg (0.23 Ib/ton)
if fuel oil were used. Similarly, the Option II value for the same
category would increase from 0.2 g/kg (0.4 Ib/ton) for natural gas-
firing to 0.22 g/kg (0.44 Ib/ton) for fuel oil-firing.
6-7
-------
6.9 MODEL PLANT PARAMETERS
Table 6-2 lists values of operating parameters for model plants
classified in each industrial category and subcategory. These model
plants were generated to typify new glass melting furnaces which will
be constructed in the five year period from 1978 to 1983. As such,
they form one basis for determining the ambient air impact (in
Chapter 7.0) and the control system cost impact (in Chapter 8.0)
associated with the alternative regulatory options.
The values of the parameters were based on data encoded in
the National Emission Data System (NEDS) and revised by representa-
tives of the glass industry. Presented are values for production
rate, stack height, stack diameter, stack gas exit velocity, stack
gas temperature, particulate emissions and emission rates. Exhaust
gas velocities are calculated for the model plants assuming a 10.5
percent oxygen concentration in the furnace exhaust, which corres-
ponds to 100 percent excess oxygen. Stack diameters are then
calculated to maintain a 30 fps stack gas exit velocity. Addition-
ally, the exhaust flow rates shown for fabric filter applications
are larger than the uncontrolled furnace rates because of the
addition of cooling air.
6.10 COMPARISON OF ALTERNATIVE REGULATORY OPTIONS WITH STATE
COMPLIANCE LIMITS FOR EXISTING GLASS FACILITIES
To gauge the magnitude of emission rates proposed by the
alternative regulatory options in contrast to the compliance limits
allowed by the states in which glass facilities are located, emis-
sion rates for model plants in each industry category as set in both
options are tabled with emission rates corresponding to the compli-
ance limits for existing glass facilities located in New Jersey.
The current New Jersey regulations are shown because they represent
a rough average of the states' compliance limits discussed in
Chapter 3.0.
6-8
-------
As seen in Table 6-3 for all cases, both options represent
more stringent emission rates than the New Jersey compliance
regulations. In particular for the Container category, represent-
ing the largest production capacity of all glass industry segments,
Option I represents 19 percent of the state emission value and
Option II represents 39 percent.
6-9
-------
tn
C£
LU
"So,
.Or &. £
u-u. c,f
U
o
Of
o o
§*;
I1*"
to
-ICM
to
COO
<-«CM
to
^HCM
as
CXI CM
»
CM CM
CM in
CM CM
*"*
CM CM
*x. O
ffi
(U
*5
j.
ct
o
1
e
cu
CO
CO CD*
^H tO
cn
in in
CM CO
So
*-
"
s§
CO
CD O
*-* to
Ol
tnm
CM CO
ro
tr
E^
^
OJ
"o
(U
4-*
to
CO CM
rtto
«-4 in
CO CM
CM m
cMr.;
s^
!«».*»-«.
CD CO
CM tO
CM in
^
m~
^
§
u
"tj
M
to
ovro
S^
°*§.
0»1?
^.
c*§.
o
*-< .
en co
"^o"
tn
V) CX
t^
^
0
o
4_i
t.
|
^
~IJ
nJ
M
CO
^^
tn to
si
^^
CD <»>
cn o
in to
en CD
into
^f in
if
^**
into
^F
O 'fc-'
1
+>
t-
CJ
ex
s
tj
tO
1 1
1 1
o
tn f-«
i i
i i
i i
in »
CM tn
c
o
j* !S
en
o
t.
er
o
u
CM W
OCD
CM <
CD CD
I I
«-«CM
'~
1 1
«-*CM*
"^
1 1
C
O
S'S-
C7>-*
C
o
o.
o
*r co
do
*3- CO
00
1 1
_
"^
~3
CM *r
**^
i t
c
c
j?]=
en-
_
c:
o
o'
1 t
1 1
tnro
OCO
"*
1 t
1 t
» 1
CO
t-
J-
tn j;
t
c
c
tn
V
\
o
(U tU
ra t-
o «-*
*- c:
*» o
u u
o c:
n. 13
~2
SM"
* CO
i i
«,.-.
" 0
CM «-H
" C.
s-
£i
i
_^^
Vt§
C7I-
C
U
+J
tx
CJ
«5
XJ VD
" to
1 1
tn CM
"i
CO *
in ex
CO »
in c\
S-
i i
L.
tn x.
cn '
^
C
o
*J
6-10
-------
-o
QJ
o
CJ
CO
C£
UJ
«=c
D;
J
al a
O
O 1
«a 4^ ^ dJ
VI 2 ^*" ,n
£'£.££
«£«i"
uj a.
o
|||
0 C ^
is) O
a
"C £*0 3
J3 S- £
££££
o "-
o
r^ QJ
£u
fij
c s-
O 3
t> tfu
c
0 1
-t- t-^J
*-> to ' o
° **~ £ c
*~ o. ., i-
»->*- t- 3
O 0 § Lu
UJ CU
1- "So,
.a ^ *-*
_Q O ^
o
-o
£ Si^: 3
S- »-> O 2
"ra """ -M *~
u- "- c I?
o "-
*"*
(U
r^ CU
£(j
ro
4J C
C t.
O ZJ
CJ LU
ID
CO
O CD
C
1 1
CO
Cn o
« i
CO
CD CD
en CD
^^
^.
in CD
* in
^r
in CD
*r in
*~^
**
tn o
^r tn
^- >
-^.la.
^*Ji
c
o
£
T3
O
L.
C\.
TO
CO CD
"^^
I 1
cn
Ln in
CM CO
*
CM cn
**-"
CO
CO CD*
en
in tn
CM CO
*~*
^^^
r*1* CD
CM cn
__ M-
JZ
ftj
.t;
-V
nj
LO
»-* r-
to CM
t « tn
i t
in 10
₯~*S.
tn to
^*i
^- co
r-t CO
^^^_
CM VO
CD CO
»-H CO
CM tO
CD CO
-« CO
E^
QJ
a;
o
t)
LSI
*r
*-4
cn o
* *
i t
S
CO
s
cn o
CO
*"~**
S
cncD
CO
^
cn o*
CO
^
^r
*~!^-.
cn CD
CO
, .
tf>
E'^-
4-»
U
O
at
x
OJ
,v
M
LO
in
in
coo"
cn
c.
i i
O
tn o
-§
CM
CO
O
in CD
^^
CM
CO
0
CO
o
cn
"~-
tn
in
f-^ S"
o
o^
C!.
in
in
^8
C3
to
*ET
in »*-
**E-^.
4-1
ni
O
'*-
< )
ni
I/I
CD CO
1 1
o in
CO O
CO O
*"*" *i
CD n
cn CD
in vo
o in
CO CD
T *ef
^^
CD in
CO O
z
4,1
flj
n.
M
*^
"« i
to
g |
1 t
1 1
CM in
-ICM
1 I
1 1
in--*-
CM in
-» CM
I
o
QJ
O
S-
c:
(.->
tj
c:
gp-^
* CM
'O
i i
CO
*""
1 t
^c7
*^-
^c7
*^,
t i
c
o
J?!Q
en'*'
c:
*
o.
o
^^
CM^r
"«
i t
CM^"
""^
1 (
CM"^
'^*
CM^T
*^_^,
i i
c:
3
"cjlC-
c
o
n.
CD
1 1
1 1
CO*^
t-s CD
"^^
1 i
1 1
VO *-«
«3 CM
tn
t.
JtZ
'SC-
at
-
c
o
^
F
QJ
-CJ
QJ
-------
o
s
1
CO
(5
UJ
3
g
=8
01 «
Vt X>
«A O
E«
o. c=
gg
2S
gs
into
C3
lO
o
tn c>
-s
CM
to
0)
i
to
"
So
%%<
UJO.
111
-* O C\J *
s
Si
toco
OJO
rod
o
.tc
u
3
ncontrol
6-12
-------
TPD of textile fiberglass. Each of these plants is planned to increase
production in the FRP sector of the industry where capacity is projected
to grow by about 85% between 1976 and 1982,'(from 325,000 tons to 600,000
tons)'.35
Projected growth in this industry appears to indicate that some
new sources will be required to meet consumer demand. While no new addi-
tional facilities have been announced, discussions within the industry
indicate that both new and expanded facilities will be utilized to facili-
tate increased production.
This study assumes that all additional capacity required by the
industry will be provided by new facilities. Industry representatives
indicate that a typical new facility will produce 100 TPD of textile fiber
glass. It is estimated that the equivalent of 5 new facilities will be
required, in addition to the two plants presently under construction, to
meet projected demand for the industry through 1982.
8.1.6 PRESSED AND. BLOWN GLASS
8.1.6.1 Industry Structure
The primary descriptor of this segment of the industry is that each
plant manufactures glass and glassware that is pressed, blown, or shaped from
glass produced within the plant. Plants in this segment of the industry may
produce consumer and/or commercial glassware. Consumer glassware includes
products such as tumblers, stemware, tableware, cookware, pvenware, kitchen-
ware, and ornamental, decorative, and novelty glassware. Commercial glass-
ware includes products for the lighting and electronics industries and
various other fields, such as the scientific and technical market.
8-31
-------
The pressed and blown glass industry may also be subdivided into two
broad divisions: (1) the machine-pressed and blown sector, which is charac-
terized by relatively large publicly held firms that often produce products
largely for their own use, and (2) the hand pressed and blown sector, which
is comprised mainly of privately owned small firms that produce glassware
products that are generally more expensive, and are valued by the consumer
for their quality and craftsmanship.
It has been estimated that 50 firms in the industry produce approx-
imately 98% of all pressed and blown glass shipments.36 Corning Glass has
production facilities operating in all areas of this segment, and Owens-
Illinois, in most of them. More than half of the plants in the industry
produce hand pressed and blown glass exclusively. Approximately 10% of
the plants identified produce both hand-pressed and machine-pressed and
blown products.
8.1.6.2 CONSUMER GLASSWARE
8.1.6.2.1 Machine-Pressed and Blown
8.1.6.2.1.1 Geographic Location
Seven major firms were identified as producers of machine-pressed and
blown consumerware in 13 plants in 7 states of the United States. Four plants
are located in the Central part of the country, one on the West Coast, and
eight in the Eastern sector. There are fourteen plants that are primarily
in the hand-pressed and blown sector but which have been identified as having
machine-pressed and blown capabilities.
8.1.6.2.1.2 Integration and Concentration
Each of the major firms in this sector of the industry manufactures
8-32
-------
O)
a
O
C
o
o
CO
a:
UJ
UJ
a:
o"~
.O r i- F_
u_ iZ c. £
o
o
o>
r OJ
O O
4-> CI
C 1-
c
*°.^
in o
CO O
*""*
i i
t i
10
in CD
ro CD
* *
o -o
-*. O.
3;- ^^*
(U
Rj
*~
c
o
u
3
1
a.
CO
CO O
i i IO
1 t
1 1
CO
i-* O*
^J
U-
jz:
cn
1
4J
to
OV >.
**s
1 1
22
4-1
M-
S-
GJ
-a
u
to
2
CO
1 1
1 1
r-i
cn CD
to CL
EE - '
4^?
O
a>
X
*-*f
-
O
n
QJ
cj aj
= E
O 4-»
*- C
M O
- O
r-t CO
*-^
i i
i i
i i
c:
0
4-*
cn-*.
J^ JD
CD-*
C'
O
4-*
O.
0
CO UD
*^
1 1
1 1
e
0
I?]o"
cn
c
o
n.
o
i *
i i
t i
CD in
r-4 CO
V* .O
cn* *
_
c
o
<
u
U 01
5 e
O 4-»
- c
*-> o
t- o
ra C
CD m
^eo
i i
i i
. i i
s-
.c
to JD
cn '
c
o
4J
CO.
o
CM in
CM 1^
1 1
1 1
1 '
J_
-*.
cn*-'
o
o.
o
6-13
-------
o
It
5
CD
UJ
o:
LlJ
o
o
UJ
j*
to
or
LU
-3
UJ
CO
o.
o
O
§
CD
*£.
U.
O
O
to
et
P.
CO
ID
O)
££
4-> tn o
E S- «-
QJ O) -P
S- «-3 IO
S- i
3 3 3
Z 0)
in
c
1- « 1
S- c
o
Is £
<+> O
CO
s*
or
0,
>
r- «
IO G
C 0
S- T-
O) -*J
«4-^ C^>
^^ C^^
eC
r- > ^>
in en
3 tt)
T3 4->
S-
^_
.0
^"-
*l ^
in
en
^
^
XI
*
w
^^
en
^^^
t-
xT
^,
v>
CD
^
,_!
CM
Ni *
^*
CM
If)
^
^^
CM
CM
^^*
in
CO
_
o
0
oT
CD
""'
"" - ~~
*\.
'-
CM
*
^_^.
CO
tn
^^**
^_
CM
ID
CM
*
0
ID
CM
i^»^ -^
j_
CM
*
VD
^""^
CO
.
co"
*
^_^
^~
^^^
«4-
J,
VD
O
O
ID
_« ....^
*
CO
o
IO O>
gr
r»l
^^
CO
VD"
T_l
^,
CM
GO
^^,
t
o
0
-
1
c
g
^^
CO
o
IO O>
CM
VD
^_^
CO
CM
*-,
VD
CM ,
^^
0
d.
CO
o
ID
^-«^-«
C t
» 1
O IO
t "O
CO O
o
c c
CO ID
^r
l~-
^^
CO
en
CM
«d-
^
CO
^_^
CM
VD
CM
o
o
-
01
5 i
o
Q)
in s-
in QJ
Q) f
J- !->
o. O
CO
6-14
-------
7.0 ENVIRONMENTAL IMPACT
7.1 AIR QUALITY IMPACT
7.1.1 PRIMARY AIR QUALITY IMPACT
7.1.1.1 Introduction
This section assesses the air quality impact of the regulatory
options enumerated in Chapter 6.0 for new glass producing facili-
ties built in the period from 1978 to 1983. Ambient air particu-
late concentrations and contributions to the national particulate
emission inventory provide the means to quantify the air quality
impact. A computer dispersion model predicts the ambient air par-
ticulate concentration from model plants of each industrial cate-
gory (Table 6-2) for the following cases: an emission rate from an
uncontrolled furnace; an emission rate which matches the State
Implementation Plan (SIP) regulations, as typified by current New
Jersey regulations; and emission rates matching the numerical
limits of the regulatory options. The other element of measuring
air quality impact, the contribution to the national particulate
emission inventory, is determined by proportioning the uncontrolled
furnace emissions to the degree emission reduction realized by the
SIP regulations and by the regulatory options. The air quality
assessment consists of comparing the ambient air particulate con-
centrations and the particulate emission inventory impacts result-
ing from model plant's meeting the regulatory options with those
from uncontrolled furnaces, with those from facilities whose emis-
sions satisfy the SIP regulations, and finally comparing the impact
of each regulatory option.
Table 7-1 shows, for each glass category, projected amounts of
glass produced from new facilities at the end of the 5 year period
from 1978 to 1983 and the amounts of particulate emissions released
by these facilities, if uncontrolled. These figures are derived
7-1
-------
Table 7-1. GLASS INDUSTRY GROWTH IN THE PERIOD FROM ,1978 TO 1983
Uncontrolled
Particulate
Emissions
Annual
Growth
Rate
Industry
Category
New Growth
Tons x 10
Tons x 10
Container
Pressed
and blown:
soda-lime
Pressed
and blown:
other than
soda-lime
Wool
fiberglass
from the production rates, the annual growth rates, and the uncon-
trolled emission rates listed in Chapter 3.0.
The Industrial Source Complex Dispersion Model Short-Term Com-
puter Program (ISCST) was used to calculate the magnitudes and
locations of maximum ground-level particulate concentrations for
each of the regulatory options.1 The ISCST program corresponds
to the EPA single source model (CRSTER) modified to include the
effects of aerodynamic downwash on plume dispersion. No effects of
fugitive emissions were taken into account in this study. Addi-
tional program inputs were meteorological information from the
Greater Pittsburgh Airport and the stack parameters listed in Table
6-2.
In.order to simplify analysis, the current New Jersey regula-
tions were chosen to typify the norm of SIP particulate emission
7-2
-------
regulations. In all glass categories except the Pressed and Blown:
soda-lime category (50 ton per day furnace), the SIP allowable
emissions for model plants are less than the uncontrolled furnace
emissions. For the Pressed and Blown: soda-lime category, the SIP
allowable emissions for the 50 ton per day furnace are essentially
identical to the uncontrolled furnace particulate emissions.
7.1.1.2 Container Glass Category
As displayed in Table 7-2, the maximum ground-level particu-
late concentration from the model plant varies with the control
equipment installed for each regulatory option. Implementation of
regulatory Option I would reduce the 24-hour maximum particulate
concentration emitted by an uncontrolled furnace by 93 percent with
a fabric filter and by 91 percent with an electrostatic precipita-
tor. Option I also reduces 24-hour maximum particulate concentra-
tions allowed by the SIP regulation by 82 percent with a fabric
filter and by 78 percent with an electrostatic precipitator. For
all these estimations, the maximum particulate concentration occurs
300 meters from the stack at the plant boundary. Based on an an-
nual arithmetic average, the 0.6 jug/m3 particulate concentration
resulting from the uncontrolled furnace is reduced to 0.1 Mg/m3,
both by a fabric filter sized for Option I and by an appropriately
designed electrostatic precipitator. All three particulate maxima
occur 300 meters from the stack.
Control equipment which can be selected for Option II are
fabric filters, venturi scrubbers, and electrostatic pracipitators.
On a 24-hour basis, the fabric filter reduces the maximum particu-
late concentration emitted by an uncontrolled furnace by 85 per-
cent; the venturi scrubber reduces the concentration by 45 percent;
and, the electrostatic precipitator reduces the concentration by 82
percent. The fabric filter and electrostatic precipitator sized
7-3
-------
Table 7-2. CONTAINER GLASS CATEGORY
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Case
Uncontrolled furnace
SIP regulation
Fabric filter meeting
Option I - 92% control
Fabric filter meeting
Option II - 84% control
Venturi scrubber meeting
Option II - 84% control
Electrostatic precipitator meet-
ing Option I - 92% control
Electrostatic precipitator meet-
ing Option II - 84% control
24 Hour
Maximum
Parti c-
ulate
Concen-
tration
(ug/m3)
10.8
4.5
0.8
1.6
5.9
1.0
1.9
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Annual*
Maximum
Parti c-
ulate
Concen-
tration
(^g/m3)
0.6
0.3
0.1
0.1
0.6
0.1
0.1
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.3
0.3
* Arithmetic mean
for Option II reduce the maximum particulate concentration allowed
by SIP regulation by 64 percent and 58 percent, respectively. The
particulate concentration 300 meters from a venturi scrubber stack
is predicted to be greater than the particulate concentration from
7-4
-------
a model container glass furnace meeting the SIP participate emis-
sion regulation. On an annual basis, the maximum particulate con-
centrations from stacks equipped with a fabric filter or electro-
static precipitator sized for Option II are 0.1 /^g/m3, represent-
ing an 83 percent reduction from the uncontrolled furnace. Con-
trolling emissions by venturi scrubber does not reduce maximum
particulate concentrations from the uncontrolled case for the
annual average calculation.
Implementing the SIP particulate emission regulation would
reduce the uncontrolled furnace maximum particulate concentrations
by 58 percent on a 24-hour basis and by 50 percent on an annual
average basis.
Although dependent on selection of control equipment, Option
I, in general, represents a 50 percent reduction in 24-hour maximum
particulate concentration over Option II. On the annual average
basis, there is no estimated reduction in the maximum particulate
concentration from Option I as compared to Option II.
As shown in Table 7-1, the projected additional contribution
to the national particulate emission inventory from uncontrolled
container glass facilities brought on-stream in the period from
1978 to 1983 is projected to be 2.4 x 109 grams (2,700 tons).
Meeting SIP emissions reduces this amount to 1.0 x 109 grams
(1,100 tons). Implementing Option II reduces the uncontrolled fur-
nace emissions by 84 percent -- a reduction of 2.0 x 109 grams
(2,300 tons). Option II represents a 62 percent reduction in the
SIP particulate emissions and captures 6.2 x 108 grams (680 tons)
more particulate than container glass furnaces controlled to the
SIP levels. Option I control reduces uncontrolled furnace particu-
late emissions 2.2 x 109 grams (2,500 tons), an amount equaling a
92 percent reduction of the uncontrolled emissions. As compared to
the SIP allowable emissions, Option I captures 8.1 x 108 grams
7-5
-------
(890 tons) more participate, which represents an 81 percent reduc-
tion of of the SIP allowable emissions. Option I control reduces
Option II particulate emissions by 50 percent or 1.9 x 108 grams
(210 tons).
7.1.1.3 Pressed and Blown: Soda-Lime Category
In this section ambient air analyses are made for two charac-
teristic sizes of furnaces a 50 TPD small furnace and a 100 TPD
larger furnace, as shown in Tables 7-3 and 7-4 respectively. How-
ever, the contribution to the national particulate emission inven-
tory is based on the entire category. As mentioned previously, the
particulate emissions allowed by SIP regulations are assumed to be
identical to the uncontrolled furnace emissions from a 50 TPD fur-
nace. For this case, the impacts of the controlled to the SIP
regulations and the uncontrolled emissions are also identical.
For the 50 TPD furnace size (Table 7-3) Option I and II am-
bient particulate concentrations depend on the control technique
used. On a 24-hour basis, the maximum ground-level particulate
concentration for uncontrolled furnaces and emissions controlled to
SIP regulations is 2.7 tig/m*. Implementing Option I would reduce
the maximum particulate concentration by 93 percent with a fabric
filter and by 89 percent with an electrostatic precipitator with
all maxima occurring 300 meters from the stack. Both types of
control equipment reduce the annual average maximum particulate
concentration below 0.1 ng/m3, although in this case, maxima
occur 500 meters from the stack.
Selecting equipment to satisfy Option II for the 50 TPD fur-
nace size would reduce the 24-hour maximum particulate concentra-
tion from uncontrolled furnaces and furnaces controlled to the SIP
regulation by 81 percent with a fabric filter and by 78 percent
7-6
-------
with an electrostatic precipitator. The annual average maximum
particulate concentrations from both devices is less than 0.1
at 500 meters from the stack. .
Table 7-3. PRESSED AND BLOWN: SODA-LIME 50 TPD
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Case
Uncontrolled and SIP
regulation
Fabric filter meeting
Option I - 92% control
Fabric filter meeting
Option II - 84% control
Electrostatic precipitator meet-
ing Option I - 92% control
Electrostatic precipitator meet-
ing Option II - 84% control
24 Hour
Maximum .
Partic-
ulate
Concen-
tration
(Mg/m3)
2.7
0.2
0.5
0.3
0.6
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
Annual*
Maximum
Partic-
ulate
Concen-
tration
(Mg/m3)
0.2
<0.!
<0.1
<0.1
<0.1
Dis-
tance
(km)
0.5
0.5
0.5
0.5
0.5
* Arithmetic mean
On a 24-hour basis, Option I represents, in general, a 55
percent reduction in emissions as compared to Option II emissions.
On the annual average basis, Option I and Option II present identi-
cal impacts.
For the 100 TPD furnace size (Table 7-4), Option I reduces the
24-hour maximum particulate concentration from the uncontrolled
furnace case by 94 percent using a fabric filter and by 88 percent
7-7
-------
using an electrostatic precipitator. Option I reduces the maximum
concentration as compared to the SIP case by 92 percent using a
fabric filter and by 83 percent using an electrostatic precipi-
tator.
Table 7-4. PRESSED AND BLOWN: SODA-LIME TO 100 TPD
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR,
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Ca c P
Uncontrolled
SIP Regufation
Fabric filter meeting
Option I - 92% control
Fabric filter meeting
Option 11-84% control
Electrostatic precipitator meet-
ing Option I - 92% control
Electrostatic precipitator meet-
ing Option II - 84% control
24 H o u r 1 A n n u a 1*
Maximum I
Partic-
ulate
Concen- 1 Dis-
tration
(yg/m3)
3.4
2.4 '
0.2
0.6
0.4
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.7 0.3
Maximum
Partic-
ulate
Concen-
tration
(yg/m3)
0.2 '
0.1
^0.1
*0.1
-cO.l
^0.1
Dis-
tance
(km)
i "
0.3
0.3
0.3
0.3
0.3
0.3
*Arithmetic mean
Installing either control system sized for Option I reduces the
annual averaged particulate concentration from 0.2 pg/m3 for an
uncontrolled model furnace and from 0.1 pg/m3 for a furnace
controlled to the SIP regulations to less than 0.1 Mg/m3.
Implementing Option II with a fabric filter realizes an 82 and
a 75 percent reduction in the uncontrolled furnace and the SIP
7-8
-------
regulation 24-hour maximum particulate concentrations, respective-
ly, as compared to a 79 and 71 percent reduction realized with an
electrostatic precipitator. Annual maxima averages show identical
results as the Option I estimation for this furnace size.
On the average, Option I emission control results in a 54 per-
cent reduction in the 24-hour particulate concentration as compared
to Option II emission control, but on an annual basis, Option I and
Option II yield identical maximum particulate concentrations.
Uncontrolled new furnaces in this category, commissioned in
the 1978 to 1983 period of interest, are projected to release an
additional 3.1 x 108 grams (340 tons) of particulate to the
atmosphere. The following quantities are based on the assumption
that 26 percent of the pressed and blown glass is made in 50 TPD
furnaces.
Particulate control to the SIP regulation by the 100 TPD fur-
naces reduces the uncontrolled particulate emissions to 1.6 x 108
grams (180 tons). Option II reduces the uncontrolled amount by 1.9
x 108 grams (210 tons); Option I reduces the the uncontrolled
amount by 2.1 x 108 grams (230 tons). Option I controls 4.5 x
107 (50 tons) more than the state regulation and Option II con-
trols 2.7 x 107 grams (30 tons) more than the state regulation.
Option I, therefore, captures 1.8 x 107 grams (20 tons) more
particulate emissions than Option II.
For the 50 TPD furnaces in this category: Option I reduces
the uncontrolled particulate emissions by 7.7 x 107 grams (85
tons); Option II reduces the uncontrolled amount by 6.8 x 107
grams (75 tons). Thus, Option I controls 9 x 10^ grams (10 tons)
more particulate than Option II. As previously noted, the SIP
regulation and the uncontrolled furnace emissions are identical.
The contribution to the national particulate emission inven-
tory is based on the entire category. Thus, overall, Option I
7-9
-------
reduces the uncontrolled particulate emissions by 2.9 x 10** grams
(315 tons); Option II reduces the uncontrolled amount by 2.6 x
grams (285 tons). Option I controls 2.7 x 107 grams (30 tons)
more particulate than Option II.
7.1.1.4 Pressed and Blown: Other Than Soda-Lime Category
The analysis for this industrial category parallels the Press-
ed and Blown: soda-lime analysis in that the impacts for two fur-
nace sizes are estimated (a 50 TPD and a 100 TPD pull rate). How-
ever, the analysis of the "other soda-lime" subcategory differs
from the "soda-lime" subcategory in that the emissions allowed
under the SIP provisions for the 50 TPD furnace are not identical
to the uncontrolled furnace particulate emissions. The ambient
particulate concentration estimations are illustrated in Table 7-5
for a 50 TPD furnace and in Table 7-6 for a 100 TPD furnace. The
particulate emissions allowed by the SIP vary with process (or
production) rate, and so, to determine the SIP impact on national
particulate emission inventory, the proportion of each furnace size
must be determined. The impact is estimated assuming 26 weight
percent of the additional glass produced in this subcategory from
1978 to 1983 is processed in the smaller sized furnace. This
assumption is based on National Emission Data System information
for the entire Pressed and Blown (N.E.C.) classification.2
For a 50 TPD furnace (Table 7-5), the uncontrolled emissions
result in a 24-hour maximum particulate concentration of 10.7 //g/m3
The allowable particulate emissions under the SIP regulations
reduce the uncontrolled furnace particulate emissions by 70 per-
cent. Controlling particulate emissions to Option I levels reduces
the 24-hour maximum value for an uncontrolled furnace by 93 percent
and the .SIP emissions by 78 percent. The uncontrolled furnace
7-10
-------
Table-7-5. PRESSED AND BLOWN: OTHER THAN SODA-LIME 50 TPD
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE 'CONCENTRATIONS
Case
Uncontrolled furnace
SIP regulation
Fabric filter meeting
Option I - 95% control
Fabric filter meeting
Option II - 90% control
Electrostatic precipitator meet-
ing Option I - 95% control
Electrostatic precipitator meet-
ing Option II - 90% control
24 Hour
Maximum
Partic-
ulate
.Concen-
tration
teg/m3)
10.7
3.2
0.7
1.3
0.7
1.0
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.3
Annual*
Maximum
Partic-
ulate
Concen-
tration
(^g/m3)
0.8
0.2
<0.,
0.1
<0.1
0.1
Dis-
tance
(km)
0.5
0.5
0.5
0.5
0.3
0.3
* Arithmetic mean
annual average maximum particulate concentration of 0.8
reduced to less than 0.1 ng/rn^ by Option I control.
is
Option II reduces the 24-hour maximum particulate concentra-
tion of an uncontrolled furnace by 88 percent with a fabric filter
and by 91 percent with an electrostatic precipitator. On a 24-hour
basis Option II - a fabric filter reduces the SIP concentrations by
59 percent and an Option II - electrostatic precipitator, by 69
percent. On an annual average basis Option II control devices
realize an 88 percent reduction in particulate concentrations from
uncontrolled furnace emissions and a 50 percent reduction in the
7-11
-------
Table 7-6. PRESSED AND BLOWN: OTHER THAN SODA-LIME TO 100 TPD
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Case
Uncontrolled furnace
SIP regulation
Fabric filter meeting
Option I - 95% control
Fabric filter meeting
Option II - 90% control
Electrostatic precipitator meet-
ing Option I - 95% control
Electrostatic precipitator meet-
ing Option II - 90% control
24 Hour Annual*
Maximum
Partic-
ulate
Concen-
tration
(pg/ni3)
13.9
2.5
0.8
1.5
0.9
1.9
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.3
Maximum
Partic-
ulate
Concen-
tration
to/m3)
0.9
0.2
<0.1
0.1
<0.1
0.1
Dis-
tance
(km)
0.3
0.3
0.3
0.3
0.3
0.3
* Arithmetic mean
particulate concentrations from furnaces controlled to the SIP
regulations.
For the 100 TPD furnace (Table 7-6), the relevant comparisons
for 24-hour particulate concentrations are: Option I - fabric
filter, a 94 percent reduction in the uncontrolled case, a 68 per-
cent reduction from the SIP provisions; Option I - electrostatic
precipitator, a 94 percent reduction in the uncontrolled^case, a 64
percent reduction from the SIP provisions. Comparisons for Option
7-12
-------
II are: Option II - fabric filter, an 89 percent reduction in the
uncontrolled case, a 40 percent reduction from the SIP provisions;
Option II - electrostatic precipitator, an 86 percent reduction in
the uncontrolled case, a 24 percent reduction from the SIP provi-
sions. On an annual basis, Option I control devices reduce concen-
trations to less than 0.1 Mg/m3; Option II devices reduce concen-
trations to 0.1 Mg/m3, representing an 89 percent reduction for
uncontrolled furnaces and a 50 percent reduction for the SIP regu-
lations. All maxima occur 300 meters from the stack.
Implementing Option I reduces Option II 24-hour maximum par-
ticulate concentrations about 50 percent.
As listed in Table 7-1, particulate emissions from new fur-
naces in this subcategory, if uncontrolled, are projected to con-
tribute an additional 3.5 x 108 grams (390 tons) of particulate
to the national particulate emission inventory. The SIP provisions
decrease this amount to 7.4 x 107 grams (80 tons), assuming 26
weight percent of the glass will be manufactured in 50 TPD fur-
naces. Implementing Option II controls capture 3.2 x 108 grams
(350 tons) more particulate than the uncontrolled furnace emis-
sions, for a 90 percent reduction. As compared to the SIP provi-
sions, Option II controls 3.6 x 107 grams (40 tons) more particu-
late for a 50 percent reduction. Implementing Option I reduces
particulate emissions by 3.3 x 108 grams (370 tons) more than the
uncontrolled furnace case, 5.4 x 107 grams (60 tons) more than
the SIP provisions, and 1.8 x 107 grams (20 tons) more than the
Option II controls. Percent reductions for these comparisons are:
95 percent; 75 percent; and 50 percent, respectively.
7-13
-------
7.1.1.5 Wool Fiberglass Category
As seen in Table 7-7, the impact of Option I controls on the
24-hour maximum particulate concentration depends on the control
system selected. A fabric filter designed for this option reduces
the uncontrolled furnace maximum particulate concentration by 94
percent and the SIP allowable emissions by 48 percent. An electro-
static precipitator sized for Option I reduces the uncontrolled
furnace value,by 93 percent and the SIP provision value by 39 per-
cent. On an annual basis, Option I controls reduce uncontrolled
furnace concentrations by 94 percent and the SIP concentrations by
50 percent.
Table 7-7. WOOL FIBERGLASS
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Annual*
Maximum
Partic-
ulate
Maximum
Partic-
ulate
Concen-
Uncontrolled furnace
SIP regulation
Fabric filter meeting
Option I - 98% control
Fabric filter meeting
Option II - 96% control
Electrostatic precipitator meet-
ing Option I - 98% control
Electrostatic precipitator meet-
ing Option 11-96% control
* Arithmetic mean
7-14
-------
Systems meeting Option II emission limits also vary in air
quality impact. A fabric filter reduces 24-hour maximum particu--
late concentration from an uncontrolled furnace by 88 percent,
whereas, an electrostatic precipitator reduces the uncontolled fur-
nace value by 86 percent. For both control devices, the 24-hour
maximum particulate concentrations are higher than the value for
the SIP allowable emissions. On an annual basis, Option II control
devices reduce maximum particulate concentrations from uncontrolled
furnaces by 89 percent with a fabric filter, and by 83 percent with
an electrostatic precipitator. Maximum particulate concentrations
resulting from these devices match or exceed those resulting from
the SIP regulations.
Uncontrolled furnaces in this category are projected to con-
tribute an additional 1.8 x 109 grams (2,000 tons) to the nation-
al particulate emission inventory. The SIP regulations lower
these emissions to 2.1 x 10** grams (230 tons). Implementing
Option II controls captures 1.7 x 109 grams (1,920 tons) more .
particulate than the uncontrolled furnace emission ~ a 96 percent
reduction. Option II represents a 65 percent reduction in the SIP
emissions or captures 1.4 x 10^ grams (150 tons more particu-
late). Option I controls realize a 98 percent reduction in uncon-
trolled furnace emissions (1.76 x 109 grams or 1,960 tons of
particulate captured), an 83 percent reduction in the SIP particu-
late emissions (1.7 x 108 grams or 190 tons of particulate cap-.
tured), and a 50 percent reduction in Option II emissions (3.6 x
107 grams or 40 tons of particulate captured).
7.1.1.6 Flat Glass Category
Only the air quality impact of electrostatic precipitator sys-
tems are computed for the Flat Glass category. Option I reduces
7-15
-------
the uncontrolled furnace 24-hour maximum particulate concentration
by 89 percent, the SIP provision 24-hour value by 54 percent, and
the Option II 24-hour values by 49 percent. The uncontrolled fur-
nace annual average maximum participate concentration of 0.6 M9/m3
is reduced by Option I control to less than 0,1 fig/!"3
Option II reduces the uncontrolled furnace 24-hour maximum
particulate concentration by 78 percent and the SIP provision by 10
percent. This option, on an annual basis, reduces the uncontrolled
furnace maximum particulate concentration by 75 percent. The
annual average value for Option II equals the SIP regulation maxi-
mum particulate concentration.
Uncontrolled furnace emissions from new furnaces are projected
to contribute an additional 3.6 x 108 grams (400 tons) of par-
ticulate to the national emission inventory. Meeting the SIP pro-
visions drops this amount to 8.8 x 107 grams (97 tons). Imple-
menting Option II reduces uncontrolled furnace emissions by 2.9 x
108 grams (320 tons), equaling an 80 percent reduction. Option
II also reduces the SIP provision emissions by 1.6 x 107 grams
(18 tons), for an 18 percent reduction. Option I reduces uncon-
trolled furnace emissions by 3.2 x 108 grams (360 tons), for an
90 percent reduction, reduces the SIP emissions by 5.2 x 107
grams (57 tons), and a 59 percent reduction in the SIP allowed
emissions, and reduces Option II emissions by 3.6 x 107 grams (40
tons), or a 50 percent reduction in Option II emissions.
7.1.1.7 Summary of Air Quality Impact
As shown in Tables 7-2 to 7-8, for all glass categories,
implementing Option I would reduce the 24-hour maximum particulate
concentfation from uncontrolled furnaces by about 90 percent. Con-
trol to Option II levels reduces this uncontrolled furnace maximum
7-16
-------
Table 7-8. FLAT GLASS
MAGNITUDES OF AND DISTANCES TO MAXIMUM 24-HOUR
AND ANNUAL AVERAGE PARTICULATE CONCENTRATIONS
Case
Uncontrolled furnace
SIP regulations
Electrostatic precipitator meet-
ing Option I - 90% control
Electrostatic precipitator meet-
ing Option II - 80% control
24 Hour
Maximum
Partic-
ulate
Concen-
tration
fcg/m3)
20.5
5.0
2.3
4.5
Dis-
tance
(km)
. 0.3
0.3
0.3
0.3
Annual*
Maximum
Partic-
ulate
Concen-
tration
(Mg/m3)
0.6
0.15
«u ,
0.15
Dis-
tance
(km)
0.3
0.3
0.3
0.3
* Arithmetic mean
by about 85 percent. Option I, in general, reduces the SIP allowed
24-hour maximum particulate concentrations by about 80 percent, ex-
cept for the Wool Fiberglass and Flat Glass categories, where the
reduction averaged about 50 percent. Option II controls reduce
concentrations predicted under the SIP regulations by about 60 per-
cent in the Container, Pressed and Blown: soda-lime, and Pressed
7-17
-------
and Blown: other than soda-lime categories. Impacts from imple-
menting Option II controls in the other two categories do not dif-
fer from the SIP provision impacts. In general, Option I repre-
sents a 50 percent reduction in the Option II 24-hour maximum
particulate concentration.
Total particulate emissions from uncontrolled furnaces placed
onstream from 1978 to 1983 in all glass categories is projected to
be 5.2 x 109 grams (5,740 tons). Meeting the state regulations
(SIP) lowers this total to 1.5 x 109 grams (1,670 tons). Option
I reduces uncontrolled emissions by 4.8 x 109 grams (5,370 tons);
and, Option II by 4.6 x 109 grams (5,000 tons). As compared to
the state regulations, Option I reduces additional particulate
emissions by 1.1 x 109 grams (1,200 tons); and, Option II by 0.9
x 109 grams (990 tons). Option I reduces Option II emissions by
3.4 x 108 grams (370 tons).
The ambient particulate concentrations resulting from model
plants (Tables 7-2 to 7-8) are much less than the national primary
and secondary air standards for total suspended particulates. How-
ever,, with the exception of the Pressed and Blown: soda-lime cate-
gory, the maximum partculate concentrations on a 24-hour basis from
all uncontrolled furnaces exceed the Prevention of Significant
Deterioration - Class I increment of 10 M9/m3. The maximum par-
ticulate concentrations resulting from these uncontrolled furnaces
do not exceed the Class II increment of 37 M9/m3.
7.1.2 SECONDARY AIR QUALITY IMPACT
As all of the energy supplied to the control systems is in the
form of electricity, use of these systems transfers the source of
particulate emissions from the glass manufacturing facility to the
electric utility plant. The amounts of this secondary air impact
can be estimated from the electrical requirements of implementing
7-18
-------
the regulatory options and the SIP regulations as presented in Sec-
tion 7.4. The bases for these estimates are the range of energy
impacts listed in Section 7.4 and a utility boiler emission factor
of 0.1 pounds of particulate for one million Btu's of heat input as
reported in the New Source Performance Standard for Fossil Fuel-
Fired Boilers. Embedded in the energy impacts is a utility boil-
er thermal efficiency expressed as 10,000 Btu/kWh representing an
efficiency of 34 percent. Results of these calculations are:
Option I 9,300 kg (10.2 tons) to 25,800 kg (28.4 tons)
Option II 9,300 kg (10.2 tons) to 32,300 kg (35.6 tons)
SIP provisions 8,300 kg ( 9.1 tons) to 29,500 kg (32.4 tons)
In all cases, the particulate emissions from the utility plant
are two orders of magnitude less than the uncontrolled glass par-
ticulate emissions.
7.2 WATER POLLUTION IMPACT
The glass manufacturing process has minimal water pollution
potential. Any water, added to batch materials to prevent dusting,
generated by chemical reaction, or sprayed into the furnace exhaust
for cooling, is vaporized and is swept out of the stack as vapor.
Option I presents no water pollution potential as control sys-
tems meeting this option do not discharge water streams.
The only control system meeting Option II which presents any
potential water pollution potential impact is the venturi scrubber.
Although each venturi system discharges approximately 1.3 x 10~4
m^/s (2 gpm) of waste containing about 95 percent water, the total
water pollution impact is negligible as few glass container facili-
ties are anticipated to utilize this control system.^
In summary, control of particulate emissions is expected to
pose a negligible source of water pollution. State regulations
will not be different than these impacts.
7-19
-------
7.3 SOLID WASTE IMPACT
Add-on control systems generate solids as a byproduct of low-
ering glass melting furnace particulate emissions. At present,
these solids are not an economically attractive source of chemi-
cals, and they are not expected to be a usable source in the fu-
ture. In general, the solids can be disposed of through recycling
back into the glass melting process or through landfilling. Recy-
cling is technically feasible for many types of glass since the
chemicals in the waste are also present in the raw batch materials.
Landfilling has been popular in the glass industry since no addi-
tional equipment is required to dispose of the solids for most
glasses. Plant visits showed that landfilling was uniformly used
as the method of solid disposal.
There are certain limitations or disadvantages to these meth-
ods of disposal. Recycling may be limited in extent because of the
composition of the solid waste not being compatible with the batch
recipe of a specific product. Preparing the solids for landfill ing
may be required for solid wastes containing potentially hazardous
dusts. Also, if available landfill area were restricted, an alter-
nate method of disposal would be necessitated.
The amounts of solid waste generated in the control of parti-
culates from glass manufacturing is identical to/the amount of par-
ticulate removed from stack gases. Therefore, Option I will gener-
ate 4.8 x 109 grams (5,370 tons), Option II will generate 4.6 x
109 grams (5,000 tons); and meeting the state regulations will
generate 1.5 x 109 grams (1,670 tons). Comparing Option I and
Option II for solid waste impact shows them to be essentially equal;
Option I requires disposal of 7 percent more collected particulate.
Option I and Option II require about three times more solids col-
lected to be disposed of than that required by the state
7-20
-------
pns (SIP). These amounts are negligible as compared to the
totals. './-,;
** ; ' .*
,*"*** - -
S/Either recycling or landfiiMfig present minimal adverse envi-
,?*,'.- "'" ' f "* . ; / i ' '
^nfrental impact,- Totally recycling fhe',sol ids collected in the
control systems has no adverse .impact;. Landfill ling operations must
meet the state regulations, thereby'rhihimirtn^rthe potential for
adverse environmental impact. * -. * .' ; ' :. '»
7.4 ENERGY IMPACT
The energy impact assessment of ^theJ^gliTatory options and of
.-,; ", = _»: '^^f^^ f~- -- 3&*' * - "-. - n
the SIP provisions consists of dl|terimTn>ng; the numbers of model
glass melting furnaces built to produce 'th'e Amounts of glass listed
*&--(,,, "*;£*
in Table 7-1 and then calculating ['the'. energy! required by control
>j,..:-^ /"'; '^'ifK -'£&& " j.-^.:-^^ -' «.
' !"
,.. - ..- -
systems installed on these new furnaces^c.'l^fmatihg th!6"4otal
energy consumption of all industr.faTreategpcies for several mixes
of installed control equipment provides-a;r^Wge of energy impact
for each regulatory option and fdr^tjje; SIP regulations. ,^«*;*
-:-"-/-'"- V*.tr^;.-^^** ' ' ' ' "' *
Assuming that 26 weight pefcent,pf-|p£e^^^n§-blown glass is
produced in the smaller, 50 TPD fMr^^^epjra's^eStfmated from infor-
mation in NEDS, the integral num6lr\Tif-.-rn'ocJefe^lSi^S., melting furnaces
(refer to Table 6-2), on which the eriefi^f imp|l|-is based, are:2
Container Glass ' -> ' V~*f ::^:f^C?5 Furnaces
Container Glass
Pressed and Blown:
Pressed and Blown:
Soda-Lime 50
Soda-Lime
Pressed and Blown:
Other than Soda-Lime 50 TPD
Pressed and Blown:
Other than Soda-Lime 100. TPD
Wool Fiberglass
Flat Glass
2 Furnaces
6 Furnaces
1 Furnace
7-21
"'.- '''£$%££$'&.'.
&£jfM~ - f;~:JZ.I&i&*
-------
Electricity supplies the energy required by the control sys-
tems. The major portion of the energy operates fans which maintain
the pressure drop across the system. For fabric filter and venturi
scrubber systems, the minor amounts of energy used in equipment
such as solids conveyors and pumps are negligible as compared to
the amount used in powering fans. The power required by the fans
is considered to represent the energy impact of these control sys-
tems. For electrostatic precipitators, electricity used to charge
the plates is added to the fan energy requirements to arrive at the
energy impact. As all-electric melting has limited applicability,
the energy impact of this control system is not assessed in this
section.
The energy requirements for control equipment are extrapolated
from published values assuming that the fan power scales with the
exhaust flow rate and that the energy to charge the electrostatic
precipitator plates scales with the solids collection rate.5 The
energy impacts of each control;system, expressed as kWh/kg (Btu/
ton) are summarized in Table 7-9. These values represent energy
requirements at the glass melting furnace. Utility efficiencies
will be accounted for later in this section. Since the pressure
drop in fabric filter and venturi scrubber systems remains constant
for the regulatory options and the SIP regulations, the assessed
energy impact does not vary between them.
The energy requirements for several combinations of control
equipment within each glass category provide a range of values of
energy impact. The equipment combinations used in the assessment
and the energy impacts associated with these combinations are shown
in Table 7-10. Again, these values refer to the energy utilized at
the glass melting furnaces and do not account for inefficiencies in
electricity production. Electrostatic precipitators use the least
amount of electricity of any control system and therefore,
7-22
-------
Table 7-9. ENERGY REQUIREMENTS FOR CONTROL SYSTEMS
Glass
Industry
Category
Container
Pressed and Blown:
Soda Lime 50 TPO
Pressed and Blown:
Soda Lime 100 TPD
Pressed and Blown:
Other Than Soda Line
50 TPD
Pressed and Blown:
Other Than Soda Lime
100, TPD
'Wool Fiberglass
Flat
Fabric Filter
Satisfying
- Option I
- Option II
- SIP
kWh/kg
xlO-2
4.09
5.06
5.06
5.06
5.06
3.51
-
(Btu/ton)
x 105
(1.27)
(1.57)
(1.57)
(1.57)
(1.57)
(1-09)
-
V e n t u r 1
Scrubber
Satisfying
- Option II
- SIP
kWh/kg
xlO"2
6.42
-
, -
-
-
-
-
(Btu/ton)
xlO5
(1.99)
-
-
-
'.- .
-
Electros
Option I
kWh/kg
xlO-3
6.67
7.96
8.00
8.16
8.38
6.51
1.09
(Btu/ton)
xlO4
(2.07)
(2.47)
(2.48)
(2.53)
(2.60)
(2.02)
(3.37)
tatlc Prec
Option II
kWh/kg
xlO"3
6.67
7.96
7.96
8. 12
8.35
6.45
1.08
(Btu/ton)
xlO4
(2.07)
(2.47)
(2.47)
(2.52)
(2.59)
(2.00)
(3.34)
i p i t a t o r
S I P
kWh/kg
xlO'3
6.58
-
7.96
8.09
1.64
6.42
1.07
(Btu/ton)
xlO4
(2.04)
-
(2.46)
(2.51)
(2.58)
(1.99)
(3.32)
7-23
-------
Table 7-10. CONTROL EQUIPMENT COMBINATIONS AND ENERGY REQUIREMENTS
ase 1 25 Electrostatic Precipitators
ase 2 20 Electrostatic Precipitators
5 Fabric Filters
Container
Case 3 18 Electrostatic Precipitators
2 Fabric Filters
5 Venturi Scrubbers
Case 1 4 Electrostatic Precipitators
Pressed and Blown:
Soda Lime
SO TPD
Case 2 2 Electrostatic Precipitators
Fabric Filters
Case 1 6 Electrostatic Precipitators
Pressed and Blown:
Soda Lire
100 TPO
Case 2 4 Electrostatic Precipitators
2 Fabric Filters
Case 1 1 Electrostatic Precipitator
Pressed and Blown
Other Than Soda
ire 50TPD
Case 2 1 Fabric Filter
Case 1 2 Electrostatic Precipitators
Pressed and Blown
Other Than Soda
Lire 100 TPD
Case' 2 2 Fabric Filters
Case 1 6 Electrostatic Precipitators
Wool
Fiberglass
Case 2 3 Electrostatic Precipitators
3 Fabric Filters
Case 3 6 Fabric Filters
Case 1 1 Electrostatic Precipitator
Flat Glass
7-24
-------
represent miminum energy impact. The container glass equipment
combination of fabric filters and venturi scrubbers are chosen to
reflect an approximation of the currently operating mix of control
equipment. As discussed previously, venturi scrubber systems can-
not be installed to meet Option I. Uncontrolled glass melting fur-
naces in the Pressed and Blown: soda-lime (50 TPD) meet SIP
provisions.
Summing the energy consumption over the industrial categories
yields the energy impact range for each regulatory option and for the
SIP regulations. Implementing Option I requires from 20.4 x 106
to 56.8 x 106 kWh (70.1 x 109 to 193.5 x 109 Btu) at the glass
melting furnace. Assuming a utility efficiency expressed as 10,000
Btu/kWh required for 3,400 Btu/kWh (34 percent) and a heat content
of a barrel of oil of 6.023 x 106 Btu, the impact of Option I on
electrical utilities ranges from 60.0 x 106 to 167.1 x 106 kWh
(206.2 x 109 to 569.1 x 109 Btu) which is equivalent to 34,200
to 94,000 barrels of oil. Implementing Option II requires 20.4 x
106 to 71.2 x 106 kWh (69.9 x 109 to 242.8 x 109 Btu) at the
glass plants corresponding to a demand on utilities of 60.0 x 106
to 209.4 x 106 kWh (205.6 x 109 to 714.1 x 109 Btu) or 34,100
to 118,600 barrels of oil. Similarly, the SIP provisions require 19.7 x
106 to 68.9 x 106 kWh (67.3 x 109 to 235.2 x 109 Btu) provided
to the glass plants, which translates to 57.9 x 106 to 202.6 x 106
kWh (197.9 x 109 to 691.8 x 109 Btu or 32,900 to 114,900 barrels of
oil.
Because total energy requirements depend more on the types of
control systems selected than on the degree of emission reduction
achieved, the ranges of energy impacts of the regulatory options
and the SIP regulations overlap making comparisons between them
unquantifiable. However, two conclusions are apparent. In gener-
al, meeting the SIP provisions will require less energy than the
7-25
-------
furnaces classified in the Pressed and Blown: soda-lime subcate-
gory. Also, increasing the use of venturi scrubber systems in-
creases the energy impact of Option II and the SIP provisions.
Although no comparisons are made between the ranges of energy
impact of the regulatory options and the SIP provisions, these
energy ranges can be compared to the energy required to produce
glass. Using published values for the Container Glass, Flat Glass,
and Pressed and Blown Glass Standard Industrial Classifications,
the energy requried to operate control systems for these classifi-
cations ranges from 0.2 to 2 percent, 0.2 percent, and from 0.1 to
0.5 percent of the energy needed to produce glass, respectively.6
7-26
-------
8.0 ECONOMIC IMPACT
Chapter 8 contains four sections. In Section 8.1, the role of the
glass industry in the U.S. economy and the structure of the industry are
described. Tie four segments of the glass industry Flat Glass, Container
Glass, Wool Fiberglass, and Pressed and Blown Glass are identified and
characterized. Textile Fiberglass, a sector of the Pressed and Blown seg-
ment, was considered sufficiently large to identify and characterize as a
fifth segment. Several aspects of each segment are discussed: geographic
distribution, integration and concentration, import/export considerations,
demand determinants, supply considerations, and estimations of new sources.
In the Section 8.2, installed capital cost and annualized costs for
control of particulate emissions from new glass furnaces have been estimated
*
for 27 combinations of model plant size, control option, and industry segment.
Section 8.3 describes briefly other cost considerations and their
impact on the economic analysis of particulate emission control systems.
Section 8.4 contains an analysis of the economic effect of imposing
NSPS controls on furnaces within new primary glass plants. It was hypothe-
sized that no adverse effects on new plant construction would result. The
hypothesis was supported by the findings of the analysis, which indicate
that imposition of NSPS controls will not impact significantly on a new
grassroots primary glass manufacturing plant investment decisions. It may
therefore be concluded that construction of new plants should not be
impeded.
8-1
-------
8.1
INDUSTRY ECONOMIC PROFILE
8.1.1 Introduction
Glass making is one of the oldest industries, and generations of glass
makers have practiced and revised this craft. It is now an industry that is
highly sophisticated, still improving technologically, diverse, and essential
to the quality of life. Several firms in the glass industry are numbered among
the giants in business and industry internationally. As such, they are essen-
tial to both domestic and world economy. Their contribution to medical research,
protection of the environment, and engineering technology have made essential
contributions to everyone's well-being and comfort.
The scope of the industry still provides room for the artist who forms
dreams in glass to be shared with those who value creativity and craftsmanship.
It includes the tough spirit of the entrepreneuer who has found a way to make
our free-enterprise system work in a nation of big business.
As might be anticipated with an industry so large and of such long-
standing, the diversity of its products and enterprises cannot be fully
assessed. While large firms in the industry are publicly held, many of the
firms are small and privately held. Public reporting is not always disag-
gregated in such a way that it contains the researcher's desired specificity.
Privately-held firms report nothing publicly, and industry experts and rep-
resentatives become the source of reportable data.
This study attempts to assess the economic impact of NSPS on the
primary glass manufacturers within the glass industry. This part profiles
the overall industry.
The following order for the study was selected: (1) Flat Glass, (2)
8-2
-------
Container Glass, (3) Wool Fiberglass, (4) Textile Fiberglass, and (5) .
Pressed and Blown glassware. The study recognizes that textile fiberglass-
is identified as one sector of the pressed and blown segment. It was judged
to be of sufficient magnitude to justify inclusion in the study as a separate
consideration.
8.1.2 FLAT GLASS
8.1.2.1 Geographic Location
The 27 primary manufacturing plants now operating in the domestic
flat glass industry are located in 14 different states, which range across
the entire country. There are 4 plants in the state of Pennsylvania, and 3
plants each in the states of California, Tennessee, and West Virginia. A
fourth plant is under construction in California and is expected to become
operative in third-quarter 1978. ,.
Traditionally, the primary criterion for plant location has been
proximity to raw material sources. However, the plant currently being
constructed in California was apparently placed there to take advantage
of the growing market in the Western states.1
8.1.2.2 Concentration
The U.S. flat glass industry is highly concentrated, with only 7
companies participating in the primary manufacturing sector: PPG Industries,
Inc.; Libbey-Owens-Ford Co. (LOF); Ford Motor Co., Glass Division; Guardian
Industries Corp.; ASG Industries, Inc.; C-E Glass Division of Combustion
8-3
-------
Engineering, Inc.; and Fourco Glass Co. The four largest companies, PPG,
LOF, Ford, and Guardian, controlled 86.7% of total domestic production
capacity in 1977.2 PPG dominates the construction market and Libbey-Owens-
Ford dominates the automotive market.
The Justice Department filed a civil antitrust suit on May 10, 1978,
challenging the proposed acquisition of the Glass Division of Combustion
Engineering, Inc. by Guardian Industries Corp., alleging that the proposed
acquisition would eliminate competition between the companies in the manufac-
ture and sale of flat glass and would increase concentration in the U.S. flat
glass industry.^
8.1.2.3 Integration
The companies engaged in the primary manufacture of flat glass also
further process the glass by tempering and/or laminating. In most cases,
these companies fabricate a number of flat glass products for the construc-
tion and automotive markets.
In addition, four of the seven companies also manufacture products un-
related to flat glass and flat glass products. Ford manufactures automobiles
and light trucks, tractor parts and components, and communication and elec-
tronics systems. PPG manufactures textile fiberglass, chemical coatings and
resins. Combustion Engineering designs, manufactures, installs, and services
steam generating equipment for the electric utility industry,- provides design
engineering, and construction services for the chemical, petrochemical and
petroleum industries, and provides equipment, products and services to other
industrial markets. Guardian is engaged in photo processing activities and
8-4
-------
the manufacture and marketing of insulation materials.
Vertical integration in the industry includes the mining and processing
of raw materials. Captive distributorships and captive markets are other
characterstics of certain segments of the industry.
8.1.2.4 Import and Export Considerations
In the late 60's, foreign imports of flat glass offered severe compe-
tition to domestic products, particularly in the sheet glass market. At the
instigation of members of the flat glass industry, the tariff rate reductions
stemming from the 1967 General Agreement on Tariff and Trade were re-examined.
The ultimate result was the restoration in 1970 of higher tariff rates on
sheet glass imports. - .
The year 1974 was the turning point in the flow of flat glass imports,
and from 1974 through 1978, the balance of trade remained consistently favor-
able to the U.S. During 1978, the rate of growth for imports is expected to
surpass the rate of growth for exports, but the balance of trade is expected
to remain favorable to the U.S.4 The principal market for flat glass exports
is Canada, which is also the principal supplier of flat glass imports. Imports
and exports to and from Canada in 1976 totaled $26 million and $72 million,
respectively.
8.1.2.5 Demand Determinants
Demand within the flat glass industry is derived primarily from the
automotive market and the residential and non-residential new construction
8-5
-------
market. Demand, therefore, fluctuates in accordance with the economic acti-
vity of these two industries. Historically, flat glass shipments tracked
the 1974-1975 downswing into the 1976-1977 upswing in the automotive and con-
struction industries rather closely. Current flat glass industry forecasts
are based largely on expected new car and new construction demand, with the
outlook for 1978 being favorable; flat glass product shipments for 1978 have
been predicted to reach $1.98 billion, a gain of 8.5% over 1977.5
A third major industry from which flat glass demand is derived is the
secondary construction market (repair and remodeling), which one industry
representative has decribed as "a very important source of glass .demand
(which) acts as a smoothing influence on the fluctuating automotive and
construction markets".6
Demand is also influenced by federal and state regulations. The
Consumer Safety Product Commissions Standard 16 CFR 1201 ( Safety Standard
for Architectural Glazing Materials) became effective July 6, 1977. This
Standard requires all glazing materials used in high traffic areas of both
public and private buildings, including private homes, to be impact resistant
as defined in the provisions of the Standard. Although some delays have
occurred in fully implementing the Standard, it is expected to significantly
influence the demand for tempered and laminated glass.
Even more significant from the viewpoint of overall demand for flat
glass is the fact that a number of states have already adopted energy-effi-
cient building code standards that call for double glazing in areas having
cold climates. Not only is this concept of double glazing expanding into
warmer climates as a method of saving cooling energy, but the concept of
triple glazing is also gaining favor in residential remodeling projects.
8-6
-------
Other demand-creating developments in the energy conservation area are
solar daylighting and heating systems and retrofit glazing systems, which
add a second panel of glass to existing single-glass windows in commercial
buildings.
On the other side of the demand ledger is a continuing trend toward
the use of thinner glazing materials in the construction industry and thinner
windshields and side and rear windows in the automotive industry. The action
of many insurance companies in removing windshield and other car glass from
comprehensive coverage, and in instituting $50/$100 deductible provisions /is
also expected to have an adverse effect on automotive glass sales unless the
industry can develop successful counter-measures.? A possible offsetting
factor, however, is that there are nearly 130 million vehicles traveling
American roads this year, .a 20% increase over the second largest windshield
replacement year of 1973.8
Similarly, any lessening.of demand for glass in the automotive indus-
try caused by the downsizing of cars is expected to be offset by compensating
changes in car design and by the demand created by the trend toward custom-
izing recreational vehicles.
In each of the markets discussed above, there appears to be no readily
available substitute for glass. In a few rather limited applications, plastic
may be substituted, e.g., in solar applications. Intra-industry, qualities
of glaze or lamination may substitute for one another in use, but not in
price. Demand, then, becomes a matter of consumer taste in a price-sensitive
market.
8-7
-------
8.1.2.6 Supply Considerations
Estimated capacity for each of the 27 primary manufacturing plants
now operating in the domestic flat glass industry is shown in Table 8-1.
Flat glass production in 1977 was 2,594,000 tons.9 From 1972 to
1976, quantity shipments of float, plate and sheet glass grew at an average
annual rate of only 2%, but the real growth of flat glass product shipments
is expected to maintain a compound annual .growth of 4.2% for the next 5
years
10
Industry representatives have indicated that the flat glass indus-
try operated at close to full capacity in 1977 with industry capacity being
somewhat reduced, however, by the unusual amount of downtime and plant
closings attributable to industry upgrading.^ The past few years have
been a period of technological change for the flat glass industry, as sheet
glass and plate glass production were phased out in favor of more efficient
float glass production. An example of this kind of changeover is Guardian's
new float glass plant in Kingsburg, California which is intended to replace
unprofitable sheet glass plants and is expected to increase the company's
glass-making capacity by more than 50%.
The distribution of projected capacity growth between new and
existing plants is unkown. However, on the basis of 1977 production and
current estimated growth, if a worst-case scenario is assumed, where all
additional capacity must come from new plants, the equivalent of two new
plants will be needed to meet projected growth and capacity considerations
through 1982.
8-8
-------
Table 8-1
FLAT GLASS PLANTS
Producer
PPG Industries, Inc.
Libbey-Owens-Ford Co.
Ford Motor Company,
Glass Division
Guardian
ASG Industries, Inc.
C.E. Glass Division
of Combustion
Engineering, Inc.
Fourco Glass Co.
Plant Location
Fresno, California
Mt. Zion, Illinois
Cumberland, Maryland
Crystal City, Missouri
Carlisle, Pennsylvania
Meadville, Pennsylvania
Wichita Falls, Texas
Ottawa, Illinois
Lathrop, California
Laurenburg, North Carolina
Rossford, Ohio
Toledo, Ohio
Dearborn, Michigan
Tulsa, Oklahoma
Nashville, Tennessee
Carleton, Michigan
Jeannette, Pennsylvania
Greenland, Tennessee
Kingsport, Tennessee
Floreffe, Pennsylvania
Fullerton, California
St. Louis, Missouri
Cinnaminson, New Jersey
Fort Smith,
Clarksburg,
Bridgeport,
Arkansas
West Virginia
West Virginia
Capaci ty
400 TPD
450 TPD *
400 TPD
400 TPD
900 TPD
800 TPD
1,000 TPD
400 TPD
450 TPD
750 TPD
1,000 TPD
450 TPD
400 TPD
1,000 TPD
1,500 TPD
900 TPD
270 TPD *
900 TPD *
385 TPD *
400 TPD
70 TPD *
195 TPD *
500 TPD
225 TPD *
200 TPD *
450 TPD
*These estimates represent reported sheet, plate, and/or rolled glass capacity;
other estimates are measures of float capacity.
Source: Source Assessment: Flat Glass Manufacturing, U.S. Environmental
Protection Agency
8-9
-------
8.1.3 CONTAINER GLASS
8.1.3.1 Geographic Location
Container Glass manufacturing plants are placed as close as possible
to the markets they are intended to serve and plants are sometimes dedicated
almost entirely to servicing the needs of one nearby customer. In 1977,
there were 123 Container glass manufacturing plants located in 30 states
were identified. Clustering occurred on both the East and West Coasts, in
the North Central industrial states and, to a somewhat lesser extent, in the
South, which appears to be the fastest growing area.
8.1.3.2 Concentration/Integration
The high degree of concentration in the container glass industry is
evidenced by the fact that the four largest container glass companies are
currently reported as accounting for 56% of product sales, with the next
four largest companies accounting for an additional 21%.'2
In 1976, Owens-Illinois, Inc., was reported to have a 28.9% share
of the container glass market, with Brockway Company, Inc. having the next
largest share, 11.4%.13 Other companies ranking among the first 10 in 1976
market share included: Anchor Hocking, Thatcher Glass, Glass Containers
Corp., Kerr Glass, Indian Head, Ball Corp., Chattanooga Glass, and Midland
Glass.
The major firms in the industry are highly integrated, multi-product
manufacturers. These firms have substantial non-container volume in the
8-10
-------
United States, significant foreign operations, or both. Container glass
is distributed by contract with other companies who, in turn, use it to
package their products. Design and styling capability exists in each major
firm for the introduction of new shapes and forms of container glass.
8.1.3.3 Import/Export Considerations
There is relatively little export activity in the container glass
industry -- only about 1% of annual total value of product shipments ($33.1
million in 1976)14 is accounted for in this manner. Import activity is even
more limited, with imports having less than 1% share of the domestic market.
In both instances, foreign trade is considered economically unfeasible
because of high transportation costs and the easy availability of similar
«
products from local resources.
The major markets for U.S. exports are Canada, Venezuela, Australia,
and the United Kingdom. Major suppliers of imports are France, Italy, Canada,
and Mexico. Mainland shipments of container glass,to Puerto Rico, although
not considered exports, are worthy of mention because of their relative magni-
tude, $19.4 million in 1976.15
The tariff rates currently in effect on imports to the U.S. are the
end result of a number of successive reductions carried out under the pro-
visions of the 1967 General Agreement on Tariff and Trade.
8.1.3.4 Demand Determinants
Glass is one of the earliest and most universally accepted packaging
8-11
-------
materials. It competes in the market with other packaging materials, par-
ticularly plastics and metal. The market shares shown in Table 8-2 indicate
that the market share for glass has remained relatively stable from 1970 to
1976, and is projected to remain at that level through 1980. The demand for
metal cans has also remained relatively even, with a slight gain projected
for 1980. Share of market for plastics, however, has grown about 30% in the
1970-76 period, with a much slower rate predicted in the future.
Container glass demand is derived primarily from three major market
segments: food, beer, and soft drinks. Wines and liquors have a lesser
share of market, and relatively minor segments of the market include toilet-
ries and cosmetics, household and industrial chemicals, and drugs, medicinal,
and health containers.
The beer market, wh.ere glass competes only with metal cans, has been
the major area of growth for container glass over the past 10 years, with an
annual growth rate of over 7.5%.16 Shipments of non-returnable beer bottles,
which account for more than 95%17 of total beer bottle shipments, are responsible
in large measure for this gain. The popularity of this type of container is
expected to result in continuing gains unless restrictive legislation is enacted
on a wide scale. Non-returnable glass bottles have a held a significant cost
advantage over metal cans since 1975, when can prices rose sharply. In 1977,
this price differential disappeared as the container glass industry was impacted
by rising fuel and labor costs. However, it appears that container glass may
again gain a significant price advantage over metal cans as the rising costs of
steel and aluminum are fully passed on to the can companies. Prices of con-
tainer glass are expected to be lower at the end of 1979 than they have been at
any year-end since 1967.18
8-12
-------
Table 8-2
SHARE OF TOTAL PACKAGING MARKET
Paperboard
Metals
Plastics
Paper
Glass
Wood
Textile
1961
37.9
25.0
5.3
15.6
8.6
4.5
2.5
1970
34.0
27.8
9.1
13.7
9.8
3.8
1.5
1976
33.5
27.5
12.0
11.9
9.8
3.6
1.2
1980
±^\j\j
31.5
29.2
13.1
11.5
9.8
3.5
1.0
Source: American Glass Review, May, 1977, p. 12.
8-13
-------
Plastic bottles are not expected to impact on the beer industry
very soon because of the problems inherent in the chemical composition of
the plastics that can be used for these purposes.
In the soft drink market, contianer glass competes with both metal
cans and plastics. Glass shipments for the soft drink industry grew explo-
sively in the period between 1960 and 1970. -Thereafter, container glass lost
market share rapidly. However, after a gain of only 1% in 1976, container
glass shipments to the soft drink industry rose by 4% in 1977.l It is anti-
cipated that this sector will remain a source of growth for Container glass
over the next few years. The major competitors to container glass in this
market are metal cans, but plastic containers are making some inroads in the
area of larger-sized soft drink bottles, chiefly in the 32-ounce and over
sizes, which are cost competitive with glass and metal containers.
*
Prices of smaller-sized plastic bottles are still higher than those
for a comparable glass product. Plastic containers currently hold less than
1% of this market. However, the weight advantage of plastic containers over
container glass, coupled with expected reduction in production costs in
next few years, will probably result in a market share of 5-10% by the early
1980's20 unless legal restrictions against plastic containers are enacted.
The food industry continues to be the largest single user of con-
tainer glass, although, there has been Itttle growth in demand for this market
in recent years, and shipments were down by about 1% in 1977. This downward
trend apparently reflects the lack of growth in the food industry itself,
coupled with increasing competition from plastic containers.
*
In the wine and liquor segment of the beverage market shipments of
liquor bottles rose about 12% in 1977, after declining 3% in 1976. Wine
8-14
-------
bottle shipments rose about 4% after a 3% drop in 1976. This reflects a
trend toward larger-sized bottles in recent years, which is currently being
reversed by the switch-over to metric sizes that is required to be implemented
in full by January 1, 1980. Unit volume shipments should rise as the number
of containers required to package the same gallonage increases".
Container glass.in other markets, lost ground moderately in 1974 and
1975, showed a sharp increase in 1976, and settled back to a modest gain in
1977.
Demand determinants other than price and the availability of close
substitutes include a number of functional qualities that are either not
possessed by competing metal and plastic containers or are possessed in a
lesser degree. Chief among these qualities are resealability, inertness,
and reusability. Resealability is a desired characteristic in all segments
of the container glass market, because it helps to maintain the freshness
and quality of the container's contents and because it prevents or retards
spoilage and contamination. Among container materials, glass is uniquely
inert. Consequently, it can be used to package materials that might react
dangerously or unpleasantly with plastics or metals. Reusability, defined
as both refillability and returnability, is particularly important in the
beverage segment of the market. A number of legislative actions are cur-
rently pending which may make reusability mandatory. At the present time,
only container glass meets the requirements.
Other important demand determinants are taste, disposable personal
income, population, and age. The influence of taste as a determinant is
evident as the aesthetic appeal of glass, with its highly visible clarity
and purity, is translated into market appeal at the shelf level, particularly
8-15
-------
in the food, cosmetic, and wine segments of the market.
Rising disposable personal income impacts the container glass industry
favorably in several markets. Toiletries, cosmetics, wines, and liquors are
non-essential goods that react strongly to gains and losses in personal dis-
posable income. As a complementary good, the demand for container glass.
associated with these items fluctuates accordingly. Although demand in these
segments of the market decreases as personal disposable income decreases,
another segment of the market picks up some of the slack containers for
food items that are packaged for use at home and home canning jars and
bottles exhibit a minor degree of compensatory gains.
Population growth is a self-evident determinant of demand for con-
tainer glass, particularly in the food and beverage segments of the market.
Here, too, age is a determinant of demand as it controls the proportions
of the consumer population that are included in the groups eating baby
food or drinking soda, beer, wine, or liquor.
In addition to these primary determinants of demand, there are a
number of secondary forces that, empirically, have influenced demand in
certain segments of the market. These include such events as bad weather,
crop failures strikes, rising raw material costs, and energy shortages.
Finally, there is another powerful force operating in the marketplace
that is unique to the industry - the Glass Packaging Institute. This trade
association is active, well-organized, articulate, and influential. Among
its many activities is involvement in major consumer TV and trade adverti-
sing, market research and public relations programs. The stated objectives
of these programs being "to increase preference for glass packaging among
8-16
-------
prime influences in the trade and among consumers and to give the con-
tainer glass industry an aggressive, positive and socially conscious image
among its members and "the public".21
8.1.3.5 Supply Considerations
In J.977, production of Container glass amounted to 303.5 million
gross units, and domestic shipments were 307.8 million gross units, valued
at $3,674.8 mi 11 ion.22
The average annual growth rate in unit volume was 2.2% for the
period between 1971 and 1975. In 1975 and 1976 growth rates were 5% and
3%, respectively. Growth is expected to continue at a real annual compound
rate of 2.3% between now and 1982, with shipments reaching 340 million
gross by 1982.23
Severe cost pressures were evident in the industry in 1977. The
glass industry is highly labor intensive, and labor costs represent about
a third of total costs. A front-end-loaded labor contract negotiated with
the unions in 1977 resulted in a 13% increase in labor costs for the year.
Energy, which accounts for about 9% of glass industry costs, rose 20-25%.24
The desire of container glass manufacturers to relocate production
facilities closer to customers (for example, breweries) led to a wave of
expansion in the years 1972-1977. Industry analysts estimate that the
industry is now operating at about 91-92% of practical capacity and has
not worked at full practical capacity for a number of years. This, taken
in conjunction with the trend toward mandatory deposit laws, the inroads
8-17
-------
being made by plastics in some areas, and the trend toward lighter-weight
bottles leads these analysts to believe that no new expansion or construc-
tion of new facilities will be undertaken in the near future, except by
self-manufacturers,2^
The capacity of a typical new container glass plant has been speci-
fied by industry representatives to be 500 TPD. The three new plants that
are currently under construction and the expansions already underway may
result in significant excess capacity if demand is less than anticipated.
However, 1977 capacity equated with 1977 shipments of 307.8 million gross
units and with 1982 shipments projected at 340 million gross units, the
equivalent of seven additional new plants would be needed if a "worst-case"
situation existed and all additional required capacity had to be met by
new plants.*
8.1.4 WOOL FIBERGLASS
8.1.4.1 Geographic Location
Wool fiberglass is manufactured in 20 plants in 10 different states.
Five additional plants are used by one manufacturer to further fabricate its
product lines, a process that the other manufacturers incorporate in their
manufacturing plants.
Most of the industry's plants are located in the Eastern half of the
United States, but there are 4 plants in the Central states and 2 on the West
Coast. It is the practice in the industry to locate plants near major popu-
lation centers in order to insure the availability of an adequate labor force
and to take advantage of the variety of transportation needs, offered in urban
*This estimate assumes that 23.6 gross units
weight container in 1977.
8-18
= 1 ton, based on the average-
-------
areas.
8.1.4.2 Concentration and Integration
There are only 3 firms operating in the major markets of the wool fiber-
glass industry, with 1 new firm currently operating on the periphery of the
market. Owens-Corning Fiberglass, Inc. is the acknowledged industry leader,
claiming 50% of market share. Certain-Teed Products Corp. and Johns-Manville
Corp. split the remaining market share in almost equal proportion.26 These
three are large companies with integrated multi-product lines.
The industry is oligopolistic in structure, with no small regional
producers. Economies of scale are important and a large capital investment
is required for successfu} entry into the industry. The German firm of Knauf,
which recently became the 4th member of the industry, did so by purchasing a
high-density insulation plant in Indiana from Certain-Teed. The plant was
available only because Certain-Teed had been ordered to divest itself of the
facility by-the courts at the instigation of the Federal Trade Commission,
which currently has the industry under scrutiny.27
8.1.4.3 Import/Export Consideration
International trade is not an important factor in the wool fiber-
glass industry. Exports represent a very small percentage of total ship-
ments, with Canada being the primary market. Imports do not provide any *
competition for domestic manufacturers.
8-19
-------
8.1.4.4 Demand Determinants
Wool fiberglass is used primarily as building insulation and also in
acoustical ceiling tiles, heating and cooling pipe and duct insulation, and
in process equipment and appliance insulation.
Wool fiberglass has captured about 52% of the total insulation market,
and accounts for more than half of all residential insulation sales. Its main
competition can be broken down as follows. (1) Mineral wool, the principal
residential insulation material until fiberglass emerged in the 1950's..
Mineral wool carries the advantage at high temperatures since fiberglass
breaks down at more than 450°F. Mineral wool maintains a .12% share of the
housing area, but has been used more extensively in nonresidential markets.
(2) Cellulose, a macerated paper product, has grown explosively in the market
in the past two years because of the increasing unavailability of fiberglass
in many areas, and because of the extreme ease of entry into and the high
returns of the cellulose business. (3) Polyurethane and urea-formaldehyde
foams, still not cost-competitive with fiberglass, are finding increased use
in the building markets.
Generally speaking, the demand for wool fiberglass as an insulating
material is related to its price competitiveness with other products, the
ease with which it can be handled and installed, and the fact that space
is not a constraining factor in ceiling and floor areas. While supply has
tightened, so that its share of the various markets has fallen somewhat, it
is expected to maintain the status quo in most areas.
8-20
-------
In residential construction, new housing continues to be the largest
source of demand for structural insulation. Forecasts of demand are based
upon expansion of housing starts, amount of insulation used per unit, and
change in the housing mix.
Mobile homes also provide a growing market for insulation. Though
mobile home shipments remain well below peak levels of 1972-1973, the
demand for insulation has been stimulated by rapid increases in per-unit
use.
Residential retrofit remodeling is another significant source of
demand. Attempts to measure the number of housing units in this country
which are insufficiently insulated yield a figure of 24.0 million to 27.2
million. Reinsulation activity seems to be accelerating, and incentives to
reinsulate should continue to be compelling during the next several years.
In the non-residential market, wool fiberglass maintains a relatively
small share where it faces competition from wood fiberboard, tectum, gypsum,
perlite board, foam glass, and ceramic insulation material. Heavy density
wool fiberglass holds approximately 50% of the appliance insulation market,
where it competes with mineral wool in situations of intense heat and with
foam and foam board in cold appliances. In pipe insulation, a growing market
in recent years, wool fiberglass is the preferred insulating material for
fluid-carry systems of 0 to 450°F.
In air-handling, rigid fiberglass ductwork has been consistently
capturing a sizeable market share from galvanized steel. As an insulator,,
it retains heat and cold and does not "sweat". It retains approximately
8-21
-------
one-third of the ceiling board and tile market, where it competes with
gypsum, foam board, cork and wool fiber board. In nonresident!al roof
insulation, Owens-Corning, the only firm that produces it, competes with
a wide range of foams and expanded minerals.
8.1.4.5 Supply Considerations
Estimated capacity for the wool fiberglass industry in 1977 was
1,443,000 tons, 98.8% of which was utilized in producing shipments of
1,425,000 tons28 (see Table 8-3). This total was 43% greater than
shipments for 1976. Industry analysts indicate that volume of shipments
will peak in 1978, decline in 1979 as new single-family housing construc-
tion declines, and stabilize at approximately a 6% growth rate for the
industry between 1979 and 1982.29
Certain-Teed is building a new plant in Chowchilla, California
and production is scheduled to begin in 1979. No plans for new facilities
have been announced by any company beyond 1979. Consistent with industry
projections, it appears that increases in demand will be met by existing
facilities. The industry has demonstrated the ability to increase produc-
tion by 4 to 5% a year through furnace rebuild.30 Announced new capacity
plus industry expansion is estimated to be sufficient to meet consumer
demand in 1982.
However, for purposes of this study, and in order to present a
worst-case analysis, it is assumed that all additional requited capacity
8-22
-------
CO
CO
1
UJ
CD
oo >-1
1 Li-
co
_J
O
QJ O
1 ^
^ o
CO
1
z.
UJ
D_
i i
CO
«
-(-» -Q
CU -^.
S- 01
ro 3
Q.^^
CL, ro
c =E >
o
i
i *
ro ' -
13 d) r
V) Zl -i-
C i E
i t ro
:> -te-
ns
-t->
,0 ^
to
-M -Q
C i
CU
E
CU
IE IE
co-
4-> .0
C i
cu --^
S- 0)
ro ^
Q_t~~
o. ro
c:
o
)->
frt QJ p >
*=ji i
1/1 ro
S^SS
s_
CU
r" , *
4-> to
O 4->-n
C i
CU
Q-i
-C 2C
co
s- "cu"
Q.r
co.ro
O =C >
43
ro
13
tO CU f^-*
C 3 »
t t i S
ro
i > &₯
s-
o ^~-
Z3 CO
S- (-> JD
(-> Ci
CO CU
CLi
r- *i
CO '
cn
oo
CM
feq-
LO
o
LO
CM
to
<^-
O
r-i
O
to
CM
fa*
cn
t^
LO
t t
*ff
co
o
<£>
<-H
t <
CM
fa*
LO
CM
cn
fa*
CO
oo
LO
to
cn
^H
CM cn
to to
CM CM
rH 00
CM cn
CO P-.
CM CM
to oo
r*^ oo
o o
1 1 T 1
cn oo
CO O
CM OO
OO LO
P^» CD
r**» r^
i < t-H
oo >=f
i 1 LO
10 LO
r~- LO
CM CM
CM CM
oo oo
^- CO
0 0
* 1 1 1
CM OO
to co
to to
cn cn
r 1 t-H
oo
CM
o
CM
t 1
00
^J-
CM
I-H
! 1
1 1
CM
oo
o
cn
PN*
1 t
p*^
LO
LA
LO
oo
CM
0
CO
oo
1 (
LO
CO
cn
I t
LO CD
cn CD
CM OO
CM CO
LO LO
LO LO
oo oo
oo to
O CO
CM r-l
t 1 . <
00 CM
3- LO
CO OO
to CM
!"*" t""i
cn cn
<< < 1
to » i
LO LO
t-H t--
LO LO
CM CM
to to
P^ LO
LO to
r 1 t-H
P^ LO
CM -=i-
cn o
cn cn
t 1 r t
o o
oo co
CM CM
OO t 1
LO p^
CM CO
p^ oo
t-H OO
LO P^
t1 t-H
O i-H
00 CM
oo oo
t 1 «5tf-
P*^ O"»
O t-H
CM CM
P-. OO
CM CO
to to
LO «3-
"* LO
CM CM
CM P~
CO P-~
t 1 to
CM CM
0 LO
CO LO
CO O
i-H
t-H CM
cn cn
t-H t-H
oo
cn
CM
CO
co
LO
t^J-
o
cn
1 1
-
t^J-
^j-
oo
00
cn
^J-
CM
LO
CM
oo
to
CM
LO
cn
CD
oo
cn
t-H
t-H
00
p^.
cn
t 1
LO
oo
oo
o
o
LO
oo
=1-
cn
t 1
[^
cn
oo
CM
o
I 1
oo
t-H
CO
CM
cn
CM
oo
cn
oo
oo
CM
t-H
i-H
cn
1 1
oo
o
to
LO
to
LO
to
t-H
33-
I)
LO
I 1
=}
cn
CM
CM
LO
to
oo
LO
t-H
oo
00
00
o
r-H
t-H
LO
01
r-l
o oo
t-H «=!
"3- 0
P-~ OO
t 1 to
CO CM
t-H
CM O
cn LO
cn co
t 1 CM
CO O
tO CM
LO IO
to o
LO LO
oo -3-
OO CD
O LO
to p^
t-H 0
"=1- CO
CO OO
CO O
-H CO
p^. cn
«^j- CD
CO CD
OO t-H
i-H CM
jr
4->
§
S-
CJ3
-a
c
3
UJ O
to P^ r-i
P**» P^ £;
cn cn o
t 1 r-H O 1
^5
p^ t-1
«3- p-
0
CO CM
CM to
t-H f 1
^5
cn to
P^. CO
^^
i i cn
to cn
S5
^* I 1
, .
OO CO
t 1
t 1 to
CM CM
CL)
ro
E
to
0)
LO to JZ
o
^*- to . c:
'
,
r-
CO P- i-
CU
to cn 35
t 1 t-H
rt
-l->
i
I"
V5 ro
O CO Q.
Q)
CM CM CD.
t 1 t-H
cu
0
s-
CU E
S- 0
ro cj
m r*>« r**«
cri P-
CO 1 1
LO O CU
CO -4-* tO P** O
cu i/i cn cn s-
t-> ro «-H t-H 3
ro cu o
m s co
8-23
-------
will be met by new facilities. Industry representatives indicate that
typical new facilities will produce 200 tpd. The equivalent of seven new
facilities will be required in addition to the new facility currently under
construction in California in order to meet projected demand through
1982.
8.1.5 TEXTILE FIBERGLASS
8.1.5.1 Geographic Location
Textile fiberglass is produced by 6 firms in 16 plants in 7 states
in the United States. Three plants are located on the West Coast, 2 are
located in Texas, and the remaining plants are in the eastern portion of
the country. A high concentration of plants is apparent in North and
South Carolina and Tennessee where nearly half of the existing facilities
are located. Announced plans for new facilities indicate that the South-
Central portion of the country will be developed next for the industry's
market.
8.1.5.2 Concentration and Integration
The textile fiberglass industry is highly concentrated, with three
major firms dominating the industry: Owens-Corning with 58% of the capacity
in the industry; PPG Industries, 31%; and Certain-Teed, 6%. The remaining 5%
of industry capacity is shared by Johns-Manville, Reichold Chemical, and
8-24
-------
United Merchants. With the exception of the latter two firms, the remaining
firms produce textile fiberglass as a part or division of a multi-product
integrated product line. Owens-Corning is the leader in the textile industry,
with PPG Industries is the leader in flat glass, with fiberglass as a secon-
dary line of production. Both Certain-Teed and Johns-Manville are integrated
producers of building materials. Certain-Teed has served the textile fiber-
glass market with fiberglass yarn imported from Japan, but will soon commence
U.S. production for a fully integrated product line.
The textile fiberglass industry is only 40 years old. It was eleven
years after Owens-Corning introduced the product before a second firm entered
the industry. Capital investment requirements for entry into the industry
are large and have precluded the entry of many small firms. Similarly, few
firms have the capital required to engage in the extensive research and
development activities necessary to exploit this market. In primary end-
product usage, textile fiberglass competes in a market dominated by the
aluminum and steel industries, two of the largest and oldest industries in
the nation. Costs associated with the marketing of a relatively new product
in an old and entrenched market area, in and of themselves, are sufficient
to preclude the entry of small firms into the industry.
8.1.5.3 Import/Export Considerations
At present, no product with the same properties as textile fiber-
glass is ^ported into the U.S. Imports of textile fiberglass, mostly
8-25
-------
from Japan, are small. Certain-Teed has been a major importer, but it
is anticipated that Certain-Teed's volume of imports will decrease as its
new domestic facility comes on stream.
Only a small percentage of domestically produced textile fiber-
glass is exported by multi-national firms in order to strengthen the
existing markets of foreign subsidiaries or to assist them in the develop-
ment of new foreign markets. International affiliates of Owens-Corning
and PPG Industries service the European market.
8.1.5.4 Demand Determinants
The major end uses of textile fiberglass materials are fiber-
glass-reinforced plastic (FRP), tire cord, and decorative and commercial
*
fabrics. These account for 94% of the textile fiberglass market, FRP
being the major end-product.31 Other, less extensive, end-uses are
electrical wiring and applicances, and paper and tape reinforcement.
As a fibrous reinforcement for plastic, fiberglass presently
competes with only aramid and carbon, which have the capability of produ-
cing exceptionally hard composites, but which are several times more
expensive.
Generally, at the present time, FRP has distinct advantages over its
competitors, for example steel and aluminum. FRP can be just as strong as
steel, and is usually less expensive on a usage basis, though more costly per
pound. It is lighter in weight, and can result in weight savings of up to
8-26
-------
50%. FRP usually requires less total energy for production than similar
items made of metal, because it offers significant opportunities for molded.
parts consolidation.
At present, FRP competes with aluminum and steel in the transportation
market, with 60% of production going to the automotive sector. Its advantage's
' ^'
are: equal strength and durability, opportunity for parts consolidation,
'* st
resultant savings in cost and energy, and corrosion resistance. The most %
important current advantage, however, is the savings which it affords in
weight. Government regulations regarding fuel economy have made lighter-
weight materials most valuable in the production of automobiles and trucks.
It is anticipated that this market for FRP will evidence significant growth
in the near future.
The marine sector of the transportation market has always been a
significant one for FRP. Until 1976, it represented. FRP's largest market;
it continues to account for 70% of all hulls and decks produced for pleasure
boats. FRP's attractiveness is due to its moldability, seamless construction,
durability, resistance to corrosion, rust, and rot. It has been estimated
that the use of FRP in marine construction resulted in an 80% maintenance
saving, in comparison to wood and steel. Experimentation with the production ,
of large ocean-going vessels may influence demand for the product upward.
FRP has many uses in other areas of construction. One such area,
the corrosion-resistant segment, should be the most rapidly growing FRP
end-market during the next several years. FRP does not corrode, rust,
absorb moisture, or conduct electricity. FRP is being increasingly used
for pipes in chemical and other process industries that have corrosive
8-27
-------
environments, as well as for fluid transmission and sewer-water systems
applications. FRP pipe costs the same as ductible iron pipe, but more
than pipe of concrete, asbestos-cement, and steel. However, FRP pipe is
easier to install, because of its lighter weight, and is more durable.
Underground and aboveground tanks constitute about 20% of the non-
corrosive products sector. Although the initial installation of an FRP
tank can be more costly than one of steel, on a life-cycle basis, FRP is
considerably less expensive.
Other end markets of FRP include electrical insulators, electrical
apparatus, sports equipment, power tool housings, airplane stabilizers,
ladders, and so forth.
Non-FRP textile fiberglass uses include tire cord and decorative
fabrics, and commerical industrial applications.
The market for tire cord is shared by fiberglass, steel, and nylon.
Textile fiberglass competes with steel in the radial tire sector and with
nylon in the bias-belted sector. Although fiberglass radials are less
expensive, slightly lighter in weight, and offer better fuel performance
because of lower rolling resistance, steel radials continue to hold the
largest share of the original equipment market. Fiberglass tires hold only
15% to 20% of the tire replacement sector, which is twice as large as the
original equipment sector.
In the decorative and commercial fabrics market, textile fiberglass
competes with a variety of synthetic yarn, especially the polyesters. The
upholstery market has not developed significantly, and the curtain-drapery
market has been impacted by synthetics.
Solid cyclical growth is expected in industrial and commercial appli-
cations. Fiberglass has most recently replaced cotton in wire insulation
8-28
-------
because of its ability to withstand higher temperatures. Fiberglass alse
is beginning to be used as a cement and concrete reenforcement, although
asbestos fiber retains clear price advantages in this area. Another emerg-
ing market for textile fiberglass is as mat material for residential
roofing, because of its durability and fire resistance, and because asphalt
requirements can be reduced by about 25%.
The unique properties of textile fiberglass are only beginning to
emerge. Only in the past ten years have engineering and technology advanced
to the point where highly sophisticated and intricate products can be expec-
ted to grow with technological advances and with the new product development
that results.
8.1.5.5 Supply Considerations
Estimated capacity in the textile fiberglass industry in 1977 was
475,000 tons. In the same year, shipments totalled 420,000 tons, utilizing
88.4% of industry capacity.32
Table 8-4 suggests an increase in volume 'shipped, between 1976 and
1977, of approximately 15%. Analysts for the industry predict an increase
in volume shipped in 1978 and a decline in 1979. However, average annual
growth rate, between 1971 and 1982 is estimated at 9%. FRP is anticipated
to account for approximately 81% of total textile fiberglass shipments by
1982.33
Owens-Corning and Certain-Teed have announced new facilities in
Texas which will become fully operative at the end of 1979.34 /\t fu-j-j
operation, the Owens-Corning plant will have production capacity of 100
8-29
-------
Table 8-4
SHIPMENTS OF TEXTILE FIBERGLASS
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977E.
Compound Growth
(Least Squares).:
1965-77
1970-77
Shipments
(mil. Ib.)
287
323
309
394
477
433
478
575
695
662
547
728
840
Rates
10.3%
7.8
Value
$128.0
149.8
137.5
173.1
227.4
192.7
214.6
256.9
304.2
315.9
288.1
399.6
500.6
12.4%
12.4
Apparent
Value/1 b,
$.446
.464
.445
.439
.477
.445
.449
.447
.438
.447
.527
.549
.596
1.9%
4.4
Source: Commerce Department, Merrill Lynch estimates,
8-30
-------
TPD of textile fiberglass. Each of these plants is planned to .increase
production in the FRP sector of the industry where capacity is projected
to grow by about 85% between 1976 and 1982,'(from 325,000 tons to 600,000
tons J .35
Projected growth in this industry appears to indicate that some
new sources will be required to meet consumer demand. While no new addi-
tional facilities have been announced, discussions within the industry
indicate that both new and expanded facilities will be utilized to facili-
tate increased production.
This study assumes that all additional capacity required by the
industry will be provided by new facilities. Industry representatives
indicate that a typical new facility will produce 100 TPD of textile fiber
glass. It is estimated that the equivalent of 5 new facilities will be
required, in addition to the two plants presently under construction, to
meet projected demand for the industry through 1982.
8.1.6 PRESSED AND. BLOWN GLASS
8.1.6.1 Industry Structure
The primary descriptor of this segment of the industry is that each
plant manufactures glass and glassware that is pressed, blown, or shaped from
glass produced within the plant. Plants in this segment of the industry may
produce consumer and/or commercial glassware. Consumer glassware includes
products such as tumblers, stemware, tableware, cookware, ovenware, kitchen-
ware, and ornamental, decorative, and novelty glassware. Commercial glass-
ware includes products for the lighting and electronics industries and
various other fields, such as the scientific and technical market.
8-31
-------
The pressed and blown glass industry may also be subdivided into two
broad divisions: (1) the machine-pressed and blown sector, which is charac-
terized by relatively large publicly held firms that often produce products
largely for their own use, and (2) the hand pressed and blown sector, which
is comprised mainly of privately owned small firms that produce glassware
products that are generally more expensive, and are valued by the consumer
for their quality and craftsmanship.
It has been estimated that 50 firms in the industry produce approx-
imately 98% of all pressed and blown glass shipments.36 Corning Glass has
production facilities operating in all areas of this segment, and Owens-
Illinois, in most of them. More than half of the plants in the industry
produce hand pressed and blown glass exclusively. Approximately 10% of
the plants identified produce both hand-pressed and machine-pressed and
blown products.
8.1.6.2 CONSUMER GLASSWARE
8.1.6.2.1 Machine-Pressed and Blown
8.1.6.2.1.1 Geographic Location
Seven major firms were identified as producers of machine-pressed and
blown consumerware in 13 plants in 7 states of the United States. Four plants
are located in the Central part of the country, one on the West Coast, and
eight in the Eastern sector. There are fourteen plants that are primarily
in the hand-pressed and blown sector but which have been identified as having
machine-pressed and blown capabilities.
8.1.6.2.1.2 Integration and Concentration
Each of the major firms in this sector of the industry manufactures
8-32
-------
integrated multi-product lines. The sector is highly concentrated with 7
firms predominating: Corning Glass, Owens-Illinois, Anchor Hocking, Brockway,
Federal Glass, Bartlett-Collins, and Jeanette Corp. Corning produces the
largest-selling line of moderately priced dinnerware in the United States.
In 1977, 27% of the firm's net sales were in consumerware. Owens-Illinois,
through its Libbey Glass Division, is the leader in the introduction of new
products nearly 300 during 1977. Anchor Hocking leads this sector of the
industry in total glassware production. In 1977, 24% of Anchor Hocking's net
sales were in glass tableware. Sales volume for each of the remaining firms
identified is only about 10% of that of any of the three largest firms.37
8.1.6.2.1.3 Import/Export Considerations
The consumerware sector is heavily impacted by imports from a wide
variety of foreign countries. Table and kitchenware represent the largest
category of import in the pressed and blown segment of the glass industry.
Two of the multi-national companies, Corning and Owens-Illinois, have foreign
subsidiaries which supply consumerware to the European market. Export volume
is relatively small, therefore, from the two firms. Anchor Hocking, of the
three largest firms, is a volume exporter. In 1977, Anchor Hocking experienced
a record year for profitability from the export market. Much of Anchor Hock ing's
consumerware is transported by ship for foreign distribution. Historically, the
shipping industry has been impacted by labor disputes. The export market is
significantly impacted by the volatility of the shipping industry.
8.1.6.2.1.4 Demand Determinants
In this sector, consumer taste and relative price between similar items
are the major determinants of demand. Corning Glass and Anchor Hocking produce
8-33
-------
large lines of similar heat-resistant table and kitchenware. Owens-Illinois
and Anchor Hocking produce similar tumbler and stemware items. Brand loyalty
plays a large part in which firm's products are chosen by the consumer.
The availability of a wide variety of substitutes from the import
market, especially in the tableware, tumbler, and stemware categories, provides
a highly competitive structure in this sector. The impact of plastics and
paper tableware products cannot be assessed.
Consumer spending for'these products is small in relation to personal
disposable income. Therefore, consumers are highly sensitive to variations
in pricing for similar products. Demand appears to be highly price-elastic.
8.1.6.2.2 Hand-Pressed and Blown Glassware
8.1.6.2.2.1 Geographic Location
There are 84 firms that have been identified as producing hand-pressed
and blown products in 90 plants in 17 states. The greatest concentration of
plants is located in the Ohio-Pennsylvania-West Virginia area where 53 of the
participating plants are located. West Virginia has 28 firms producing in 30
plants.
8.1.6.2.2.2 Integration and Concentration
Many of the firms participating in this sector are small and privately
owned, producing in one plant only. Firms are integrated in most lines of
production, such as tableware, tumblers, and stemware. However, many firms
8-34
-------
produce hand-made illuminating glassware, such as lamp globes and bases,
which are sent to original equipment manufacturers for final assembly and
marketing. A small number of firms also produce machine-pressed and blown
items. Most hand-pressed and blown products are sold through manufacturers'
representatives. Finished stemware is packaged and sold from the factory.
There is little apparent concentration in this sector of the industry.
Manufacturers own their own molds and select their own designs or patterns.
In the past, customers placed special orders, and products were made to order.
The more recent trend is toward annual contract production. The sector is
characterized by entrepreneuership.
8.1.6.2.2.3 Import/Export Considerations
The major impact on this sector of the industry comes from imported
products. With import prices increasing because of devaluation of the dol-
lar, domestically produced items are now selling on the U.S. market at prices
equal to or slightly lower than imported counterparts. Mexican products are
sold in California at rates lower than in other parts of the country because
freight rates are low. As trade with Eastern Europe expands, however, greater
quantities of hand-blown stemware and crystal may be imported.
A small percentage of hand-pressed and blown products is exported to
Canada. Exports will continue to be small because of high freight costs and
their impact on costs of production in an industry where labor costs have
already accounted for 50% of the selling price.
8-35
-------
8.1.6.2.2.4 Demand Determinants " "
Hand-pressed and blown glassware supplies that portion of the market
that demands uniqueness and craftsmanship. Because large capital investment
in machines and in multiple molds is not justified in this sector, the volume of
production and the variety of items produced are low. Products are valued by the
consumer because they are in relatively short supply and because the image
of quality individual craftsmanship is attached to them.
j*
Machine-pressed and blown products are not perfect substitutes for
hand-processed items; they are high-volume, generally less expensive items,
produced under quality control conditions, without individual attention to
every item. Imported hand-processed items provide the most viable substi-
tute. Plastics have made some impact on the novelty glass portion of the
market. To the extent that uniqueness and the value of craftsmanship deter-
mine demand for hand-processed products, plastics are not perfect substitutes.
Hand-pressed and blown products represent luxury spending from per-
sonal disposable income. They may represent major expenditures. In periods
of economic uncertainty, consumers may defer purchases of these items or may
purchase a lower-cost, though imperfect, substitute. Demand appears to be
highly price-elastic.
8.1.6.3 COMMERCIAL GLASSWARE
8.1.6.3.1 Lighting and Electronics
This sector of the .industry includes automotive lighting glassware;
search light and other lenses; electronic tube blanks; tubing and cone for
8-36
-------
electric light bulbs, flourescent and neon lights; bowls and enclosing
globes, lamp chimneys, lamp parts, and shades.
8.1.6.3.2 Glass Tubing: Lighting and Electronic
8.1.6.3.2.1 Geographic Location
In this category, 4 firms were identified as manufacturing tubing in
16 plants located in 10 states. One plant is located in the Central Area of
the United States, and .the remaining 15 plants in the Eastern section.
8.1.6.3.2.2 Integration and Concentration
Four firms are identified with this highly concentrated category of
the industry: Corning Glass, General Electric, GTE Sylvania, and Westing-
house. These firms produce multi-product lines and are integrated in areas
other than glass production, as well.
Investment capital requirements for equipment in this category is so
large, as is volume demanded, that small flourescent and neon light manufac-
turers find it more economically feasible to buy from a large firm than to
produce glass tubing themselves. The major producers are large multi-product
firms, and light tubing manufacture does not account for a significant portion
of their sales volume.
8-37
-------
8.1.6.3.2.3 Import/Export Considerations
.Because-glass tubing for lighting is such a high-volume-of-production
category, import impact is minimal. A high percentage of tubing for lighting
is used captively by the firms that produce it, and is sufficient for their
own use. Neither imports nor exports appear "to significantly influence the
domestic market for lighting tubing.
Tubing for electronic tube blanks is influenced to some degree by
imported, assembled, end-product capacitors, resistors, and other forms of
electronic glassware components.
8.1.6.3.2.4 Demand Determinants
Tubing for the lighting industry is the high-volume, low-priced cate-
gory in this sector. Flourescent and neon tubing are the lighting end-products.
Demand for the product is largely from the non-residential lighting market,
and is determined by the infinite need for the product. Nothing substitutes
for glass in lighting, and as a result demand would appear to be relatively
price-inelastic.
Electronic tubing accounts for the smallest portion of the glass
tubing sector. Demand for the product is specialized and determined by
the precision manufacturing needs of companies that purchase it for their
own applications. This category is a very-small-volume, high-selling-price
area of production. As a result, price appears to be price-sensitive.
8-38
-------
8.1.6.3.3 Incandescent Light Bulb Blanks
8.1.6.3.3.1 Geographic Location
Two firms produce incandescent light bulb blanks in 7 plants located
in five states. All plants are located in the Eastern region of the United
States.
8.1.6.3.3.2 Import/Export Considerations
Because of the high volume of production in this industry domestic-
ally, .neither imports nor exports significantly influence the domestic
market.
8.1.6.3.3.3 Integration and Concentration
Corning Glass and General Electric produce virtually all of the
incandescent light bulb blanks manufactured in the United States for both
residential and non-residential use. Both firms are large, integrated,
muIti-product, multi-national companies.
As it is in other areas of lighting manufacture, capital investment
requirements for entry into this market are huge. These light bulb blanks
are produced on the Corning ribbon machine, the cost of which is estimated at $25
million. The very large inventory of finished goods that must be maintained
further limits available capital. The characteristics of this category sug-
gest a highly oligopolistic industry in which few firms participate and the
capital requirements are so vast so as to preclude the entry of small firms.
8-39
-------
8.1.6.3.3.4 Demand Determinants
There are no viable substitutes for,an incandescent bulb blank. Demand
for the product is consistent with demand for electrical lighting as a way of
life.
Consumer taste plays some very small role as a demand determinant in
the end product market. Preference for other forms of lighting, such as
flourescent or neon tubing for residential use, influences demand. Mercury
vapor and high-pressure sodium vapor lights, superior to incandescent lights
in the amount of energy consumed, are new products being developed. To the
extent that product substitutes are limited, demand appears to be relatively
price-inelastic.
8.1.6.3.4 Television Tube Envelopes
8.1.6.3.4.1 Geographic Location -
Four firms produce TV tube envelopes in six plants in Pennsylvania,
Ohio, and Indiana. Two plants are located in Pennsylvania, three in Ohio,
and one in Indiana.
8.1.6.3.4.2 Integration and Concentration
Corning Glass, Owens-Illinois, RCA, and Lancaster Glass are the only
firms participating in this category of the industry. Lancaster Glass is a
custom manufacturer of an integrated multi-product line of special purpose
products. RCA is a major manufacturer of television sets and other home and
8-40
-------
industrial appliances and components. Corning and Owens-Illinois are large,
fully-integrated manufacturers of glass multi-product lines and related
products.
8.1.6.3.4.3 Import/Export Considerations
The market for TV sets, the end-product for TV tube envelopes, is
sharply impacted by the availability of foreign imports. So much so, that
in 1977 the U.S. and Japan entered into an "Orderly Marketing Agreement,"
which limited the import of Japanese color TV sets for a three-year period.
From 1975 through early 1977, imports and their subsequent sales had risen
dramatically. A petition was filed by Corning Glass in behalf of all of
the domestic firms in the 'industry and eleven trade unions, before the
International Trade Commission. This resulted in the "Orderly Trade Agree-
ment." Corning, through its participation in COMPACT, the Committee to
Preserve American Color Television, monitors the Agreement for the industry.
8.1.6.3.4.4 Demand Determinants
Demand for TV tube envelopes is derived from demand for TV sets. Tele-
vision sets are a luxury item representing a large expenditure from personal
disposable income. The state of the general economy is an indicator of how
demand for a luxury item will rise or fall, depending on whether consumers
defer purchases until a more economically auspicious time.
While there are available substitutes for domestically produced TV sets,
8-41
-------
there are no available substitutes for TV tube envelopes. To the extent that
this TV component represents a minor portion of final product value and is with-
out a substitute product, price elasticity of demand is considered relatively
inelastic.
8.1.6.4 SCIENTIFIC AND TECHNICAL GLASSWARE
This sector of the industry includes industrial and technical glass-
ware, laboratory glassware, pharmaceutical glass products, ophthalmic lens
blanks, and optical glass tubes, rods, and cones. All of these products are
considered here in two sections: Glass Tubing: Scientific and Technical;
and Optical Glass.
8.1.6.4.1 Glass Tubing: Scientific and Technical
8.1.6.4.1.1 Geographic Lo'cation
Tubing for medical and pharmaceutical is produced by 4 major firms
in ten plants* located in seven states. With the exception of one plant
located in Illinois, all of the plants are located in the Eastern section of
the U.S.
8.1.6.4.1.2 Integration and Concentration
The degree of concentration is apparent in the small number of firms
participating: Owens-Illinois, Corning Glass, Schott Optical Glass, and
Wheaton Glass. Owens-Illinois and Corning, the two largest participating
*A number of these plants produce multi-product lines.
8-42
-------
firms, produce'a fully integrated multi-product line of glass tubing and rod
and tubing products. High demand items such as ampules, vials, and syringe
cartridges are included as are components for highly sophisticated biomedical
instruments. Producers in the industry sell to distributors, to small plants
and to one another for the production of precision glass products, ophthalmic
lenses, and small specialty job orders.
8.1.6.4.1.3 Import/Export Considerations
Tubing and rod and tubing are imported into the U.S. The extent to
which they penetrate the market for medical-pharmaceutical products cannot
be estimated from available public data.
Owens-Illinois and Corning have international subsidiaries in four
European countries which supply the European market. Exports primarily give
additional supply to foreign subsidiaries or open new markets for foreign
affiliate sales. - .
8.1.6.4.1.4 Demand Determinants
Demand in this category is influenced by the specialized nature of
the products being manufactured. Growth and advances in technology in
the medical and health-related industries will influence demand. Viable
substitutes for some products in this area include plastics and ceramic
glass. In that both Owens-Illinois and Corning Glass maintain full produc-
tion capabilities in both of these areas, substitutability of other goods
is not anticipated to influence demand in this market.
This category of the glass tubing sector represents the highest
volume of production in the sector. Products are medium-to-high-priced,
8-43
-------
and viable substitutes produced by the same firm create a demand which
appears to be relatively price-inelastic.
8.1.6.4.2 Optical Glass
8.1.6.4.2.1 Geographic Location
Five firms in this category produce optical glass products in 8 plants
located in New York, Massachusetts, Virginia, and Pennsylvania.
8.1.6.4.2.2 Integration and Concentration
The presence of only five major firms in the industry indicates a high
degree of concentration. Corning Glass, American Optical, Bausch and Lomb,
Eastman Kodak, and Schott Optical Glass predominate in the industry. Each of
the firms is integrated in multi-product-line manufacturing.
8.1.6.4.2.3 Import/Export Considerations
The major firms in this sector of the industry are also engaged in
end-product manufacture of a wide range of analytical, technical, electronic,
and health-related diagnostic instruments. The sector is highly specialized,
with products being produced largely by contract arrangement or to fill
intra-company needs. To that extent, small inventories exist and no real
impact is experienced from imported end-products.
No significant export market exists for the glass itself, except to
add supply to an existing foreign subsidiary. End-product export reflects
the specialized contract nature of this category and is relatively insigni-
ficant.
8-44
-------
8.1.6.4.2.4 Demand Determinants
Optical glass and instruments are a highly specialized, high-priced,
low-volume category. Demand is influenced by any growth of the electronic
market, and by an expansion in government scientific and research programs.
Optical glass for use in analytical instruments experiences increased demand
when environmental, industrial, and aerospace programs are initiated or ex-
panded. Any advance in biomedical research techniques brings about increased
demand in this category.
Plastics have become viable substitutes in some areas of ophthalmic
and instrument lens production. Plastics are, however, more expensive than
glass in a sector where prices are high for glass products. In uses where
weight is a factor, plastics have become a viable substitute.
The highly specialized, high-priced, low-volume nature of this cate-
gory of the industry appears to indicate that price elasticity of demand is
relatively inelastic.
8.1.6.7 Supply Considerations
Public reporting of shipments for the plants in the Pressed and
Blown segment was available only in terms of dollar amounts, and not by
volume of shipments. Neither existing capacity nor capacity utilization
rate could be estimated. Because of the lack of specificity and disaggregation
of the data, another methodology employing different assumptions was used
to determine the number of new grassroots plants needed to meet industry
growth for the next five years. .
Future growth in the pressed and Blown segment of the industry appears
closely allied with general growth in the economy.38 Items produced in the
consumerware sector and much of what is produced in the hand-Pressed and Blown
8-45
-------
category represent luxury purchases or purchases that may be deferred
depending on the state of the economy. Also, recent drop-offs in sales of
domestically produced TV sets, because of import penetration into the
market, resulted in severe curtailment in the TV envelope category of the
industry. For these reasons, this section of the study assumes that the
pressed and blown industry will grow at the same rate as GNP. 6NP is
projected to grow at an annual growth rate of 4% for the next 5 years.39
The number of plants in each category of the industry were estimated,
and the GNP growth rate applied for 1978-1982. Projections were adjusted
to accommodate the typical size of the new plant identified in each category.
The results are shown in Table 8-5. Industry sources indicate that during
the 1960's, major retrofit took place in plants in this segment of the
industry, and that additional capacity created by those expansions has
continued to be underutilized. New sources have not been announced and it
is assumed that projected production demands may be met by existing sources.
In order to present a worst-case scenario, however, this study assumes that
all projected demand will be met by new sources.
One new plant each will be required in machine pressed and blown
consumerware, incandescent bulb blanks, and TV tube envelopes. Two each
are required in glass tubing, optical glass and hand-pressed and blown
consumerware. It is estimated that this number of new facilities will be
adequate to meet projected demand through 1982.
8-46
-------
CO
O)
-Q
ro
U3
a
«c
a
LU
LU
CC
O-
t/5
LU
CJ
o:
ra
o
CO
o
LU
«=c
OJ
00
cn
O 3
a-
s_ co
CD a:
a
S
3 CO
z: u
i.
3
O
oo
O)
M- Q
O.
i- OJ
i O
I- i-
JD 3
tO O
-Q
03 C
a.
(C i
c a)
o
r- re
4-i S-
U CU
3 >
T3
a. o
o
a_
s_
O
CD
CO
O
LO
O
O
O
o
o
CM
o
cn
QJ
0)
O
a>
r-
O)
a.
X
0)
4J
3
a
r-
CO
O
O
IT)
O
to
CM
LO
O
LO
o
o
o
cn
co
CM
CM
umerware
to
c
o
QJ
C
c*
O
fl3
s:
o
-o
fO
-o
QJ
to
to
QJ
S_
Q.
1
-a
03
rc
QJ
fU
QJ
E
3
to
E
O
to
cu
.a
3
CU
Q.
O
"a
E
LU
^>-
1
r
CO
JQ
3
CQ
+J
E
CO
O
to
CO
a
E
(O
0
E
t i
to
03
r
CD
I
fO
0
t|
4-^
a.
o
cn
E
^
o
3
H-
CD
to
-E
-t-J
-M
JS
O
r
)->
Q.
to
to
O
o
r«s.
M cn
O
8-47
-------
REFERENCES
1. 1977 Annual Report, Guardian Industries Corp., pp. 2, 3.
2. Daily Report for Executives, The Bureau of National Affairs, Inc.
Washington, DC, May 11, 1978.
3. Ibid.
4. U.S. Industrial Outlook 1978, Department of Commerce, p.24.
5. Ibid, p. 23.
6. "Glass Demand Up for '78 Says Ford's Gonzaga", Glass Dealer, Vol.
28/No. 1, January 1.9.78, p.ll.
7. "Annual Industry Forecast", Glass Dealer, Vol. 28/No. 1, January,
1978, p. 38.
8. Ibid, p. 34.
9. Current Industrial Reports; Flat Glass: Fourth Quarter, 1977,
Bureau of the Census, March, 1978, p. 1.
10. Department of Commerce, Op_.__Crt., p. 24-25.
11. "Annual Industry Forecast", Glass Dealer, Vol. 28/No. 1, January, 1978,
pp. 36-38.
12. Department of Commerce, Op. Cit., p. 205.
13. Industry Review: Container Industry - Quarterly Outlook, E. F.
Hutton Institutional Department, New York, July, 1977, p. 28.
14. Department of Commerce, Op. Cit., p. 206.
15. Ibid.
16. Industry Surveys: Containers Basic Analysis. Standard and Poor,
New York, March, 1978. P. C121.
17. Ibid.
18. Container Industry, Seventies Research Division, Merrill, Lynch, Pierce,
Fenner, and Smith, Inc., June, 1978, p. 10.
19. Standard and Poor, Op. Cit.. p. 122.
20. Ibid.
8-48
-------
21. Glass Packaging Institute 1977, Glass Packaging Institute, Washington,
DC, p. 3.
22. Current Industrial Reports: Glass Containers: Summary for 1977,
Bureau of the Census, June, 1978, p. 1.
23. Department of Commerce, Op. Cit.y pp. 203,- 206.
24. PBT Views/March 1978, Prescott, Ball and Turben, Cleveland, Ohio,
p. 3. ; -
25. Industry Analyst, Arthur Stupay.
26. Barrens, April 24, 19.78,.p. 11.
27. Ibid.
28. Institutional Report, Merrill, Lynch, Pierce, Fenner, and Smith, Inc.,
December 9, 1977, p. 19. j :
29. Ibid., p. 17. ,
30. Ibid., p. 19.
31. Annual Report: Certain-Teed, 1977, p. 2; Annual Report: Owens-Corning,
1977, p. 4; Ohio Industry Review, December 28, 1977, p. 1.
32. Institutional Report, Merrill, Lynch, Pierce, Fenner, and Smith, Inc.,
December 9, 1977, p. 23
33. Ibid., p. 31. .... '
34. Ibid., p. 29. ;
35. Ibid., p. 30.
36 Energy Efficiency Improvement Target for SIC 32: Stone, Clay, and
Glass Products. Volume 2 Draft Report, Federal Energy Administration,
Washington, DC, 25 June 76, Appendices pp. 29-30. . ..:,
37. Annual Report: Owens-Illinois, 1977; Annual Report: Anchor-Hocking.
1977.
38. Source Assessment: Pressed and Blown Glass Manufacturing Plants.
January, 1977, p. 90.
39. The Data Resources U.S. Long-Term Bulletin. "The Economic Outlook for
1975-1990." Summer, 1976.
8-49
-------
8.2 COST ANALYSIS OF ALTERNATIVE CONTROL SYSTEMS
8.2.1 Introduction
Installed capital costs and annualized costs for the control of
paniculate emissions from new glass furnaces have been estimated
for 27 combinations of model plant size, control option, and
industry segment. The 27 combinations are listed in Table 8-1.
Typical parameters for uncontrolled exhaust from the se-
lected model plants for each industry segment are given in Table
8-2. These parameters were estimated by the U.S. Environmental
Protection Agency (U.S. EPA) with data obtained from plant
representatives, design calculations, and various reports of
source tests.
The regulatory options for particulate emissions and the
emission limitation from the New Jersey State Implementation Plan
(SIP), which is used as a reference to represent current require-
ments, as a baseline case, are shown in Table 8-3. The baseline
process weight regulation for glass furnaces is shown in Figure
8-1. With the exception of the 50-tons/day model plants in the
Pressed and Blown: soda-lime segment, controls would be required
in order for the model plants to comply with the process weight
regulation for glass furnaces in the baseline case even without a
new emission limitation. The regulatory options do not entail
any monitoring costs.
8.2.2 Capital Cost Estimates
Two major categories of costs have been developed: in-
stalled capital costs and total annualized costs. The installed
8-50
-------
TABLE 8-6. .CONTROL COMBINATIONS
Industry
segment
Container
Flat
Pressed and
Blown and
(borosilicat
opal, and le
Pressed and
Blown (soda-
lime)
Wool Fibergla
Model plant
size, tons/day
250
700
100
e, 50
ad)
100
50
100
50
100
50
ss 200
200
Regulatory option
I
0.2
0.2
0.3
0.5
0.5
0.5
0.5
0.2
0.2
0.2
0.2
0.4
0.4
II
0 . 4
0.4
0.4
0.6
1.0
1.0
1.0
1.0
0.4
0.4
0.4
0.4
0.8
0.8
Control device
ESP
Fabric filter
Scrubber
ESP
ESP
: ESP
Fabric filter
Fabric filter
ESP
ESP
Fabric filter
Fabric filter
ESP
Fabric filter
8-51
-------
0.
i
§
PI
I
s
00
3
J5 C
3<;s
3-H h
i|r
i ^
S4* 4*
1-5
8"
9 k.
S-
3 o
h
4g
I
h
II
Cbg
t-ttl
S
Is
*j
1
IndiMtry
« r-
S 8
o o
S 3
ft o
c« «»
0 O
o o
g S
* *
* S
s s
3
s g
Pi f»
Container
Plat
ft ft
e o
« <
at m
o, e
o a
I* *4
e e
i i
M a?
" "
4 M
4 0
O O
« «4
S S
^
«
JS
1
*
PreiMd and 1
(Boroiilioati
n rt e
e «t
e o e
r« n «
t «n o
w M e
e o o
a e o
0 O 0
0 0 )
« m
o e o
0 «l 0
n i
i| -
S Q
KM 5
o. i
M
S
c
1
1
1i
3
o
1
Co
4
g
ft
a
1
i
3
5
S
8-52
-------
w
3
O
H
CO
W
H
S
W
H
EH
fc>
O
H
EH
(^
«
O
fa
CO
§
H
E-i
CM
O
s
o
EH
D
O
CO
00
w
e
V
5 fl
Q) Q) C
C IH 0
H-H 4J
fl) O*«Q
01 4) rH
« IH
n
CM
(0
rl H*
S H o
H O\
3 -H XI
IH *
0 H C
* °
|S<
01 4J -!
a) a
06 0
^
c
V
0^
0)
j n in in «M CM ">
a o o o o o o
ia
0)
a
c
rH > rH 10
O Q) J3 E13
-P rH
T3 (fl *O ^-» O»
Co CO) IH
M 10-H <8 H 0
9) rH -H XI
C Tl -H "0 iH -H
H wo] a) i l
o> en
0) -r4
N
4J
ra
0)
a
i O
rH Vj
a 4J
a. c
a)
O 04
o H
c w
3 -a
rH 0) 0)
IB £ 0)
> H e
a .a
8-53
-------
00
<*
o
o
o
o
o
o
0)
o
(0
M-l
3
o
10
M-l
o
O -C
Ul
. o
II
o
oi
CO
CO
iJ3
0)
01
-------
capital cost of each control device includes the purchase price
of the major and auxiliary equipment, the cost of site preparation,
equipment installation, design engineering, the contractor's fee, and
interest during construction as shown in Table 8-4. The cost of
each control system is estimated by adding the installed costs of
each major piece of equipment, i.e., the control device, fan,
ductwork, and dampers. The cost of a stack is not included
because it is considered to be part of the glass-making process
and not an additional cost for particulate control. The equip-
ment cost of fabric filters is based on the estimates in Refer-
ence 1 for custom baghouses with reverse-air or shaker cleaning.
The equipment cost for the scrubber option is based on informa-
2
tion provided by FMC Corporation. The equipment cost of elec-
trostatic precipitators (ESP's) is based on, estimates relating
plate area to cost for conventional plate and wire precipitators.
Installation costs and indirect costs are based on published
information and engineering judgment. The total installed
costs calculated in this manner for fabric filter control systems
and ESP control systems are shown in Figures 8-2 and 8-3, re-
spectively. The estimated cost for the option of installing a
scrubber in the container segment is shown in Figure 8-4.
These figures also show actual control system costs as
reported by industry in response to EPA inquiries (Section 114
letters). All estimated costs and reported industry costs have
been indexed to January 1978 dollars using the chemical engi-
neering cost index. The reconciliation of estimated costs and
8-55
-------
TABLE 8-9.
COMPONENT CAPITAL COSTS ESTIMATED
SEPARATELY- BY MODULE
Direct Cost Components
Equipment
Instrumentation
Piping
Electrical
Foundation
Structural
Sitework .
Insulation
Paintings
Buildings
Indirect Cost Components3
Field overhead
Contractor's fee
Engineering
Freight
Offsite
Taxes (5% of material)
Allowable for shakedown
Spares
Contingency (20% of total)
Interest during construction
a Each component cost estimated separately depending on
equipment involved.
8-56
-------
> £
o _
UJ P
O D
IT)
CM
o
o
o
«t
o
o
01
«c
O
A
O
U3
O
CM
CM
O
O
o
in r
o
0}
-P
en
>i
to
O >i
l-l TJ
-p d
c -P
o w
o
5-1 -H
Q) A
I i .jj
H
H C
M-I -H
o 'O
M CO
M-l
-------
CO
LU CO
Q
Ul O
i O
I o-
CO Ul
LU O£
O D
O
O
C3
»
O
O
tf>
CM
O
O
O
CM
4J
Q£
LU
S£
Q.
IT)
CM
00
I
It)
CM
SUBL top 8Z6L AWfWVr 'ISOO IVlIdVO QHUVISNI
8-58
-
-------
industry costs is considered to be reasonable despite the large
variation in reported industry costs. Such variation is not
unexpected, considering the broad range of site-specific design
parameters and retrofit variations inherent in the reported
industry costs.
The reported industry costs for fabric filter systems are
plotted as a function of total cloth area in order to eliminate
variations in air to cloth ratios and flow rates. It can be seen
that the reported cost values are essentially equally scattered
about the estimated cost line. The reported industry costs cover
a broad range of design conditions and it is not known what
auxiliary costs are included by each company. As discussed later
in Section 8.5, the variation in installation costs can be sig-
nificant depending upon specific plant layout and furnace con-
figuration.
The reported industry costs for ESP systems are plotted as
a function of total plate area in order to eliminate variations
in specific collection area and flow rates. As in the case of
fabric filters, there is overall agreement between the estimated
and reported values. Several installations, however, have re-
ported costs significantly lower than expected for the plate area
reported. Discussions with plant representatives reveal several
reasons for these lower costs. For example, one of the units is
reported to be improperly designed and has operated for only
about 2 months over a 2-year period. Another of the units is
reportedly designed to achieve only 50 percent efficiency. Other
8-59
-------
reasons for the low costs in general include the ease of in-
stallation in a convenient location close to the furnace, low
installation labor costs, and the inability to accurately allo-
cate individual unit cost in multi-unit installations.
Figure 8-4 shows the reported industry costs for scrubbers
compared to the one scrubber case which was estimated. Also
shown are estimates for a scrubber system quoted in Reference 8
and inflated from a base reference year of 1974. No costs of
water treatment are believed to be included in the reported
industry costs and therefore, the water treatment costs were
deleted from the estimated value which is shown. There are in-
sufficient data to make any firm comparison of estimated and
actual costs, but the location of the points suggests that there is
good agreement between the reported and estimated costs.
Estimated costs in all cases are based on the use of heavy-
duty equipment, liberal installation allowances, and 20 per-
cent contingency. No attempt was made to include costs of
research and development, cost of land, possible loss of produc-
tion during equipment installation, or losses during startup.
It can be seen from Figure 8-3 that the relationship between
cost and the size of the control system follows an increasing
logarithmic relationship such that a doubling of plate area
results in a cost increase of about 1.5 times. This behavior is
often referred to as the "0.6 rule" because the proportional
increase in cost is equal to the ratio of plate areas raised to
the 0.6 power. The fabric filter curve in Figure 8-2 more closely
8-60
-------
o
o
o
D
D
D
co
_J 0 LU
co
LU
^fc ^^ ^^ ^^
O Z3 LU
u_ u_ LU _i cc o
O LU f O I LU
LU s: H-I a: g£
CD 3E LU O
2: O I t I a.
i
d
^H d
o -P
^i CQ
c w
O-H
ox;
-p
M
a c
f) -H
M QJ
O 01
CO d
co d
-P o
to
o
o
a>
rd 4J
4J t$
to £
G-H
H -P
CO
T3 Q)
Q)
£_l ^J
a-H
0)
I
oo
tn
fe
8-61
-------
follows a 0.7 power relationship. There are insufficient data to
define the scrubber curve in Figure 8-4. These relationships can
be used to estimate the costs of control for furnaces larger
and/or smaller than the model sizes used herein.
To determine the capacity of the air pollution control
system, the volume of exhaust gas from a furnace must be known.
Factors affecting exhaust volumes include: furnace size; pull
rate,' drafting system; combustion efficiency; checker volume; and
furnace condition. Exhaust volumes for comparable conditions,
however, are roughly proportional to pull rate. Therefore, given
a specific air-to-cloth ratio or specific collection area (SCA),
the control costs for a given .furnace size can be roughly esti-
mated using the appropriate power rule.
8.2.3 Annualized Cost Estimates
The total annualized cost consists of three categories:
direct operating cost, indirect operating cost, and capital
charges. Direct operating costs include:
0 Utilities, including electric power and process water
0 Operating labor
0 Maintenance and supplies, including chemicals , and
0 Solid waste disposal
Indirect operating costs include payroll overhead and plant
overhead. Plant visits showed that landfilling was uniformly
used as the method of solid-waste disposal. The industry there-
fore, receives no dust recovery credit.
8-62
-------
Water pollution control costs arise only when a scrubber is
installed in the Container Glass segment. Industry response to EPA
inquiries indicates that the predominant method of scrubber
wastewater disposal is discharge to municipal treatment plants.
A wastewater treatment system is included with the scrubber
control option because pretreatment standards and effluent stan-
dards for the glass industry are expected to require treatment by
1981.
Capital charges include: depreciation; interest; property
taxes; and insurance. Depreciation and interest are computed by
use of a capital recovery factor (CRT), the value of which de-
pends on the operating life of the device and the interest
rate. (An operating life of 15 years and an annual interest rate
of 10 percent are assumed.) Taxes and insurance costs are esti-
mated at an additional 4.0 percent of the installed capital cost
per year. The values used to calculate annualized costs are sum-
marized in Table 8-5.
Annualized costs for the operation of control devices are a
function of the number of operating shifts per day. For purposes
of this study, all facilities were assumed to be in operation 95
percent of the time, thereby requiring 8320 operating hours per
year. The operating labor required,is based on having one man on the
day shift for the dust-handling operation. For the small units,
collecting less than 300 Ib/day, the operating labor requirement
is reduced by one-half. No operator is included for the scrubber
option.
8-63
-------
^
E
|y|
Ej
CO
CO
."1
8
E"t
o
CJ
y^l
o
H
Q
1
ol
a
CO
CO
0
u
o
pi|
H
S
B
23
H*
*rfl
"4
CALCULATION OF
c
o
r4
4J
10
H
o
H
10
u
IM
O
"8
5
1
Cost component
-.
Direct operating cost£
*
o
1
CO
w
«
EH
10
0^
o to
0 JC
o 3
in n
(N 0
0 O
X X
id id
§, a
M M
0) 0)
Qj ft
01 0)
U) 0)
s a
y 3
I I
Utilities
Hater
Electricity
LJ
0 0
^^ 3
o o
»c
f^ O
X co
flj X
0) O
03 M
M O
O fl
X f-4
C 4!
a o
* 3
O' 13
tj (jj
S o
f itf
3 0
Z 01
Operating labor
Direct
Supervision
n
4J
u
o
u
C total capital
o
OP
vo
0)
4>
*
Maintenance and supplj
Labor and material
_-
C
^
0
in
o
H *
id v
H -g
M 3
C labor and mate
b of collected d
0 §
00 O
tn e
H <
Supplies
Solid waste disposal
|
in
in
>
X
a
0)
n
3
§
Q
I
§'
O O
(0 -P
U)
Sodium carbonate for
ber
Indirect operating co:
H
O
H
"a
a.
3
(0
«
0)
u
id
0)
JJ
^4
s
IM
0
Of
o
in
ta
H
a
M M
£ operating labo
£ operating labo
o o
<*> <*>
0 0
in
04 O
CO d*
r-4 T
Capital recovery
Taxes and insurance
M
O
A
^ m
. 3
>i M
v, id w
£ H 0
H«M
0 O
(N T3 M
00 00 f4
r« c
C H
H H
id u id
0) 3 01
a c
O >i
irl 0) *C
4J
H O O
O 0) VO
4J OH
o a -a
o -a
en id
0)
50) (Q
U 4J
C 01 C
O U 3
0)
*O t)V4 Q)
0) 0) O<
01 Hi M
id to
A _C H
H 10 O
LJ *W
c «
o -a u
H >1
a Electrical consumpt
Labor rate based on
required is 2920 h/
handling.
o
0)
n
a
a>
4J
(0
M
4J
B
H
4J
0)
U
8,
0
H
o
^
H
M
in
te
4J
C
H
cr
0)
. S 0
0) ' M
4J 0) 4J
10 O fl
M C O
0) U
M M
O 0) B
a
-------
8.2.4 Description of Control Systems
The control systems for which costs have been estimated are
fabric filters, ESP's, and a combination packed-bed/venturi
scrubber system, all of which are technically capable of achiev-
ing the various emission reductions. Each system includes all
appropriate auxiliary equipment such as: fans; motors; drives;
pumps; ductwork; dampers; walkways; and ladders. In general,
equipment selection is based on heavy-industry specifications.
Ductwork is 1/4-in. carbon steel with allowance for expansion
joints and access doors. Fans have radial tip blades with
totally enclosed motors. Motor size is based on cold startup.
Dampers are included for emergency bypassing.
Fabric filter costs are based on the air-to-cloth ratios
shown in Table 8-6. The fabric filters are insulated to maintain
the operating temperature at 405°F and prevent condensation in
the baghouse. Glass bags are used at this elevated temperature.
No cooling or other preconditioning of the furnace exhaust gases
is included.
Electrostatic precipitator costs are based on a conventional
plate and wire design. The ESP's are insulated and covered. The
required collection area is based upon the data given in Table 8-
6. No cooling or preconditioning of the furnace exhaust gas is
included. The larger ESP's and fabric filters are equipped with
dust-handling conveyors and storage bins, even though the total
dust collected is less than 1 ton/day. The relatively large size
of the units requires more than one hopper and consequently a
8-65
-------
H
CO
0
O
M
a s
o
0-0 C
01 iH
«i.
M
M-H v
O D1 (D
O -Q
O w ja
(4 9
M4 M
U
0
a
01
JS H (8
P-H 0)
O 9 CO
i-4 O1 9
U 01 O
U CT>
H o l
H-H a
O 41 m
25 C Ci
(d o
-4 -P
CU
4J
0)
i>
CO
Ll
Indust!
ja
oo
M PI
0)-0
r-l i-4 ft
.. 2
tn in
vo vo
0 O
O
o o vo
o o tn
M
to
vo
CM H iH O iH (M
000
o o in
*
si
It
ja oT
o id
C O . u
id -H *~
'O ^ (d
0) 03 0}
01 OfH
4J n u
«J 0) 0
CU CU CU
I-l H 11
u3 w"J ^
00 00
01 11 r» es v oo ^r eo o
000
o in o
-1 N
e
O in
iH in
J3 id
0~ 01
C 01 U
id S 4)
o'lH ^H
0) 1 i Si
H V«f
Sc5
Jj *
9 C
Oi 0
0) *H
m
0) ]>
5 Q
D c
B ot
o i>
iQ O*
0) C
K w
g-a
oT id
»,
^ w
O ^
10 *§
S'3
9 W
a)
01 C
Qj'H
w
li
5
tn a)
jj* *(
4J
id >t
UA
4i ft)
O TJ
-4-H
U >
0
O M
P a
«s
to e
id o
01
M *O
id 01
to
§5
So
U M
iH
H n
o c
°.s
at 4J
i to meet the regulatory
gineering Division, OAQPS,
Z c
X M
.(-1
g -a
Er e
oJ «
V)
0)
o **
10 *§
23
9 10
a)
0) C
«j Q
C.V*
03
(3 e
W
^
m a)
j.,-1 tj
4J
10 >>
MX)
JJ 0)
O TJ
-4 -H
U >
oS
4J a.
H -^
10 -S
h
to e
id o
01
M-O
id 01
to
e «
.2-°
Ua>
0 M
i-l *
r-l
e
0
H
jj
U
1-1
1
M
j3
9
M
O
CO
J3
1
M
a
1
1
4J
U
0)
c
o
-H
JJ
id
o>
H
E
01
O!
id m
0) O
id 4J
cs o
id
>i
en M
c 0
ca id
9 iH
9
C tn
O 01
H M
4J
id oi
9 A
tr-P
"M
c o
O >H
01
US
O H
5 9
1 O1
£ 01
O M
n
4J a
9 «
0) ai
o w
e
S«
4J
IH id
iH
a a
4)
4J U
Id^H
r-l
5
,Q
id
u
H
it
§
4J
S
1
^
2
8-66
-------
conveyor system. For the small units with one hopper, only a
dumpster box is provided. Cost of containers such as bags or
drums for transport of the collected dust to a disposal site is
insignificant relative to the total annualized cost. No dust ag-
glomeration or other processing of the collected dust is consid-
ered.
A scrubber system is used only in the. Container Glass segment of
the industry. Costs for the scrubber installation are based on
the system illustrated in Section 4. Equipment cost is based on
2
information provided by FMC Corporation. The capital cost and
annualized cost of water treatment are based on a system treating
a blowdown stream from the scrubber of 7200 gal/day. The
system consists of precipitation with lime and coagulants, clari-
fication, filtration, and sludge thickening. Sludge from the
wastewater treatment system will contain heavy metals and cannot
be disposed of indiscriminately. A sludge disposal cost of
$5.90/ton is included.
The estimated total costs of each control system are shown
in Tables 8-7 to 8-11. Not all of the control costs are attri-
butable to the regulatory options since all of the facilities, except
four, require controls to meet the baseline SIP regulations. The
costs of these SIP controls must be deducted from the total
control costs to obtain the cost attributable to the regulatory
options. For estimating baseline SIP control costs, it is as-
sumed that either the same type of control device is used to meet the
SIP as is used to meet the regulatory option, or an ESP is used
8-67
-------
ca
H
§
D
fe
CO
CQ
O
e
!S
«
O
fa
S.H
« s
EH W
O O
U H
CO
pa
EH «
Kd M
f4 S
D H
0<
H EH
EH S
PS O
2°
§
CO
EH
CQ
O
U
Hi
g
H
P4
I
00
I
*H
o*-
4* C
JJ
U H
f-H
3S
a c
c a
M O
r> oo
t-
a a
S 0
o e
c a
M fl
jj
0 a
C "-4 1-4
090
> O H
H *H JJ
O 14 O
o a o
u a.
IH
o
«
a
B
0
CO
H
fl
0
IB
3
O
H
VI
S.
I-l JJ
O B
OnO
HO
O
B
5 x
14
o"o
0
o
Jj
o
o
o
X
a
14
IB
^4
o
e
jj
B
O
U
I
^,
(0
1
fH
^
.*
f*
CU
o
cu
Ul
£
a
cu
M
co
£4
O
Cu
H
ca
&4
cu
0
cu
Ul
H 0
O 0
H -«4
4J >
B 0
g-o
«
*s^
lj) gj
N C
O H *H O
s ato jj
H
S
B-.
O
Sjjg
I
Q«O
vo r- t- in ao
co m in CM CM
1C » 1-4 O
f-4 i-l
A 0
fH CM CM O 00
m CO O C* rH
O> VO » CM O
1-4 r-l
» C CM
H 0 en
O O tj*
a IH >
H 0 0
O oo IB
O H
o o
i e
t" 4V
in
ve
e
o a
01
x
jj 0
*5
H S
rt O
SIU U-
Cl O
8-68
-------
CO
w
1
tf
D
fo
w
w
o
«
o
EM
i-q
§
EH
S EH
O 2
U W
S
H O
EH W
3W
D EH
U i<
H J
r
cu
(i,
O
to
EH
CO
O
O
J
£
H
CM
<
CJ
I
00
w
i-q
CQ
o
o
a oo o
0) r-
pH 4J -H
to ca
*» o
B o a
c a
m v
a a
GrH i-T
0) 3 0
> o n
H -H *»
U -U C
0 V4 O
9) 18 O
U4
U «-l
0
«
a
c
o
m
a
H
g
a
01
9
0
j i
M
10
a.
^ 4J
01 C
"§2
rH
X
a
u
10
i-*
O
a
^j
c
o
u
01
a
>
'O
\
A
>,
a
X.
D>
K
o
a,
H
CO
,
^
a,
o
0.
M
IA
.
E-I
cu
O
0.
i-l
0)
8
a<
CO
38
U -H
C 0)
u
a
»\
0} 03
N C
rt O
E an jj
14
0 fi
u C
18 -~
3 Q.
tji O
£
in m
0 «s-
V ^H
00 !£
r- M
m i-«
^i f>
r- r*
O4 CM
0 0
m eo
00 09
r* r*
o o
i-4
-------
4
°3
U 0
u
o o
0-H
H
1*
w
iHH
r* 0
Si
a e
e o
M U
O CO
H »CK
5 a
o
a u c
C 0
H ^3
a o
a 4J
o 0 »
Ci-li-l
O 3 O
> 0 H
H *H ,
§
a
>i
IB
o>
Jc
fH
s
I
o
H
e
u
o
JJ
0
-1
3
cn
£
0
o
o
I-l
X
b
0
H
H
li
o
a.
H
a
f .
o
3,
H
»
B
^4
O
Q<
4
U5
EH
CU
O
Oi
H
H
^
I
I
a
c
4J
§
H
JJ
a
O
r» i iH r>r»iN i-l
fHint^ to vino M
COM* 10 t^cnm o
eotNcn CM oim^ v
CN-CNCN M l-4iHr4 i-l
o a
«v« « tn in in in
CD CO CO 09 1-liHlH i-l
OOO O H 4>
Sm 0-H 0-H CO CO 0-H 0-H
(4 b IM b IM u U h>o
M o eo
ssa
U i» 4J O
0 m B O
co e
§
&
I
9
§-H 0 W Cj-H 0 O
4J O § -a u « o
** >i 4) C ** >t
rt 0 OX -H M0a41
UUU 0 1414UJ3
S S55 .
41 853 §
«U~ '" " O Oi**5***
M00*4 H00IA
M.-jJJJ -9 «»
8-70
-------
w
w
s
»
HH
W
w
3
o
w ^
s s
w
05 S
O O
fa W
w
tJ
« w"
EH S
§2
°i
W Q
B O
d^
D
^ 5
H 5
E-i O
t« ^
< CQ
&
Q
fa 3
O <
03 Q
t< W
W CO
O W
° §
J CH
IS
H
B
in
rH
I
CO
CQ
<
EH
!!
a u
o o
OrH
rH
o a
01
»~4 i"4
-H O
IB h
41 41
a e
S8
O oo
0) r-
rH 4JrH
3a
0
a o e
.M <-,
a 4)
a 4)
C rH rH
5 3 0
> 0 U
H -H *J
P JJ C
O b O
4) (8 O
«-i a.
U *w
O
tn
C
o
0
n
H
G
tu
01
IB
3
O
H
4J
10
a.
4) C
O fl
.82
&
S
IB
rH
3
cr
0)
Bi
t~
e
>-a
J
B,
O
O
4J
O
O
O
rH
X
B
H
rH
rH
41
C
at
u
S
>
as
\
.Q
>.
IB
*O
\
01
rH
O
H
4J
C
o
o
»
N
H
a
§
H
a
o
o
o
rH
M
a
«
P_}
0
£4
O.
O
IH
to
s
0
a.
H
.
O.
o
a.
CO
£
o
a.
ca
.
u
H
0
(B
O
a
c
o
*J
o
rH^lO 00 f> ^ rH r*
jj o ja o
o i*^ to 10 c*i ^ i-i r**
.H o> c>» rH r> VD r> vo
H rH iH
« a
CO CO CO CO
** "* ^ *
(N « \O \O C4CMO4 d
cncocn co mcnui en
.-. r-(
r** ^* ^* P* in *rt i/) trt
Uk4Uh U^lUij
*H O ""1 1) -H <1) -i-l 4>
M -P M 4J W4J>44J
W W .-u a
a rH 41
-icon
H O 4>
a &
> 41 e M
« > 0) U
4) -U
O-H a e
rt & ><<0
u a
an o
*J M U
5O 41 O
*> *» IH
rH 41
O "d-H H
X VH a
a A
3 u M
SJll .
IB no
a 3 a
HO a>
Si O » »
P O r-
O -rt
o. 1-1 a
n B >i
U JS'rt O
4J 4J e
e o a a
IBrH OI-H
u o
B \4J-H
H w a w
HOW
4) IB O 0
§J3 IB-H
4J o -g
O-H-H B
J-t o
a, a. a,
M u a a
tn 41 « o
4J kl
4) rH 4J U
91
IM
T3 O
U
0 S
H O
4J-H
a o
4> 41
o
OrH
oS
"8
a B
a
g t
to
a «>
5-..C
19 *»
9) O
rH B
H O
a
S
ts
M
«8 1
Jl 3
« 41 O
U U £
0 4J JJ
0,09
O -U
a s
Eo A
a
o
o n
TJ8.
3-
IB rH
B
§°
O<
O B
82
5S
M
a 4J
Cl
4J e
§>,
H O
41 B
a 4)
«-H
O
-------
I
W
D
fa
CO
e?
2 O
o m
u co
H pq
|j «
p$ H
*> B
OHO
a a o
«4 a
HIM
o
B
B
0
H
m
B
1
S
a
I
a
a
cu
1-1 u
3!
X
e
5
-i
o
41
§
8
S
a
s
a
a
o
=U
O
CU
IH
09
^
O
cu
M
01
O
cu
M
09
,
§
cu
03
3§
b-H
u >
J«
.1
a B
2!
O-H
<-4 4>
O>O
» 0 i-l M
O ^ ^ W
in <-i
eo vo in p>
in o to to
vo vo M in
CM i-t CM i-t
*V VO ^ AD
in n in tn
) in H in
N «J-H CO (0-1
O 0 0 O
o o o o
CM CM CM CM
M H H M
I
4J U
O IP H*
a IM J o
0 4J H Dt
2J§ii«
S640.?
14 OJ-H
2 §1*1
g-H-rt « »4
a *>
JS£T5I>,
3MM CO
83.183
oSSoj?
U U >iU]>-4
*S B O
b o B Oin
otai jsc
^"Sfio
b B « o
H BCU b-H
B IB M 0) 4J
MH09J:.9
III.3
8-72
-------
that can be more readily sized to the lower SIP efficiency re-
quirements. Even at the low efficiency required in some cases to
meet SIP, a centrifugal collector will not be suitable because of
the extremely fine particle size of glass furnace emissions.
8.2.5 Modified/Reconstructed Facilities
The cost for installing a control system in an existing
plant that has been modified, reconstructed, or expanded (given
the same exhaust gas parameters) is greater, as a result of spe-
cial design considerations, more complex duct runs, and similar
features.
Estimating this additional installation cost or retrofit
penalty is difficult because of many factors peculiar to the
individual plant. Configuration of equipment in the existing
plant governs the location of the control system. Depending on
process or stack location, long ducting runs from ground level to
the control device and to the stack may be required. A sizable
increase in costs may be incurred if the control equipment must
be placed on the roof, which may require steel structural sup-
port. Modification of the electrical substation may be required
to accomodate an increased electrical load. Other cost components
that may be increased because of space restrictions and plant
configurations are: contractor's fees and engineering fees. These
fees, estimated at 15 percent and 10 percent, respectively, under
normal conditions, can be expected to increase to 20 percent and
15 percent, respectively, for a retrofit. These fees vary from
place to place and job to job depending on the difficulty of the
job, the risks involved, and current economic conditions.
8-73
-------
The requirement for additional ducting can vary consid-
erably, depending on plant configuration. For purposes of this
study, it is estimated that approximately 50 percent more ducting
may be required to install a control system in an existing plant.
If the space is tight within the plant, it may be necessary
to install the control equipment on the roof. It is estimated
that a roof-top installation could double the structural costs.
It is estimated that 10 "percent is required to tie the system
into the process. This work would most likely have to be done at
premium-time wage rates in order to avoid extensive downtime and
loss of production.
Consideration of these additional cost factors shows that
the costs of a retrofit installation may be expected to run
approximately 20 percent higher than the cost of a new installa-
tion.
8.2.6 Cost-effectiveness
Incremental annualized costs (that is, costs due solely to
the regulatory options) are divided by the incremental quantities
of particulate removed to obtain cost-effectiveness quotients.
Tables 8-12 to 8-16 list these cost-effectiveness quotients for
each of the 27 control combinations. The quotients for all the
control alternatives are plotted by industry segment in Figures
8-5 through 8-8. It is clear that the quotients vary both with
the plant capacity and the control alternative. The quotients
vary from $0.39 to $6.30/lb removed, and are strongly influenced
by the SIP control efficiency relative to the regulatory option
control efficiency.
8-74
-------
O
EH
>1
H*O
4} (U A
«IT
S§
1-1
rH JJ
ScS1
3
« 13
4J OJ IH
BI-I N n >,
O IB-H 4J X
g JJi-H O O
OO^OO
ki a a o o
SC rH
C >
HI ia
WTJ -o
JJ O rH M
g'£3^
i -H -n o
3 ig ao
rl 3 IB O
U C U rH
c e «
HI 10
(0 A
It . ^
S V - >.
0) O 01 V,
e a) ^ o
Q) b tfl O
M T| o O
O TJ O rH
C «
HI
a o>
Eg JJ
«i a a A
CrH . H rH
O 3 rH >.
> O rax
JJ JJ O -W
(U *0 S C
w O. U 3 O*
MH Jd
U UH
o
rH
O 0)
hi 0
4J-H
C >
O 0)
O TJ
>,
a
rH 4J * ^v
o e a 01
a a N c
S-i-as
^
K
o c -^
CdO
O «3 S
« " « 0) C
J< o -5
O «
« « IH
^ 4J TJ ffl
3- S-H &
41 O O
rl VI 4J
gvjj o
_ -H O rl
09 43 u *r4
c a TJ
o So, c
rl 01 M -H
O S,(A
« 0 B
B ^"S 41
! 8- 5
* rl O
rH ^ O C
W O 0 HI
«J A 0
0}
0)
8-75
-------
CD
(-1
00
u i
a '
5H4J
S-H
§
*o
81
^HO
e 0.0
9 oo
o » >i
««><
Id BO
3 00
a oo
ss
* *
« 9>H >
> O 0,-X
H-H > e
8h g-H
a 5 c
IH QiM 9
44 *
il
oca
a a N
OrH'H
X a a
9 PU
CT>O
o in
at »
14
o
15
S
I
9,*
a a
B -O
r- i-i
S 2
1
tf
o &
a
I
1
1
1
fi
§ 8 t!
H « a
m 4t t*
I
M we o
g. 8^ &
" s«
, I
« g °
§ §&
M
(a
s
8-76
-------
I
00
IS
^
n*a
Q,>rH
? §
a u
rH -U
O-H O<
OCX
3
i-H 73
«J » - K
i) iH M 01 >i
C .S-H .UX.
Q) * rH 0> O
i o n oo
0) 4J 3 OO
h C rH
8 g w
M
0-
rH 1Q » U
fflrH >.
4* N fl \
C -H4J O
01 i-H-rl O
6 a ao
iS 3 IflrH
U C O v>
u c
C it)
3 A
> u
c .u a >i
41 O -W V.
fai-oi o
u-o o
h ~*o o
O 13 rH
e «
M
a 01
n -u
01 « o ja
01 3 rH >,
> 0 I8<
H -rl > (0
4J 4-1 O -U
O U E 1
v a a c
»-i D, h 3 t7>
U
o o
O T)
19
a
rH 4J «V^
01 C 01 01
a « N c
O rH-H O
I a. n *>
M
s§
rH JJ
3 a
o>o
a
»
9V CM fn
^ A ^ 1^
CM in *n »-i H
m m r* o\ oi
(N 0 P* CO »
m d oo **i <*>
r> oj o\ r* in
rj
m r^« ^o M w
^ m ^o m «r
^H
M O% 00 O VO
rH \£> fN CS «H
O O O O O
SO O O O
m « M NO
m » i (*) *H (N
*>i » C4 --V n
o o o o o
o o o o o
r- oo co » oo
O CO « 00 f
O J^ O l-l 0 W
-H fl ^
(0 *H {/] C/} fl *H (0 *H
(M>M U U hM-i faU4
o o o o o
o m tn in tn
f-t
HI M l-l
41
a
o -o
a
4J 01
« t
rl «
-H >
8 0
§ rH
5 2
e >.
u a
>, g a
« -H -O
> > C
a 41 e
: 8
ll
a JJ
a a
1
01
a
e o e
-3 3 tt
JJ rH O
CU 3 U
01 01
>. es 41 e
M U iH
O 1-1 -H -U
S °5, "
rH 4J TT
3 a
S1 8-3 -
U U 4J
CT.J u
i c e 41
n o h
a jj o -H
c a -a
o u a. c
S 4J O T3
So 3
o *> -(
a. K o
M -H O C
u a o M
8-77
-------
I
w
g
cr
B
s
§
o
I
< W
13 « CO
§§£
^D
E-JO
SS
H 4J
9
3 -o
w O I8°v
*J 4J Q *>
O «J 01 C
-<-H O
S O.B.U
&e
II "3 ft
(DO
;
«
00
tn
r-
,,.
H
r»
CM
H
O
in
o
0
m
o
o
CM
a.
to
Ed
0
o
S 5 3 * S S
CM » » in in m
in r> t» Jp 2 en
in r- r* r» «n m
oo, o. «JJ2
CM co in fH ^r r*
en CM o « g CM
r-l CM 04 M CM CM
H
S n-H (8-H tn en «-H
U Cu
i! °I5 S
3 S* S.
17
h
1
a
§
,-H
':- tB
' 01
*4
§
a,
HI
CO
«"
O>-l
°s
o-u
c c
18
U 111
8.3
o
«0
8li
28
A
0
*>
o
41
V4
1
rt
1
«<
g
M
O
I
8-78
-------
H EH
HH1 3
M W
< S
U O
o w
1-3 to
.4
H
o
0 6
(Q M
rH -W
S-rt a
c x
3
4J (D . U
Si-l N 0 >i
IS-H *> \
g 4J'-' CO 0
S238 §
g H '5
M «3
la TJ o
4J (U - U
fl) -H 1" X
§rH4J 0
nj-H 0
14 3 a o
O C 1" H
c c o «
H, ig
*H M
« «
4J 4J . W
C O 0) >i
a] flj 4J \
c! H 0 §
25 o 2
O "^
c «
H
a ,
*< «H > 09
4J 4J O 4->
u n C c
**4 Q< 1^ 3 O»
W «W ^
O
Is
4Jt4
§S
O"O
fc.
id
o
U C O M
a i
S {
Q
id-H
r-l 43
a a
o< o
M
^o o m *£
« ^r r» ^
r*) o IH o
o* o* ^o 04
rH 03 CO O
r^ o <*i rH
^ ch *l* G\
in »H
O r-l (N
rH
O in rH N
in r-i
O O O O
o o o o
(N N *T <*
f^ r^ «
4J-O V
8S 4!
CT14J O
ecu
H O H
^° 5
Sfr5
25,03
° O
* o e
a o M
J 0
8-79
-------
qi/$ 'SS3N3AI1D3JJ3-1SOO
8-80
-------
00
uluj
LU
00
"* i-
00
u-u-
U.U.
03
s§
___
i i
LI
1^
CO i
CM
JL
CO
»
(O
o
14 O
I in
Q_
O II
M
CO
Q.
O
c
o
-a
a)
0)
M
(U
U)
(0
c ^
o -o
H rt
r-l C
4J »
fi H
O< flJ
o &
o
o -
0)
en 4*
W (0
o o
C-H
0) iH
>-H
"5 O
O M
(U O
4J
(0
O
00
0)
CSj
00
qi/$ *SS3N3AIlD3Jd3-iSOO
8-81
-------
00£
t« UJ
OH-
Q. 03 _l
CO <«-«
UJ U.U.
0
UJ
o
01
?= <-> 2
UJ o
CO ii
sr
^-L
CM
03
£ o O
C/1 I I HI
UJ ( l_
a: SI £
Q- O O
m
in
cr>
GO
«-i O
o
o
M.
ii
>
1-1 O
O
Lf)
CO
l
LU
a
o
'SS3N3AI1D3JJ3 1S03
'S
(0
0)
W
CO
a>
5-4
ft
0)
x:
-P
n
o
m
C
O4 0)
O o)
M
JJ ^
fi O
o e
O-H
tH
IW I
O nt
Tl
0) O
(0 (Q
(U-^
fi
(1)
4J
O
0)
M-t
>W
0)
-P
CO
O
u
r-
t
CO
0)
8-82
-------
.p
G
Q)
01
01
to
nJ
H
>;g I
Z LU _
LU ll O
ii O O
O ii CM
CD J-> Ll_
UM ^L-> I*
CO U- LU
CO
CO %<
CO &< CM
: vo w
i cn
« H "
CO ii ii _J
o o -J
_J l-l H-H LU
O I h- Q
O 0. 0. O
3 O CD 3E
U>|
cva
co'
O
51
o
o
o.
o
LO
CO
CM
0)
H
Pn
o
i
0)
.-p
M
o
4-1
01
c
o
H
ft
o
o
S-l
-p
01
en
0)
H
p
u
(U
I
-p
01
00
00
0)
M
qi/$
8-83
-------
The variation in cost-effectiveness, between industry seg-
ments is due mainly to the changing SIP efficiency. It can be
seen in Figures 8-9 and 8-10 that the results do not follow the
expected pattern of increasing cost-effectiveness (lower cost-
effectiveness ratio) with larger plant size and less restrictive
control. This expected pattern can be seen, however, in Figure
8-7 for the Pressed and Blown: soda-lime segment. In this partic-
ular case, the SIP baseline is very low, and even though the
incremental cost is large, the incremental pounds collected are
also large. Therefore, the cost advantage of larger plant size
and less restrictive standard is not overshadowed as it is in the
other cases.
The fabric filter options show lower cost-effectiveness
ratios for the more stringent regulatory option because signifi-
cantly more pounds are removed at a small increase in cost. The
cost-effectiveness of the fabric filter systems in regulatory
Option II are based on the same air to cloth ratios as in regu-
latory Option I, but with the treatment of a portion of the gas
stream corresponding to the lower required efficiency. The por-
tion of the gas stream which is bypassed is less than 10 per-
cent. It is likely that this would not be feasible in actual
practice because of the difficulty of maintaining proper flow
proportioning as furnace conditions change. In such cases, it
would be more appropriate to treat the entire gas stream. Con-
sequently, the Option II cost-effectiveness ratios in these
cases are not meaningful.
8-84
-------
LU
UJ
g
m
Q)
H
8
o
o
<->
e i «
j-.ce: H-i a;
LU H-
Zh- Z 00
00 O
h- C
000 O
n
-P
H
(d
-P
rt
O
a:
UJ
oo
8 CD
oe
3 LU
CO
UJ
O 3 I
LU O I
00 _J «t
oo CQ a
LU O
a: oo
r\ *^^
O
col
CM
o
LO
CSJ
-a
C3
O
CM
O
O
O
vr>
o
o
o
ur>
oo
o
o
0) CO
'd M
o v
e -P
to o
CO -H
0) M
fi -Q
iw
H
-P tr>
o s
Q) -H
M-l CQ
IW 3
0)
-P
CO
O
i
oo
CU
Cn
H
00
CM
*SS3N3AIlD3JJ3-iS03
8-85
-------
§
Of
t-
I
oo o
h- II I I I CO
H-o: H- co
5.1 Q.UJ
oco o '
D
W
Q. «-*
§
CO-JCOOO.
n
a. ca
iir
CM
«[
CM
Lf>
vo «r
CM rI
r-?
ov
CM
o
\O CO
O
O
o
tr>
CM
o
o
CM
O
O
o
Lf>
o
o
o
in
UJ
N
CO
-P
C
o w
o u
vS
S3
P, PI
H
rH O
0) 0)
T3 M
O CU
so
HI a«J
0-P
0) -P
en IQ
0) O
C M
0) -P
> o
H 0)
-PH
o o>
0)
M-4 C
a) -H
4J
en
8
00
0)
CO
CM
CO
CM
'SS3N3ALL33JJ3-1SOD
8-86
-------
In the wool fiberglass segment using a fabric filter, the
SIP baseline costs are also based on the concept of treating a
portion of the stream. For the same reason stated above, the
incremental cost of the regulatory options over and above the
baseline is not meaningful.
8-87
-------
REFERENCES
1. Kinkley, M. L., and R. B. Nevril, Capital and Operating
Costs of Selected Air Pollution Control Systems, EPA 45O/
3-76-014.
2 Private communication between Atul Kothari of PEDCo Envir-
onmental, Inc., Cincinnati, and Peter Czukra of FMC Corpora-
tion, Chicago.
3. Perry, R. H., and C. H. Clilton, Chemical Engineers Hand-
book, 5th Edition, New York: McGraw-Hill Co., 1973, pp.
25-16.
4. Robert Snow Means Company, Inc. Building Construction Cost
Data, 1977. Duxbury, Massachusetts, 1977.
5. Peter, M. S., Plant-design and Economics for Chemical Engi-
neers, Chap. 4. New York: McGraw-Hill Co., 1968.
6. Vibrant, C. F., and C. E. Dryden, Chemical Engineering Plant
Design, Chap. 6. New York: McGraw-Hill Co., 1959.
7. Richardson Engineering Services, Inc. Process Plant Construc-
tion Estimating Standards, Vol. 1-3. P.O. Box Y, Solana
Beach, Calif. 1978. "~
8. The Mcllvaine Company. The Mcllvaine Scrubber Manual,
Chap. IX, Northbrook, 111. 1974.
9. Statistical Abstracts of the United States, 1977. U.S.
Department of Commerce, Bureau of the Census, 1978.
10. Teller, A. J., "Control of Glass Furnace Emissions," The
Glass Industry, February 1976. pp. 15-20.
11. Arthur D. Little Inc., Environmental Considerations of Se-
lected Energy-Conserving Manufacturing Process Options:
Volume II, Glass Industry Report, December 1976. p. 33.
8-88
-------
8.3 OTHER ENVIRONMENTAL COST CONSIDERATIONS
Water pollution control is the only other set of environmental regula-
tions with any potential economic impact upon the glass manufacturing industry.*
Impact in tlie case of new grassroots plants is defined as incremental control
costs over those faced at existing plants. The increment for water pollution
control consists of added costs for moving from best practicable controls,
required of plants in 1977, to more stringent new source performance standards
(NSPS). EPA defines NSPS requirements for the various glass manufacturing
categories as being the same as best available controls with which existing
plants must eventually be in compliance.
Wastewater containing various contaminants is present in all the
glass manufacturing industries herein analyzed. However, the contaminants
differ by industries and there are different wastewater generating processes
in the industries. In general, the processes that generate wastewater are:
etching; abrasive polishing; acid polishing; washing; rinsing; cullet and
reject quenching; fume scrubbers; edge grinding; shear-spraying; cutting-,
and annealing. Considerable amounts of non-contact cooling water are also
used in most of the industries.
Contaminants present in most wastewater include suspended solids,
dissolved solids, oil and oxygen demanding wastes. Flourides are also
present in the wastewaters of TV tube, hand pressed and blown glass, and
*EX1sting_OSHA Standards on lead and arsenic also affect the industry to
a very minwextent. It is not known how the silica and lead standards
presently being proposed, will impact on the industry
8-89
-------
incandescent bulb facilities that employ acid etching and polishing opera-
tions. Lead contaminants are present in the TV tube and hand-made pressed
and blown glass facilities.
Wastewater treatment methods to meet NSPS requirements vary considerably
as a function of pollutant, glass manufacturing industry, and definition of
best practicable controls. In one industry, wool ."fiberglass, BPT already
specifies no discharge, therefore, no increment for NSPS is entailed. For
the other industries incremental control usually consists of either diatoma-
ceous earth or sand filtration. A few segments also require the addition
of gravity oil separators, dissolved air flotation and activated alumina
filtration.
Incremental water pollution control costs to meet NSPS have been ex-
pressed in Table 8-22 as a percent of product price in order to present a
measure of economic impact. Two segments with percentages of 1% or greater
are: incandescent bulbs (1.0%) and hand-made pressed and blown (2.1%). The
size plant for which incremental water pollution control costs for the incan-
descent bulb industry were available was 175 tons per day capacity. Minimum-
size new incandescent bulb plants are expected to be 400 tons per day. Due to
expected economies of scale in a larger size plant, the acutal percentage of
costs to product price is expected to be lower than the 1.0% mentioned above.
The 2.1% figure for hand-made pressed and blown was for a 5 TPD plant as
compared to the 50 TPD, model plant of the air pollution NSPS impact analysis.
The water pollution economic analysis for hand-pressed and blown concluded
that many existing plants of the size studied (5 TPD) would close. Such a
conclusion appears to justify the selection of a 50 TPD plant as the minimum
size new source. At the greater size, the control cost percentage of product
price should be considerably lower than the 2.1% for the 5 TPD plant. The
8-90
-------
Table 8-22
Flat Glass
Wool Fiber
Container
TV Tubes
Optical
Tubing
Flourescent
(INCREMENT OF NSPS OVER BPT'
Model Size
ng Segment piant (TPD)
700
lass 200
500
erglass 100
Blown
"onsumerware 50
100
100
50
ent 17^2)
100
Consumerware 5(2)
50
100
100
nt 100
Incremental Cost
as % of m
Selling Price* '
.5%
o(3)
.4
N.A.
N A (4>
"/d^
N.A.14'
.1
N.A.(4)
1.0
N.A.
2.1
N.A.
N.A.
.2
N.A/4)
(2)
(4)
that control costs and
C°StS f°r d1fferent Stze "ant than modeled
BPT requirements specify no discharge.
"1th1nU»
«
Sources: EPA Development Documents for Proposed Effluent Limitations Guidelines
and New Source Performance Standards for the Insulation Fiberglass,
8-91
-------
discounted cash flow analysis in Section 8.4 for the hand-pressed and blown
category suggests that the lower NSPS incremental water pollution control
costs, when coupled with NSPS air pollution controls, would not deter
investment in new plants.
8.4 ECONOMIC IMPACT ASSESSMENT
8.4.1 Summary
The purpose of this section is to measure the potential economic
effects of imposing NSPS-controls on new primary glass plants. As used
in the context of this document, the term "potential economic effects",
refers to the extent to which new grassroots plant construction would be
impeded in the industry if NSPS controls were mandatory.
A model plant approach was used to test the hypothesis that the
addition of NSPS controls would not affect the rate of return on assets
sufficiently to deter new grassroots plant construction in the glass
industry.
Annualized control costs per ton of glass produced were calculated
for 19 model plants, representing distinguishable segments of the glass
industry. These costs were then divided by the average selling price of
the industry's product to derive the unit price increase that might have
occurred in that industry's product had the NSPS annualized control costs
been passed through to the consumer.
This price increase was used as the criterion in a discriminant
analysis procedure that separated the industry model plants into two cate-
gories. For each plant in the first category, the relative change in
anticipated return on assets was computed by multiplying the estimated
8-92
-------
change by the plant's sales-to-assets ratio. If the relative change was no
more than 10% of the anticipated rate of return on assets, and the absolute
rate of return on assets was greater than 10% after the relative change was con-
sidered, the imposition of NSPS controls was considered to have no adverse
effect on the new plant construction decision.
Plants in the second category were subjected to an additional screen-
ing process, which resulted in some plants being eliminated as non-represen-
tative of typical new plants, sources, other plants being reclassified into
the first category, and the remaining plants being subjected to a discounted
cash flow analysis, the results of which was used to substantiate or disprove
the economic feasibility of constructing new grassroots plants in which NSPS
controls were mandatory.
The findings of this study are that the imposition of NSPS controls
will not adversely affect-new plant construction in the glass industry.
8.4.2 Methodology and Application
As an instrument for measuring the potential economic effects of
imposing NSPS controls on new primary glass plants, plant models were devel-
oped for the following industries: .flat glass, container glass, wool fiber-
glass, textile fiberglass, machine-made consumer ware, hand-made consumer
ware, TV envelope"tubes, incandescent bulb blanks, optical glass and tubing.
These models were developed by (1) accepting certain production parameters
from Section 8.2 and (2) by assigning to the respective plant model other
pertinent characteristics considered representative of grassroots new plants
in that industry.
The production parameters accepted from Section 8.2 were production
8-93
-------
Table 8-23
DATA SOURCES FOR MODEL-PLANT CHARACTERISTICS
Plant Characteristics.
Plant Investment
Average Product Price
Production Capability of Typical
New Plant
Profit Rates
Working Capital
Debt to Equity Rates
Anticipated Rate of Return on
Assets
Effective Product Yield
Data Source
Industry Representatives
Glass Packaging Institute
Department of Commerce
Industry Representatives
Glass Packaging Institute
Industry Representatives
Glass Packaging Institute
Annual Reports
Almanac of Business and Industrial
Financial Ratios
Industry Representatives
Glass Packaging Institute
Annual Reports
Industry Representatives
Glass Packaging Institute
Industry Representatives
Glass Packaging Institute
"Energy, Efficiency Improvement Target
for SIC 32: Stone, Clay, and Glass
Products", Federal Energy Administration,
June, 1976
Industry Representatives
Glass Packaging Institute
8-94
-------
capability and the number of operating days per year. Other pertinent
characteristics, such as working capital and average product price, were
derived from the data sources outlined in Table 8-23. In some cases,
industry models were developed for more than one production capability
and/or more than one type of glass.
Several important assumptions were an integral part of the plant
model for each industry:
1.
2.
The plant would include SIP controls as a matter of course;
Each plant so equipped with SIP controls would be considered
economically feasible; that is, it would generate sufficient
internal cash flow to justify the initial capital investment
and achieve a minimum acceptable rate of return;
External financing for the incremental capital needed to comply
with the proposed NSPS standard would be available?
No monetary resources other than those generated by the new
facility could be used to meet NSPS annualized control costs;* and,
NSPS control costs would not be passed on to the consumer.
Model plants were used to test the hypothesis that the addition of
NSPS control costs would not affect the rate of return on assets sufficiently
to deter new plant construction.
NSPS control costs from Section 8.2 were examined, and the most costly
control option for each industry was selected in order to measure the maximum
*In the context of this analysis, NSPS control costs are defined as only the
incremental costs above SIP control costs.
8-95
-------
£
CM
&
O
f
CO
r~*"
mm
x
g
i
o.
g
_1
V)
1
t/*
*5
CO
1
Qi
0
U_
to
ll
^t
El
t""
u.
t/
0
c
oc
1-
c
c_
LL
<-
LL
U.
to
* -oIL
* CO 0^
N 0
r T- to
tO r t -
4-> tO O O
O 3 S- CD
C C «
"C O f <
*-*'
0)
*U'1&
to o Q to
i- »-> 0
O. en O CD
to o s- o
O
C r-H
O
C>
S
r-
tt 5?
O
g_l
*~!
^^
°. 4ja-
+i^ C
O -r- tO
3 J3r r-
"O IO O. 03
O Q. T3
s- to1*- o
O. CJ O S
>
4J
(/
C
>
f~* t* ^j" to ^o 10 ^o co i**» ify ^o to ^o **p >|
S-
,
Lnr^«d-^.oir-o>f^cor^e^r>.cr>r^c
vor ioi*~r^>fi*-t'^f^'i'^'*^f^r>'f*"f>~'-'
-te-r
&. S- i-
at a> a>
* % | * ^^
i «p* «|»
U. U.U.
o. 00.0.0.0.0. o 00.0.0.0.0.0
tO-r-tOtOtOtOtOT-'j-tOtOtOOOCOt
LU S- LU Ul Ltl UJ taJ S- i-LULULULULUL
jQ ?** £}
to to to
u_ tuu.
*
oooooooooooooo*
C3LOOCDtOOtOOlOOlf>OlOOl
** Q)
O) i
4-> fO
^"^ tO 25
a> o
t/i S **"" ^
J S I/I^<'-JD' 5i
1 £ 01 r- _l T- -r- S
MiAr- -0=3 3WI 3
S «g p ,ECQ^OQ 2 w £
i .-SS -Po+Jocn o
CD ^31.0 ai c: oo c co to «J
«/» i- miZ §"o^~'o*~'ej 0)
IO Q) jX r- l/» I/I «/» l/l "O
i c rr ai T3CT3ClO E
IO !- CCtOCtOO 1
+J +J i +J UJlOl BJ c -r- "O
IOCOX OCQOCQ4J C
f o o a> 2>c c a. to
u_cj3i i *-< *-i o re
8-96
»
^^
o
r*
T3
(U
CO
S
x. a. a. a. a. cu
o to to to co 3E
14 LU UJ LU LU
43
§
c
o
g-
i
o
(O
0
o
z>o o oo +J
r> o LO CD tr> 1/1
O)
1^
_g?
O)
V-
««
O)
0}
to "g"
r 4J O
to to s-
s. 42 %
S §- , to
3 en *=
w «=
c? ca a. o
^^ O «r-
ai -P
C O) CJ tO
El *
0
o
o
S-
4J
C
8
o.
|«4
to
o
o
to
43
to
o
o
1
OJ
1
b
c
*r~
01
10
-o
»
r~
at
to
to
r-
5
o
i \
X
OJ
c:
o
0
CD
f
C
r-
«*
S
O
O
o
43
0)
I/)
.e
o
?
*
-------
possible change in the rate of return on assets. These options and the
corresponding capital and annualized costs are shown for each industry in
Table 8-24.
A discriminant analysis procedure was developed and applied as the
next step in discerning whether the change in the rate of return on assets
appeared to be of sufficient magnitude to adversely influence new plant con-
struction decisions. The criterion used in the analysis procedure was the
magnitude of the profitability on sales change produced from the absorption
of NSPS annualized control costs. (This would be referred to as potential
product price increase if costs would be passed through to the consumer.)
This profitability change was derived by dividing the estimated control costs
per ton of glass produced* by the average selling price of the industry's
product.**
The results of the discriminant analysis are shown in Table 8-25. As
can be seen, two industry categories were derived: Category I, containing
those industries for which the estimated profitability change was 1% or less;
the second , containing those for which the estimated profitability change
was more than 1%.
For industries in Category I, the relative change in anticipated
return on assets was computed by multiplying the estimated price change
by the plant's sales-to-asset ratio. If the relative change was no more
than 10% of the anticipated rate of return, and the absolute rate of return
*Estimated control cost per ton = Total Annualized Control Costs
. No. of TPY at capacity, adjusted for
effective yield
**Average selling price, as used here, represents the net sales value per
ton, f.o.b. plant, and excludes discounts and allowances.
8-97
-------
Table 8-25
DISCRIMINANT ANALYSIS RESULTS
Category I
Industry
Model Plant
Capabi1i ty
(TPD)
Control Cost/Ton
Price/Ton
Flat Glass
Wool Fiberglass
Textile Fiberglass
Machine-made Consumer Ware
TV Envelope Tubes
Tubing
Optical Glass
Hand-made Consumer Ware
700
200
100
100
. 50
100
100
100
,50
100
.3%
.7%
.3%
.6%
.7%
.8%
.IX
.2%
Category II
Industry
Model Plant
Capabi1i ty
(TPD)
Control Cost/Ton
Price/Ton
Container Glass
Textile Fiberglass
Incandescent Bulb Blanks
(Soda-Lime)
Incandescent Bulb Blanks
(Borasilicate)
Tubing
Hand-made Consumer Ware
TV Envelope Tubes
250
50
100
50
100
50
50
50
50
1.
1.
1.
3.
1.
2.
1.
1.
4%
5%
0%
1%
2%
5%
5%
1.4%
8-98
-------
was greater than 10% after the relative change was considered, the imposition of
NSPS control costs was considered to have no adverse effect on the new plant
construction decision.
Industries in Category II were subjected to an additional screening
process. Production capability of the "typical" new grassroots plant for
each industry was compared with that of the industry's model plant found in
Category II. If the production capability of .the typical plant was equal to
that of the model plant (as was true in the case of the hand-made consumer
ware 50 TPD facility), the calculated profitability change for the model
plant was accepted. If the production capability of the typical plant was
greater than that of the model plant, Category I was re-examined'for larger
plants of the same industry. A larger plant was listed in the cases of the
textile fiberglass 50 TPD facility, the tubing 50 TPD facility, and the TV
envelope tube 50 TPD facility; accordingly, the Cateogry II model was elim-
inated for these industries and further analysis of that industry was limited
to the model found in Category I. If no larger plant was listed in Category
I, the production parameter of the highest volume model plant for that indus-
try in Category II was adjusted, and the appropriate control costs were
scaled to permit the recalculation of the profitability change. This was
true in the case of the container glass 250 TPD facility, the incandescent
soda-lime bulb blank 100 TPD facility, and the incandescent borasilicate
bulb blank 100 TPD facility. .The lower-volume model plants for the incan-
descent bulb blank industry in Category II were eliminated from further
consideration. (See Table 8-26).
If the recalculated profitability change shown in Table 8-26 was 1% or
less (as was true of the incandescent soda-lime bulb blank-100 TPD facility
and the incandescent borasilicate bulb blank 100 TPD facility), the model
8-99
-------
o to
a> to
fJ
o
o
o
t
in
l-H
in
LU
OS
£_>
oc.
XX
CM C_>
I _)
OD g
ai ua
to
o
o
o
o
a
LU
S
CO
tO
+J
to
0
T3 O
O)
a
r»
O
S-
4*^
c*
O
O
a
CD
o
o
CM
fee*
o
o
o
in
o
o
o
A
o
CM
O
o
o
VO
CM
fee-
O
O
O
o
CO
fee-
o,
o
ro
r
cj ra co
O
O
in
o
o
o
o
o
in
CM
o
o
CD
O
o
u
Q.
OO
cu
o
ta
v>
o
o
c
Q)
(U
10
to
Q.
ai
to
a>
a»
o
O
10
-------
r^
CM
i
CO
cu
jQ
i"3
H-
i- 0
C ZJ to S_
OI CU C
CU OS 01 O
en m o
ro o 5-
-C CO)
O CU CD 4->
re r
Ol- O
to s-
t- to +J
cr
c
1
c:
CO
D.
CO
2:
u_
CD
^
a
i
CO
o
a.
1 H
I i |
3=
1
LU
1 1
P.
LU
CO
CO
e^
2-
O
2:
LU
a:
u_
o
LU
o**
2:
1 1
CO
LU
CD
1
0
o< cu c
$- 0
CU C O CJ
»-> O f-
ro . cu co
Di c: ca D_-
- s_ co
3 2:
-4-J
cu
a;
£>
r
r- CU
jQ CD
re c
i -C
If O
o
J_
'Q-
^^
0
CM
1
in
y |
Ij-
.O
O 4J
1- re
i o;
to i
cu to
i -t->
co to
00
«t
CM
-X
to
OJ
to
01
^- c
o ro
cu a.
rs
c: r
CO CU
> "a
cu o
a; s:
m
^ ^
*/>
to
§ ^
1^ CM
r-"
s:
a
D_
1
0
0
to
to
ro
r
CD
ro
1 1
^-s
CO
"
^_
o
- CM
1
in
fi
CO
CM
CO
to
LO
CM
CM
LO
O
a.
0 '
0
CM
*
to
to
ro
en
cu
.a
r-
Lu
O
O
~"^
~^9 ^-£>
r^^ in in
.
^p o
O LO LO
CM ,-H t-H
1 1 1
LO CM CM
r~H t _j . i i
0
r^ co to
o
UD OO
* 1 * . t 1
^ oo co
VO CD CD
co m in
CM i 1 to
r^ t-H o
CO CO *
o
Q.
H- a, cu
CD re re
OS 3:
i t
* s_ t
cu cu
01 3 §
01 to to
ro c c
r O O
CO CJ CJ
- CU CU CU
-Q T3 -a
i ro * ro
U- E a E'-^
1 Q_ 1 £3
CU CU 1 CU 0.
i- -t- CD -r-
4-> J= O -CO
X CJ t 1 O LO
^-* ^~
CM
rH
t I
1
CM
t~H
oo
LO
t^H
CO ,
CD
CM
O
CD
CO
CU
1
i.
O)
13
Ol
c
o
CJ
CU Q
"^3 ^
ro 1
1 0
TD o
c: i <
ro ^
zzz
£r>
LO
t 1
1
CM
rf
K
. ^
«
CO
o
LO
cn
CO
CO
^ ^
Q-
O
CD
t 1
CO
CU
23
1
'V Envelope
i
co
ti
i
CM
t-H
&9
CO
LO
CO
CD
in
H/
cr>
CM
C3
D.
1
CD
CD
t t
cn
c:
i
-Q
1
CO
in
t-H
CM
t i
CM
to
t 1
OO
CD
in
CO
CD
CO
S"
Q_
'
o
LO
Ol
Ol
ro
CD
re
o
Q.
CD
cn
LO
CM
r~~
K
CO
.1
CTl
CD
LO
"=f
s
cu
1
re
-o
o
CO
ncandescent
Bulb Blanks
t 1
LO
LO
1
CM
i
CO
r~.
i
CO
0
LO
to
CO
cu
re
o
"co
re
i.
o
03
ncandescent
Bulb Blanks
t t
co to
II II
CM sf
<1- CO
re
Q-
ro
u
CO
a.
co
ex
re
o
O)
cu
to
to
re
o
co
X
to
a>
o
c
8-im
-------
plant was reclassifled into Category I and the relative change in the rate of
return on assets was computed as previously described. Table 8-27 shows the
changes in rate of return on assets for the model plants falling into Category
I, before and after the second screening. If the recalculated price increase
shown in Table 8-26 remained greater than 1% (as was true in the case of the
container glass 500 TPO facility), the plant remained in Cateogry II.
A discounted cash flow analysis (DCF) was then performed for each model
plant found in Category II (the container glass 500 TPD facility and the hand-
made consumer ware 50 TPD facility). The results of these analyses were used
to substantiate or disprove the economic feasibility of grassroots new plants
in which NSPS controls had been installed. These results are shown balow in
Tables 8-29, 8-30, and 8-31. Meanwhile, Table 8-28 of this section explains
the discounted cash flow methodology used in those tables.
8.4.3 Findings
Table 8-27 shows the changes in the rate of return on assets for
the model plants falling into Category I, before and after the second
screening. As can be seen from this Table, the percentage change in the
required rate of return for model plants in all industries fell within
an acceptable range, i.e., was less than 10% of the rate of return on
assets. It was, therefore, concluded that the imposition of NSPS controls
would have no adverse effects on the new plant construction decision for
the Category I model plants.
Tables 8-29, 8-30, and 8-31 show the discounted cash flows for the
model plants that remained in Category II after the screening process was
completed (two tables for the container glass 500 TPD facility and one
8-102
-------
Table 8-28
DISCOUNTED CASH FLOW METHODOLOGY
Industry Assumptions
Data Requirements and Computational Steps
Container
Hand-made
Consumer Ware
$255 $5500
15% 13.1%
Pollution Control Incremen-
tal Costs
Profit Before Taxes and
After Pollution Control
Taxes
Profit
Investment Tax Credit (ITC)
Depreciation (years)
Equip. 15 Equip. 15
Bldg. 40 Bldg. : 33
Interest
, Debt Repayments
15 years 15 years
85% debt 50% debt
Net Cash Flow
Discount Rate
15% & 8% 15%
Discounted Cash Flow
Initial Investment
Present Value of Discounted
Cash Flow
Revenue = Average Selling Price per Ton
Profit Before Tax and Before P.C. = Revenue x
Profit Rate Before Taxes
Annual i zed Co'ntrol Costs -r (Yearly Tonnage
Capacity x Yield)
Profit Before Taxes and P.C. minus P.C. Incremen-
tal Costs
48% of profit before taxes and after P.C.
Profit before taxes and after P.C. minus :taxes
10% x Equipment Cost (incl. P.C.) -r (Yearly
Tonnage x Yield) (amount allocated in 1st and
successive years no greater than tax, owed)
(Equip, or Bldg. Value T Number Depreciable
Years) * (Yearly Tonnage x Yield) . .'
10% x Tax Rate (48%) x % Investment Funded by Debt x
Total Investment
Yearly Tonnage x Yield
% Investment Funded by Debt x Total Investment *
15 years
Yearly Tonnage x Yield
Profit + ITC + Depreciation + Interest minus Debt
Repayments
Minimum Acceptable Rate of Return on Equity or on
Assets
Net Cash Flow x Discount Rate
Total Investment = (Annual Tonnage & Yield) x %
Equity or Debt Financed (Total Investment =
Equip., Land, Bldgs., P.C. + working capital)
Sum of Discounted Cash Flows
8-103
-------
en
en
r-~ oo
r-«
o co CM
in r-« T-*
r*. I-H co
CM
co
O CO CO
CU
co
VO
co in oo
CO
o in co
CO rH 1^.
CU
co -t~t r*.
CM
o en «g-
co in co
o
cu
CM
CO
cu
co
en
co en r»-
r»«
C3 CO O
co vo co
t^- CO
cu
cu
CO
en
co co CM
en
o en 3-
co r^ co
co
t»-
CTl
VO
CO
co
in en
co co
vo
co in
o
en
CM
co
o
cu
ICO
t-t
QJ-
f-O
cu
c^-* cu
Of i.
o ta o
»-> M-
D)
73 O
V- I- CU O
3 E
cr x >> u-
Q) -I- 03
> Q CU
o m
i- 3 I
cr "« 3
faO-
O CO Cu O CU CQ
Q) i- S- 3 S- !- S-
O U- r
CU Oi (C
o O 4->
i- CU J=
o o
Q. Q. +->
(U QJ C
QJ QJ -r-
cn t/> -r cu to
8-104
-------
o
CU
3 O
O 3
_j -a
0
3= 0-
=C co
O CO
rO
Q r
LU C£J
ra o
o
C_J C
GO O
. | . I ^
Q
S-
'~? CD
d. °-
I S
0 $ 0
cn if; -
1 ;"~* r~>
00 £ Eb
CU ^ ^
i|l
°P
r-i
i i
s:
GO
GO
=c
CD
DC
LU
Z
1 i
"=C
2:
O
o
CM
cn
i i
I-H
cn
cn - .
r~i
o
cn
ti
.
.
cn
CO
cn
i i
CM
CO
cn
i i
rl
CO
i 1
.
O
CO
cn
r-H
s
CM
cn LO cn <3- cn r-^ CM <3- t-^ r-i ^i-
C£1CM CO cn =!- to CO ,_,,_,,_, tf,
r-Ht^-Cn cn r-H r- 1 ,_| ,-H <-<
op LO cn co Lotocn^j-Loooor-i
2 £i 23 °° ^"^r^Lo^HcncMt-,
rt*^- Cn oOi-Ht-lT-t ^-ir-,,-.
CJ
Q_
e3 08 o
CO « ^ CO .
CU ** CU
x a r- x 4J
to O ro c
"-«£> h>- 5 co
a,^. ^ . E X 4J
?* 5 ?* to ex ra c
h ^9 ^ ^ cn -r- i cu
O O ra O 13 Z3 £Z
§ s2 .§1 S^i s ^ ^ |
^-^Cu'-Pcl^0- 4J ' ' 3 S
01 - J- 3 S- T- SL. c/1 -t- . . S!
o<+-oi ot-cucuq- s-i-cu+->
J £% o5 ££ ig S £ & & £ -S
a. o-ca a. Q. <; j Q. I-H a a .-, Q
o r^.
f~I 2
CM r-)
O r-H
CM
r~ cn
CM r-H
O LO
cn
CM S
*
*
CD ID
co
CM ^J-
*
O VO
ai
CM LO
*
O r-H O
CM
r~~ cn to
CM LO rH
-k&
O CTI UD
r^ co co*
CM IJD i 1
O cn to
r-« cn t-i
CM r-^ CM
- ««
o to cn co co
cn CM cyi o o
«* cn cn .^j- ^j-
-fa* ±f* -to-
o
CU
c: c:
CU 3
E O
' ^ 4-1 O
CO CO
> 5
t-i 4-
^^ S O
LO 0 >>
S r-H i -»-? O)
0 U_ T- . 3
U. CO JC CT "irO S
+-> to LU :> o
-C ra ra i
CO QC O i 4-> U_
O . . -i- CO CO
-r° to to -r- CU ra
CO T~ -i- C S- O
Z O Q l < Q-
[ <
c
ro
ex
CO
c:
f-
o
-a
r~*
(U
r-
«J
O)
>^
*t~^
to
t 1
O)
o
c
CU
S-
OJ
!j:
T^3
O)
JC
to
CO
cn
t i
0
CU S-
. 3
r- CO
-!-> S-
co ra
c co
' >>
D1 O)
r- -i
cn -a
ra O)
-i<£ CO
o o
(O O
CL. 3
to
CO
co -o
ra c:
ra
co ,-cn
-c r^
-»-> cn
t i
r*> c~
o >r~
CO CU
t- O
c.
ex co
. ex i.
3 CU
to f-
-------
Table 8-31
DISCOUNTED CASH FLOW
Price/Ton
After P.C.
(500 TPI
(
1978
n $255
sfore Taxes &
P.C. (15%) 38.3
n Control
ental) 3.8
efore Taxes &
.C. 34.5
16.6
17.9
15.4
Bldgs. 1.5
Equipment, incl. P.C. 13.0
: x Tax 12.3
layments 17 . 1
1p*i *t A ^ n
Flow $4-3.0
A- t r\at n "i* \ OC 1 f\
ite (8% A.T.) .9610
-_L- r-i-.... 1 $20.4 $18.8
Taxes
Profit
ITC
Depr.
Depr.
Intere
Debt R
Net Ca
Disc.
Disc.
Initial Investment
Present Value of Discounted
Cash Flow (20 years)
(1)
(3)
$265.8
$303.2
'As supplied by the Glass Packaging Institute.
(^Differences from 1979 and beyond arise because of lower yield at plant start-up.
^Values for year 20 include $29.I/ton working capital recovery, $8.9/ton land
value and $20.8/ton undepreciated bldg. value.
8-106
-------
table for the hand-made consumerware 50 TPO facility). The DCF analysis for
the hand-made consumerware industry utilizes the rate of return on equity as
the discount rate rather than the rate of return on assets. When such a rate
is used (which is usually higher than the return on assets) the net present
value is compared to the equity portion of the investment rather than being
compared to the entire investment, which would be the case when discounting with the
return on assets. The choice of the equity rate of return was made since
ownership of small plants in this industry is typically private, and such a
rate is considered their minimum acceptable return. The sum of the discounted
cash flows exceeds the equity investment, and it is, therefore, concluded that
NSPS-controls would not adversely affect new plant construction decisions in
the hand-made consumerware industry.
The same conclusion is reached for the container industry. However, the
data does not'present as clear a picture as it did above. DCF calculations
were performed using both,; a rate of return on assets as compared to total
asset investment, and a rate of return on equity as compared to the equity portion of
the investment. The latter calculation clearly shows present value exceeding
the equity investment utilizing a 15% after-tax return, and therefore not
adversely affecting new plant construction. A significant factor in that
outcome 'is the low equity investment of 15% specified by the industry and,
therefore, a large amount of debt leveraging.
When an 8% minimum acceptable return on assets is used as the discount
rate with the data supplied by the Glass Packaging Institute (GPI), the present
value also exceeds that of the initial investment. GPI did not supply a before-
tax profit rate on sales and the results of the DCF are quite sensitive to that
rate. A rate of 15% was calculated based on the GPI minimum acceptable return
on assets, and the sales to assets ratio for a new plant. The 15% profit rate
8-107
-------
before taxes is high as>compared to the present rate on existing plants of about
B%. If an 8% before-tax rate had been used, the present value would be less
than the initial investment. However, the 8% minimum return on assets rate
supplied by 6PI must convert to a 15% before tax profit rate in order for the
8% to be achieved, given the sales to assets ratio of the 6PI data. A before-
tax profit rate that is higher than the rate which presently exists in the
industry suggests that new plants are more capital intensive and provide
improved earnings.
If such a rate is not achievable, it suggests that incentive is not
present for new container plants of this size to be built, regardless of NSPS
controls. The'change in return on assets from NSPS controls is 13% at the
15%-before-tax profit rate, and 15% at the 8%-before-tax profit rate.
8.4.4 Capital Availability
In conducting the profitability change analysis and the discounted
cash flow analysis used to test the economic feasibility of imposing NSPS
controls on new primary glass manufacturing plants, the assumption was made
that capital would be available. This section turns to the examination of
that assumption. Capital availability assessment is defined here as, the
sufficiency of cash flow to permit the repayment of debt.
Model plants in Category I were estimated to be impacted by NSPS
controls by a change in return on assets of less than 10%. Since such a
change is relatively small and since the cash flow from profits is not the
only cash flow available to service debt obligations, it is concluded that
financing could be obtained by those plants for NSPS.
For plants that remained in Category II after screening, the increase
8-108 .
-------
in capital required to install NSPS controls for the two model size plants
subjected to a DCF analysis are:
50.TPD
500 TPO
Hand-Made Consumerware
Container Glass
5.7%
10.0%
Since the DCF analyses included provisions for all debt repayment
and since the cash flows in the analyses exceeded the initial investment
at the minimum acceptable rate, it is also concluded that sufficient cash
flows exist in those plants for the NSPS capital-related costs.
8.4.5 Industry Costs .-
Table 8-32 presents the maximum number of typical new plants which
will be constructed in each of the industry segments during the five-year
period from 1978 to 1982. These projections are based on^current forecasts
of industry growth which have been applied to 1977 shipment statistics and
the results converted to new plant estimates based on typical new plant
production parameters. A methodology, described in 8.1.6.7, was used to
project demand in the .pressed and Blown segment. These estimates assume (1)
no excess practical capacity existed in any of the industry.segments in
1977, and (2) that all industry growth requirements will be satisfied
through the construction of grassroots new plants. To the extent that
1977 shipments in any segment were less than maximum practical capacity,
the number of new sources will have been overstated, and the industry
cost estimates inflated.
8-109
-------
o
co
m
oo
in
CM
CM
*
a\
CM
in
*
(.O
* CM CM
o in 15
in «* *
<8
Q.
IB
O
CM
CM
CM
O
to
o
CFl
o
O
O
Q.
-
O
U
U
(O
CM
ob
U
O)
U.
UJ
in
o
o
CM
o o
o o
r CO
o
o
CM
o
in
o
in
u
(U
CL
(U
M
O
4J
(A
(/)
(O
S3
cs
"»
vt s-
IB CD
5 -
+> 5
fO £t
r O
^_t_, tJ
^ A
S- !-
U-
o
O
(U
O
O)
T3
IB
a>
u
IB
in
18
CD
.a
3
h-
a>
c
UJ
IB
CO
=3
CO
«/)
£
a.
IB
a
(U
a +»
u c
a> a.
ra
o o
O«r-
Q.
(U >>
w »->
a>
^= ai
8-110
-------
Table 8-32 presents the maximum total annualized control costs that
will have been incurred by each industry by 1983 if NSPS controls are
installed on all grassroots new plants. These totals were derived by
multiplying the estimated number of new sources for each industry by the
appropriate annualized control cost estimate. As Table 8-32 shows, the
container glass industry will experience greater costs than any other
industry segment.
Neither the total annualized control costs for all industry segments
nor the total control costs per to.n as a percentage of selling price are
considered a major economic consequence as specified by the regulatory
analysis requirement of Executive Order 12044.
8-111
-------
-------
APPENDIX A
EVOLUTION OF STANDARDS
A-l
-------
APPENDIX A
EVOLUTION OF STANDARDS
Date
06/77
07/77
08/77
09/77
10.77
12/77
01/78
02/78
03/78
Conducted a literature search to acquire reference
materials
Contacted EPA regional offices and 26 state agencies
Contacted various control equipment vendors
Visited New Jersey Bureau of Air Pollution Control, EPA
Region II and III offices and SCAQMD
Contacted industry representatives to update plant data
Meeting with EPA and ad-hoc committee of 6PI
PES visited five glass manufacturing facilities in the
western U.S.
PES and the task manager together, visited four glass
manufacturing facilities in the eastern and central U.S.
Commenced the drafting of the SSEIS introductory chapter
Brinks mist eliminator pilot study received and reviewed
Received test reports from New Jersey
PES visited one glass manufacturing facility
Sent out eight, 114 Questionnaires
PES witnessed an EPA Method 5 test
Progress meeting with EPA
Received final 114 Questionnaire response
Working Group meeting to discuss findings of SSEIS and the
recommended standard
Commenced work on second draft of SSEIS
NAPCTAC package, consisting of the second draft of the
SSEIS was submitted
A-2
-------
04/78 NAPCTAC meeting held to review the recommended standard
04/78 to
01/79
07/78
11/78
12/78
01/79
02/79
SSEIS was reviewed and edited and the Preamble and Regula-
tion were written for submission to the Steering Committee
Meeting with PPG representatives
Meeting with the ad-hoc committee of GPI to discuss the
SSEIS and Regulation
Meeting with EPA in Durham to discuss progress of SSEIS
Steering Committee package submitted for comments on
the proposed standard
Meeting with GPI representative to discuss the SSEIS
as it went to the Steering Committee
03/79 Package submitted for AA Concurrence
A-3
-------
-------
APPENDIX B
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
B-l
-------
I p
o 10
4J -I-
E E
a>
10
O -P
O. E
CO
o -P
s- to
ta
a -p
E
(O
p
CO
I
a>
P R
P E
r- O
3: s-
T~
§E
U
i co
s- tn
ta
to
a>
O)
p
u
cu
«o
cu
>>00
i O
0 -P
S- . ro U
O 1-0)
H- (/)
cu
.J2 4-> CO
(O
cu
cu
to
cu
i ^- O OJ
O) O
M- !- O T-
s- o
o. cu
to
cu
.E "
4J CO
O CU
u
CO «r-
o -P «
E a.
o *-
O CO
3 CM
p
« S-
P CU
P U
to E
(U CU
-a > -a_>
«*
ea
CO CU
j= -P
p a
to
c h. oo
o «o
P O
c.
i- E
in _
a; ta
a> ~o
s_ c
(O
P CU
£5
X
CU >;
C\J
ro
O "O O
O) (U
o o
r- CU
10
a;
3 0) -P
i- .a o,
-o re
o J=
CC -P U
re
«»- s-
o a>
s- -o
UJ CC
(O
+-> co
E O
E U
CO CO
0)
D. i
4J ro o co
O -E E ,
CO O M S-
(J_ CO
q- E -P P
QJ -I- O C
E re
"re 13 M-"o
4-9 10 O
E to E
CO 3 t/)'i
C >
to
O -r- O CO
u s-
0)
CU T3
UJ
UJ
cu
to
to
UJ
CDr'
E re<
££:
ro CU I
Q. E CO
CU E
S- O Od
0. S-U.
r
O E CO
If- UJ'
trt E to
CU O -P
.^43
i O
CU eC
i- >, re
E 3
O !-
s- to
D. re
ca
cu
CO
ro
CO
ro
P
CO
O
ro
CU
P
O
CU
re
-p
co
(O
(f- (O
O
cn
cu
JD
re
cu
r- 10 CU
cu
a>
to
cu
re
cu
u
r- O
E re
p HI
ro
p T-
ro
r- >
X CU
UJ _1
CNJ
B-2
-------
CU
25
O
O
CO
o
I1
s
LU
O
CL E
Q. CU
3 E
co a
to re
s-
O E
r- LU
re -a
O E
o re
cu
CO
E
CU
cu
cu
O Q.
E re
M--5
o
E
O
re -a
Q. cu
E s-
i- CU
-a
CU !-
S~ in
re E
t- o
o
cu
4-5 CO
r -a
re s-
cu re
cu re
P^
1 1
. . ^
CU
E 4-5
O Q.
r- re
Q. 0
o
E
JE *r
4-5
's "S
^
T3 CU
CU "O
4-3 *r-
re to
- E
0 O
o o
to
to cu
re i.
re
CO
4-> CO
o E
re cu
Q.4-5
E to
1 ^>
CO
^~
re r
4-3 O
E S-
CU 4-3
E E
E 0
O O
S.
i- E
> 0
E .-
CU to
CO
CU !-
to .
E
cu
to
to
"o
i.
4-5
E
O
O CM
E 00
O
r- E
to o
CO -r-
*i tj
CU CU
CO
1 1
1 t >
00
o s-
- CU
Q-Q.
o re
JE
«f- 0
O
E
O
re T3
Q. CU
E S-
r- CU
a
in in
O E
0 0
o
cu
c-
o
CO
CU
to _
co CU
[ ^
i CL
o re
S- JE
4-5 O
E
0 E
O -r-
£Z CJ
o cu
i- S-
to cu
tO "O
'i 'to
CU E
o
l-H O
l-H
to
0 "~
i- CU
4-3 S_
CL re
0<4-
1
O S
I ^ ^5
O E
re re
Q.
E -E
i 4->
fH.
cu re
to
CO
"O
s-
T5
o
E
re
4-5
CO
"O
cu
CO
o .
O-r
o
s~ r*^
CL
CU O
i ^ *\
O
4 QJ
O CO
4-5 «
u r-.
re
Q, S-
E Q)
O.
E re
O JE
r- U
4-3
3 E
r «i
o -o
CL CU
i- CU
i "O
re !-
CO
CU E
;F °
» o
o
_£
3
^-*
^
O
CL
S-
O)
4-3
re
5
E CM
O
f-
to
"O E
S- O
re v-
E O
re cu
4-5 CO
CO
"C I*s.
cu
to i-
o cu
Q.4-5
p CL
s- re
Q-jE
0
cu
JE E
4-5 -r
<*- T3
O CU
S-
4-5 CU
o -o
re -i-
^E C
- 0
o
cu
-E CO
1 !-
cu
4-5
CO
re co
2
r-.
r- E
i O
O -i-
CO 4-3
E CU
O CO
CO >
"O r*^
i.
re i.
-o cu
E 4-5
re CL
4-5 re
CO .E
o
CU E
CO -r-
o
CL-0
ocu
CL CU
o
CU !-
JE CO
4-5 E
^8
o
CO
4-5 !-
O
re i
o. re
E to
- !- 0
CL
CU to
JE !
h- T3
CO
p
CU
CO
3
^_^
CD
CU
cu
E
O
CO «3"
^ r^^
re
a E
E 0
re T-
co u
cu
"O CO
cu
CO «
o r^
CL
O i.
s- cu
Q.4-5
CL
cu re
4-5 0
t- E
O !-
4-5 -a
0 CU
re s-
CL CU
pjr ii
r- *r
CO
CU E
-E O
1- 0
1
O
o
CO
^
CO
CO
o
o
J-.
o
to
^3
re CM
E CO
re
4-5 E
to o
TJ 4J
cu u
co cu
O CO
CL
O >
k~
cu cu
-E 4-5
4J Q
re
If- f-
o o
4-3 E
U !-
re
Q.-O
E CU
I""" 5ri
CU
CU T3
f^'to
CD
CJ
a.
l-H
1
z
LU
O
a:
1 1 1
X
LU
Q
l-H
COr <-~*
E re CTI
T E 5P
re cu r^.
Q. E co
CU E
s- o a:
O- S- U_
s- > en
O E CO
If. LU
to E CO
CU 0 4-5
E -r- E
i- 4-5 CU
0 E
CU «cC QJ
Agency Quid
Regulatory
Impact Stat
4-5
0
re
» i
cu
re
"cu
o
re
4-5
re
cu
to
4-5
re
a.
l-H
r
re
4-5
E
CU
l-H C
> i O
i- to
E -I 4-5
0 > CO
r- E O
4-5 LU O
O.
O
4-5
0
re
Q.
l-H
CU
1
cu
E
re
-C
4-3
re
cu
E
0
r
U
0
4-5
3
O
Q_
i_
Ol
4-5
_re
re
to
o
O.
to
cu
4-5
to
re
o
oo
cu
LU
to
4-5
to
O
B-3
-------
-------
APPENDIX C
EMISSION MEASUREMENT
C-l
-------
C.I Emission Measurement Methods
During the standard support study for glass manufacturing plants, EPA
conducted particulate emission tests at two facilities, one controlled with
a baghouse, and the other with a scrubber. The tests were run in accord-
ance with EPA Method 5 (40 CFR Part 60 - Appendix A). Method 5 provides
detailed procedures and equipment criteria, and other considerations neces-
sary to obtain accurate and representative particulate emission data.
Visible emission data were taken during the two EPA tests in accordance with
Method 9 (40 CFR Part 60 - Appendix A).
In addition to the Method 5 tests, particulate emission data were also
obtained using proposed Method 17 at the plant controlled with a baghouse.
Method 17, which was proposed in the Federal Register on September 24, 1976
(41 FR 42020), collects particulate samples on a filter located in the stack
and the samples are, therefore, collected at stack temperatures, (in this
case, 180°C). The Method 5 filter is located outside the stack and is main-
tained at a temperature of 120°C. Tbe Method 17 samples obtained by EPA at
a stack temperature of 180°C produced results 35 percent lower than the
Method 5 tests run at a temperature of 120°C. These differences can be
expected from stack emissions containing vapors that will condense between
180 and 120°C and this is considered the reason for the lower in-stack
filter results.
The remaining data base was obtained from reports submitted by State
agencies or glass manufacturing plants referenced in Chapter 4. The emis-
sion results in Reports 24, 32, 41, 43, and 44 were considered to be
representative of Method 5 testing. The emission results in Reports 25 and
31 were obtained by collecting samples with water impingement"followed by
C-2
-------
filtration. The water impingement method should collect more condensable
r
stack gas components and reactive gases than Method 5, therefore, the reported
results are considered to be higher than would be measured by Method 5.
C.2 Monitoring Systems
The opacity monitoring systems that are adequate for other stationary
sources, such as steam generators, covered by performance specifications
contained in Appendix B of 40 CFR 60 Federal Register. Ocotber 6, 1975,
should also be applicable to glass manufacturing plants except where con-
densed moisture is present in the exhaust stream. When wet scrubbers are
used for emission reduction from glass plants, monitoring of opacity is not
applicable and another parameter such as pressure drop may be monitored as
an indicator of emission control.
Equipment and installation costs for visible emissions monitoring are
estimated to be about $18,000 to $20,000 per site. Annual operating costs,
which include the recording and reducing of data, are estimated at about
$8,000 to $9,000 per site. Some savings in operating costs may be achieved
if multiple systems are used at a given facility.
C.3 Performance Test Methods
Consistent with the data base upon which the new source standards have
been established, the recommended performance test method for particulate
matter is Method 5 (Appendix A - 40 CFR 60, Federal Register,
December 23, 1971). In order to perform Method 5, Methods 1 through 4
must also be used. Method 17 (in-stack filter method) is not recommended
as the performance test because of the presence of condensable components
and variable stack gas temperatures. For glass plant emissions, in-stack
filtration does not provide for a consistent definition of particulate
matter and does not allow for the comparison of the various systems of control
(e.g., baghouses and scrubbers).
C-3
-------
Subpart A of 40 CFR 60 requires that affected facilities which are
subject to standards of performance for new stationary sources must be
constructed so that sampling ports, adequate for the required performance
tests, are provided. Platforms, access, and utilities necessary to perform
testing at those ports must also be provided.
Sampling costs for performing a test consisting of three Method 5 runs
»
is estimated to range from $5,000 to $9,000. If In-plant personnel are
used to conduct tests, the costs will be somewhat less.
The recommended performance test method for visible emissions is
Method 9 (Appendix A - 40 CFR 60, Federal Register, November 12, 1974).
C-4
-------
APPENDIX D
PRIMARY GLASS MANUFACTURERS
D-l
-------
APPENDIX D
PRIMARY GLASS MANUFACTURERS
State
Alabama
Arkansas
*
California
Col orado
Connecticut
SIC
Code
3221
3211
3221
3229
3211
3211
3211
3211
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3229
3229
3229
3229
3229
3296*
3296*
3296*
3229*
3221
3229
3221
3229
3229
3229
Manufacturer
Brockway Glass
Fourco Glass
Arkansas Glass Container
Thomas Industries, Inc.
C-E Glass
Guardian Industries
Libby-Owens-Ford
PPG
Ball Corporation
Brockway Glass
Brockway Glass
Gallo Glass
Glass Containers
Glass Containers
Glass Containers
Madera Glass Div. of
Indian Head
Kerr Glass
Latchford Glass
Latchford Glass
Latchford Glass
Owens-Illinois
Owens-Illinois
Owens-Illinois
Thatcher Glass
Arrowhead Puritas Water
Brock Glass
- The Glass Works
Libby Glass Division
of Owens-Illinois
Ray Lite-Glass
Johns-Manville
Johns-Manville
Owens-Corning-Fi berg! ass
Reichold Chemical
Columbine Glass
Pikes Peak Glass
Glass Containers
Innotech
Thermos Division of
Kings - Seeley Thermos
Thermos Division of
Kings - Seeley Thermos
Plant Location
Montgomery
Ft. Smith
Jonesboro
Ft. Smith
Fullerton
Torrance
Lathrop
Fresno
El Monte
Oakland
Pomona
Modesto
Antioch
Hayward
Vernon
Madera
Santa Ana
Los Angeles
San Leandro
Huntington Park
Los Angeles
Oakland
Tracy
Saugus
Gardena
Santa Ana
Huntington Beach
City of Industry
South Gate
Corona
Willows
Santa Clara
Irwindale
Wheatridge
Colorado Springs
Day vi lie
Tr umbel 1
Norwich
_
Taftville
D-2
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
Florida
Georgia
Illinois
Indiana
SIC
Code
3211
3221
3221
3221
3221
3221
3221
3221
3296*
3296*
3296*
3211
3211
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3229
3229
3229
3229
3229
3296*
3221
3221
3221
3221
3221
3221
3221
3221
Manufacturer
Guardian Industries
Anchor Hocking
Owens-Illinois
Industrial Glass
Thatcher Glass
Glass Containers
Midland Glass
Owens-Illinois
Certain-Teed Products
Johns-Manville
Owens-Corning Fiberglass
Libbey-Owens-Ford
PPG
Anchor Hocking
Ball Corporation
Hillsboro Glass
Obear-Nestor Glass Div.
of Indian Head Inc.
Obear-Nestor Glass Div.
of Indian Head Inc.
Kerr Glass
Metro Containers
National Bottle Corporation
Owens-Illinois
Owens-Illinois
Thatcher Glass
Eire Glass
Johnson Glass and Plastic
Corporation
Kimble Div. of Owens-Illinois
Peltier Glass
Reha Glass
Johns-Manville
Anchor Hocking
Brockway Glass
Foster-Forbes
Glass Containers
Glass Containers
Kerr Glass
Midland Glass
Owens-Illinois
Plant Location
Ft. Lauderdale
Jacksonville
Lakeland
Bradenton
Tampa
Forest Park
Warner Robins
Atlanta
Athens
Winder
Fairburn
Ottawa
Mt. Zion
Gurnee
Mundelein
Hillsboro
St. Louis
Lincoln
Plainfield
Do! ton
Joliet
Alton
Streator
Streator
Park Ridge - 3 Plants
Chicago
Chicago Heights
Ottawa
Chicago
Waukegan
Winchester
Lapel
Marion
Gas City
Indianapolis
Plainfield
Terre-Haute
Gas City
D-3
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State SIC
Code
Indiana 3221
(cont.) 3229
3229
3229
3229
3229
3229
3229
3296*
3296*
Kansas 3296*
3296*
3296*
3296*
Kentucky 3229
3229
3229
3229
3229
3229
Louisiana 3221
3221
3221
3229
Maryland 3211
3221
3221
3221
3229
3229
Massachu- 3211
setts 3221
3221
3229
3229
3229
3229
Manufacturer
Thatcher Glass
Canton Glass
Corning Glass
Indiana Glass
Kokomo Opalescent Glass
Kimble Div. of Owens-Illinois
St. Clair Glass Works
Sinclair Glass
Certain-Teed
Johns-Manville
Certain-Teed Products
Certain-Teed Products
Johns-Manville
Owens-Corning Fiberglass
Corning Glass
Corning Glass
General Electric
General Electric
GTE-Sylvania
Venezian Art Glass
Laurens Glass Division
of Indian Head
Owens-Illinois
Underwood Glass
Libbey Glass Division of
Owens-Illinois
PPG Industries
Chattanooga Glass
Columbia Glass
Carr-Lowrey Division of
Anchor Hocking
Anchor Hocking
Kimble - Terumo Division of
Owens-Illinois
Guardian Industries
Foster-Forbes
Owens-Illinois
American Optical
Emerson And Cuming
GTE Sylvania
GTE Sylvania
Plant Location
Lawrenceburg
Hartford City
Bluff ton
Dunkirk - 2 Plants
Kokomo
Warsaw
Elwood
Hartford City
She! by vi lie
Richmond
Kansas City
Wichita Falls
McPherson
Kansas City
Danville
Harrodsburg
Lexington
Somerset
Verseilles
Callettsburg
Ruston
New Orleans
Harahan
Shreveport
Cumberland
Baltimore
Baltimore
Baltimore
Baltimore
Elkton
Webster
Mil ford
Mansfield
Southbridge - 2 plants
Canton
Danvers
Ipswich
D-4
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
Michigan
Minnesota
Mississippi
*
Missouri
New Hampshire
New Jersey
New York
SIC-
Code
3211
3211
3211
3221
3221
3221
3221
3221
3221
3229
3229
3211
3211
3229
3229
3211
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3229
3229
3229
3229
3229
3229
3296*
3296*
3296*
3221
3221
3221
3221
3229
Manufacturer
Ford Motor Company
Guardian Industries
Guardian Industries
Owens-Illinois
Brockway Glass
Midland Glass
Chattanooga Glass
Chattanooga Glass
Glass Containers
Ferro Corporation
General Electric
C - E Glass
PPG Industries
Pittsburg Corning Corporation
GTE Sylvania
C - E Glass
Anchor Hocking
Brockway Glass
Kerr Glass
Leone Industries
Metro Containers
Metro Containers
Midland Glass
Owens-Illinois
Owens-Illinois
Thatcher Glass
Wheaton Glass
Friedrich and Dimmock
Kimble Division of
Owens-Illinois
Potters Industries
Thermal American Fused Quartz
Wheaton Glass
Wheaton Products
Certain-Teed Products
Johns-Manville
Owens-Co rni ng-Fi berg! ass
Glenshaw Glass
Leone Industries
Owens-Illinois
Thatcher Glass
American Optical
Plant Location
Dearborn
Carleton
Detroit
Charlotte
Rosemount
Shakopee
Gulf Port
Mineral Wells
Jackson
Fl owood
Jackson
Saint Louis
Crystal City
Sedalia
Greenland
Cinnaminson
Salem
Freehold
Millville
Bridgeton
Jersey City
Carteret
Cliffwood
Bridgeton
North Bergen
Wharton
Mill vllle
Millville
Vine! and
Carlstadt
Mintville
Millville - 2 plants
Millville
Berl i n
Berlin
Barrington
Orangeburg
Rochester
Brockport
Elmira
Buffalo
D-5
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
SIC
Code
Manufacturer
Plant Location
New York
(continued)
3229
3229
3229
3229
3229
3229
3296*
Bausch And Lomb
Corning Glass
Eastman Kodak
GiHinder Brothers
Warren L. Kessler
Super Glass
Owens-Corning Fiberglass
Rochester
Corning - 4 plants
Rochester
Port Jervis
Bethpage
Brooklyn
Delmar
North
Carolina
3211
3221
3221
3221
3229*
3229*
3229*
Libbey-Owens-Ford
Ball Corporation
Laurens Glass Division of
Indian Head
Owens-Illinois
PPG Industries
PPG Industries
United Merchants
Laurinburg
Asheville
Henderson
Winston-Sal em
Lexington
Shelby
Statesville
Ohio
3211
3211
3211
3211
3221
3221
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
Guardian Industries
Guardian Industries
Libbey-Owens-Ford
Libbey-Owens-Ford
Brockway Glass
Chattanooga
Anchor Hocking
E. 0. Brody
Corning Glass
Crystal Art Glass
Federal Glass
General Electric
General Electric
General Electric
General Electric
General Electric
Guernsey Glass
Labind Glass
Lancaster Glass
Imperial Glass Corporation,
a subsidiary of Lennox
Crystal, Incorporated
Libbey Glass Division of
Owens-Illinois
TV Products Division of
Owens-Illinois
TV Products Division of
Owens-Illinois
RCA Corporation
Rodefer-Gleason Glass
Super Glass Corporation
Mill bury
Upper Sandusky
East Toledo
Rossford
Zanesville
Mount Vernon
Lancaster - 2 plants
Cleveland
Greenville
Cambridge
Columbus
Willoughby
Logan
Bucyrus
Miles
Cleveland
Cambridge
Grand Rapids
Lancaster - 2 Plants
Bellaire
Toledo
Columbus
Perrysberg
Circleville
Bellaire
Cambridge
D-6
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
Ohio
(continued)
Oklahoma
SIC
Code
3229
3229
3229
3229*
3296*
3296*
3211
3211
3221
3221
3221
3221
3221
3221
3229
3229
3229
3229
Manufacturer
Techniglas, Incorporated
Variety
Holophane Division of
Johns-Manville
Johns-Manville
Johns-Manville
Owens-Corning Fiberglas
ASG Industries
Ford Motor Company
Ball
Brockway Glass
Brockway Glass
Kerr
Liberty Glass
Midland Glass
Bartlett-Collins
Corning Glass
Scott Glass
Scott Glass Products
Plant Location
Newark
Cambridge
Newark
Waterville -
Defiance - 3
Newark
Okmul gee
Tulsa
Okmul gee
Muskogee
Ada
Sands Springs
Sapulpa
Henryetta
Sapulpa
Muskogee
Cedars
Pocola
2 plants
plants
Oregon
3221
Owens-Illinois
Portland
Pennsyl-
vania
3211
3211
3211
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3221
3229
3229
3229
3229
3229
3229
3229
3229
3229
ASG Industries
PPG Industries
PPG Industries
Anchor Hocking
Brockway Glass
Brockway Glass
Diamond Glass
Foster Forbes
Glass Containers Corporation
Glass Containers Corporation
Glass Containers Corporation
Glenshaw Glass
Menlo Containers
Owens-Illinois
Pierce Glass Division of
Indian Head
Corning Glass
Corning Glass
Corning Glass
Corning Glass
General Electric
K. R. Haley Glassware
Houze Glass
Jeannette Corporation
Jeannette Shade And Novelty
Jeannette
Carlisle
Meadville
Connellsville
Brockway - 2 plants
Washington - 2 plants
Roversford
Oil City
Knox
Marienville
Parker
Glenshaw
Washington
Clarion
Port Allegheny
CharTeroi
State College
Wellsboro
Bradford
Bridgeville
Greensburg
Point Marion
Jeannette
Jeannette
D-7
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
Pennsyl-
vania
(continued)
Rhode
Island
*»outh
Parnl ina
Tennessee
Texas
SIC
Code
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229*
3296*
3221
3229
3229*
3221
3229*
3229*
3211
3211
3211
3221
3229*
3229*
3211
3221
3221
3221
3221
3221
3229
3229
3296*
3296*
Manufacturer
J. H. Millstein
Kopp Glass
Lennox Crystal
Mayflower Glass Works
Kimble Division of
Owens-Illinois
Kimb. Divis. of Owens-Illinois
TV Prod. Div of Owens-Illinois
Pennsylvania Glass Products
Phoenix Glass
Pittsburg Corning
Schott Optical Glass
. L. E. Smith Glass
Victory Glass
Westmoreland Glass
Owens-Corning Fiberglas
Certain-Teed
National Bottle Corporation
Corning
Owens-Corning-Fiberglas
Laurens Glass Division
of Indian Head
Owens-Corning Fiberglas
Owens-Corning Fiberglas
ASG Industries
ASG Industries
Ford Motor Company
Chattanooga Glass
Reichold Chemical
Owens-Corning Fiberglas
PPG Industries
Anchor Hocking
Chattanooga Glass
Glass Containers Corporation
Kerr Glass
Owens-Illinois
EMC Glass
Multicolor Glass
Oohns-Manville
Owens-Corning Fiberglas
Plant Location
i
Jeannette
Swissvale
Mount Pleasant
Latrobe
Philadelphia
Pitts ton
Pitts ton
Pittsburg
Monaca
Port Allegheny
Duryea
Mount Pleasant
Jeannette
Grapeville
Huntington
Mountaintop
Coventry
Central Falls
Ashton
Laurens
Aiken
Anderson
Greenland
Kingsport
Nashville
Chattanooga
Nashville
Jackson
Wichita Falls
Houston
Corisicana
Palestine
Waxahachie
Waco
Decatur
San Antonio
Cleburne
Waxahachie
Virginia
3229
Corning Glass
D-8
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
SIC-
Code
Washington
3221
3229
West
Virginia
3211
3211
3221
3221
3221
3221
3221
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3229
3296*
Manufacturer
^~""^~^^^^^^-«^^»^^^_
Northwestern Glass Division
of Indian Head
Penberty Glass Division
of Nuclear Pacific
Fourco Glass
Libbey-Owens-Ford
Chattanooga Glass
Kerr Glass
National Bottle Corporation
Owens-Illinois
Owens-Illinois
Beaumont
Blenko Glass
Brockway Glass
Demuth Glass Division of
Brockway
Colonial Glass
Corning Glass
Corning Glass
Crescent Glass
Davis Lynch Glass
Elite Company
Erskine Glass
Fenton Art Glass
Fostoria Glass
Gentile Glass
Gladding-Vitro-Agate
Hamon Handcrafted Glass
Harvey Industries
Kanawha Glass
Lewis County Glass
Louie Glass
Mid-Atlantic Glass
Minners Glass
Pennsboro Glass
Pilgrim Glass
Rainbow Art Glass
Scandia Glass Works
Seneca Glass
Earl Shelby Glass
Sloan Glass
Viking Glass
Viking Glass
West Virginia Glass Specialty
Westinghouse Electric Corp.
Paul Wissmach Glass
Johns-Manville
Plant Location
Seattle
Seattle
Clarksburg
Charleston
Keyser
Huntington
Parkersburg
Fairmont
Huntington
Morgantown
Milton
Clarksburg
Parkersburg
Weston
Martinsburg
Parkersburg
Wellsboro
Roversford
Cameron
WeTlsburg
Williamstown
Moundsville
Star City - 2 plants
Parkersburg
Dunbar
Clarksburg
Dunbar
Jane Lew
Weston
Ellenboro
Salem
Pennsboro
Ceredo
Huntington
Kenova
Morgantown
Huntington
Culloden
New Martinsville
Huntington
Weston
Fairmount
Paden City
Vienna
D-9
-------
PRIMARY GLASS MANUFACTURERS (cont.)
State
SIC
Code
Manufacturer
Plant Location
Wisconsin
3221
Foster-Forbes Glass
Burlington
* Fiberglass
D-10
-------
APPENDIX E
END-PRODUCTS OF EACH GLASS MANUFACTURING SEGMENT
STANDARD INDUSTRIAL CLASSIFICATION
'£-!
-------
APPENDIX E
END-PRODUCTS OF EACH GLASS MANUFACTURING SEGMENT
STANDARD INDUSTRIAL CLASSIFICATION
SIC NO. 3211 FLAT GLASS
Establishments primarily engaged in manufacturing flat glass.
This industry also produces laminated glass, but establishments
primarily engaged in manufacturing laminated glass from purchased
flat glass are classified in Industry 3231.
Building Glass, flat
Cathedral glass
Float glass
Glass, colored: cathedral
and antique
Glass, flat .
Insulating glass, sealed
units: mitse
Laminated glass, made from
glass produced in the same
establishment
Multiple-glazed insulating
units, mitse
Opalescent flat glass
Ophthalmic glass, flat
Optical glass, flat
Picture glass
Plate glass blanks
for optical or
ophthalmic uses
Plate glass, polished
and rough
Sheet glass
Sheet glass blanks
for optical or
ophthalmic uses
Skylight glass
Spectacle glass
Structural glass, flat
Tempered glass, mitse
Window glass, clear
and colored
SIC NO. 3221 GLASS CONTAINERS
Establishments primarily engaged
containers for commercial packing and
Ampoules, glass
Bottles for packing,
bottling, and canning:
glass
Carboys, glass
Containers for packing,
bottling, and canning:
glass
Cosmetic jars, glass
Fruit jars, glass
in manufacturing glass
bottling, and for home canning.
Jars (packers' ware), glass
Jugs (packers' ware), glass
Medicine bottles, glass
Milk bottles, glass
Packers' ware (containers),
glass
Vials, glass: made in glass
making establishments
Water bottles, glass
E-2
-------
SIC NO. 3229
CLASSIFIED
PRESSED AND BLOWN GLASS AND GLASSWARE, NOT ELSEWHERE
Establishments primarily engaged in manufacturing glass and
glassware, not elsewhere classified, pressed, blown, or shaped
from glass produced in the same establishment. Establishments
primarily engaged in manufacturing textile glass fibers are also
included in this industry, but establishments primarily engaged in
manufacturing glass wool insulation products are classified in
Industry 3296. Establishments primarily engaged in the production
of pressed lenses for vehicular lighting, beacons, and lanterns are
also included in this industry, but establishments primarily engaged
in the production of optical lenses are classified in Industry 3832.
Establishments primarily engaged in manufacturing glass containers
are classified in Industry 3221, and complete electric light bulbs
in Industry 3641.
Art glassware, made in
glassmaking plants
Ash trays, glass
Barware, glass
Battery jars, glass
Blocks, glass
Bowls, glass
Bulbs for electric lights,
without filaments or
sockets: mitse
Candlesticks, glass
Centerpieces, glass
Chimneys, lamp: glass
pressed or blown
Christmas tree ornaments,
from glass: mitse
Clip cups, glass
Cooking utensils, glass
and glass ceramic
Drinking straws, glass
Fibers, glass
Flameware, glass and
glass ceramic
Frying pans, glass and
glass ceramic
Glass blanks for electric
light bulbs
Glass brick
Glassware:, art, decorative,
and novelty
Glassware, except glass
containers for packing,
bottling, and home canning
Goblets, glass
Illuminating glass: Tight
shades, reflectors, lamp
chimneys, and globes
Industrial glassware and
glass products, pressed
or blown
Inkwells, glass
Insulators, electrical:
glass
Lamp parts, glass
Lamp shades, glass
Lantern globes, glass:
pressed or blown
Lens blanks, optical and
ophthalmic
Lenses, glass: for lanterns,
flashlights, headlights,
and searchlights
Level vials for instruments,
glass
E-3
-------
Light shades, glass: pressed
or blown
Lighting glassware, pressed
or blown
Novelty glassware
Ophthalmic glass, except
flat
Optical glass blanks
Reflectors for lighting
equipment, glass:
pressed or blown
Refrigerator dishes and
jars, glass
Scientific glassware,
pressed or blown: made in
glassmaking plants
Stemware, glass
Tableware, glass and
glass ceramic
Teakettles, glass and
glass ceramic
Technical glassware and
glass products, pressed
or blown
Textile glass fibers
Tobacco jars, glass
Trays, glass
Tubing, glass
Tumblers, glass ,
TV tube blanks, glass
Vases, glass
Yarn, fiberglass: made in
glass plants
SIC NO. 3296 MINERAL WOOL
Establishments primarily engaged in manufacturing mineral wool
and mineral wool insulation products made of such silicious materals
as rock, slag, and glass, or combinations thereof. Establishments
primarily engaged in manufacturing asbestos insulation products
are classified in Industry 3292, and textile glass fibers in
Industry 3229.
Fiberglass insulation
Glass wool
NOTE: Taken from the 1972 edition of the Standard Industrial
Classification Manual.
E-4
-------
1. REPORT NO.
EPA-450/3-79-005a
TECHNICAL REPORT DATA
(Mease read Instructions on the reverse before completing]
2.
Glass Manufacturing Plants, Background Information:
Standards of Performance
'. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Research Triangle Park, North Carolina 27711
12. SPONSORING AGENCY NAME AND ADDRESS
3. RECIPIENT'S ACCESS/ON NO.
5. REPORT DATE
June 1979
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
'"S7rTRYN°TE.SVolume I discusses the proposed stand rds and the resulting environ-
mPnta1 ?-??ieS?"^' ae!fe?I; V°lume J1' to be Published when the standardllrl
.1 discuss any differences between the proposed and promnlgat.pH
nthe dent
p
uncontrolled emissions of particulate matter from these furnaces
h Env1ronraSn1»1 imPact anS «onom1c ImpSrsSSnS? qJan??fy!
Pr°P°Sed standard and alternative control options a?e included
17.
a.
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
Air Pollution .= .
Glass Manufacturing and Processing
Emission Standards
18. DISTRIBUTION STATEMENT "
Unlimited-Available to the public free
charge from: US EPA Library (MD-35)
Research Triangle Park, N.C. 27711
of
b.IDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
19' ,S1EC4RITY CLASS (This Report}
Unclassified
20. SECURITY CLASS (This page)
Unclassified
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
c. COSATI Field/Group
21. NO. OF PAGES
278
22. PRICE
-------
-------
-------
-------
|