U.S. DEPARTMENT OF COMMERCE i
>
National Technical Information Service
PB-273 788
Source Assessment
Pressed and Blown Glass
Manufacturing Plants
Baftelle Columbus Labs, Ohio
Prepared for
Industrial Environmental Research Lab, Research Triangie Park, N C
Jan 77
-------
£?A-600/2-77-005
JANUARY 1977 Environmental Protection Technology Series
SOURCE ASSESSMENT: Pressed and Blown
Glass Manufacturing Plants
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, ULS. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
/The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been re viewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
TECHNICAL HEPORT DATA
(Pirate read luttnictioiis on ilu- reverse be/ore completinf)
1. REPORT NO.
EPA-600/2-77-OOS
3. RECIPIENT'S ACCESSION NO.
TITLE AND SUOTITLE '
SOURCE ASSESSMENT: PRESSED AND BLOWN GLASS
MANUFACTURING PLANTS
6. REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) _____^^______^^_________
J. R. Schorr, Diane T. Hooie, M. Clifford Brockway,
Philip R. Sticksel. and Dale E. Niesz
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORftANIZATION NAME AND ADDRESS
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
10. PROGRAM ELEMENT NO.
1AB015; ROAP 21AFA-013.
11. CONTRACT/GRANT NO.,
68-02-1323, Task 37
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final: 9/75-1/77
14. SPONSORING AGENCY CODE
EPA-ORD
^SUPPLEMENTARY NOTEsTask officer for this report is Dale A. Denny, Mail Drop 62,
919/549-8411 Ext 2547. Earlier related reports are: EPA-650/2-75-019a and the
EPA-600/2-76-032 series.
16. ABSTRACT
This report summarizes the results of a study to gather and analyze background information
and technical data related to air emissions from glass manufacturers producing pressed and blown glass-
ware. This includes all glassware except flat glass, glass containers, and fiber glass. The report covers
emissions from three areas within the plant: raw materials preparation and handling, glass melting, and
forming and finishing operations. Emissions from the melting furnace account for over 80 percent of
the total plant emissions. The major pollutants are NOX, SOX, and submicron particulates consisting
predominately (*80%) of mineral sulfates but can also include fluoride and borate compounds. NOX has
the largest emission factor (4.5 g/kg) with annual emissions of 57.5 x 10" g. In comparison with
national emissions from stationary sources, NOV emissions from glass melting furnaces contributes 0,17
__-_- ^ * '^ '' s ' " i ' 4 . *
percent of the total. The source severity is a measure of the potential environmental effect of air
emissions and is the ratio of the maximum average ground level concentration compared to the; primary
ambient air quality standard for criteria pollutants. For this study source severity factors determined
were largest for NOX, SOX, lead (PbO), and other partibulafes. Other emissions had a low severity
factor (*0.01).
7.
DESCRIPTORS
KEY WORDS AND DOCUMENT ANALYSIS
»*>^M^V^^^^'M^HHIBPIM*
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Assessments
Glass Industry
ontainers
Nitrogen Oxides
ulfur Oxides
Dust
Minerals
Sulfates
Air Pollution Control
Source Assessment
Stationary Sources
Particulate
Mineral Sulfates
13B
14B
11B
13D
07B
11G
08G
8. DISTRIBUTION STATEMENT
Unlimited
19. SfcCUHIT Y CLASS (Tliil Report)
Unclassified
21. I
20. SCCUHITY CLASS
Unclassified
22. PRICE
CPA Form 2220-1 (9-73)
-------
EPA-600/2-77-005
January 1977
SOURCE ASSESSMENT:
PRESSED AND BLOWN
GLASS MANUFACTURING PLANTS
by
J.R. Schorr, D.T. Hooie, M.C. Brockway,
P.R. Sticksel, and D.E. Niesz
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-1323, Task 37
ROAPNo. 21AFA-013
Program Element No. 1AB015
EPA Task Officer: Dale A. Denny
*(
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
The Industrial Environmental Research Laboratory (IERL) of EPA has
the responsibility for insuring that air pollution control technology
is available for stationary sources. If control technology is unavailable,
inadequate, uneconomical, or socially unacceptable, then development of
the needed control techniques is conducted by IERL. Control approaches
considered include: process modifications, feedstock modifications,
add-on control devices, and complete process substitution. The scale
of control technology programs range from bench- to full-scale demonstra-
tion plants.
The Chemical Processes Section of IERL has the responsibility for
developing control technology for a large number (> 500) of operations
in the chemical and related industries. As in any technical program,
the first step is to identify the unsolved problems. Each of the
industries is to be examined in detail to determine if there is suf-
ficient potential environmental risk to justify the development of
control technology by IERL. This report contains the data necessary
to make that decision for pressed and blown glass manufacturing plants.
Battelle's Columbus Laboratories was contracted by EPA to investigate
the environmental impact of the pressed and blown glassware industry, which
represents a source of emissions in accordance with EPA's responsibility
as outlined above. Dr. J. Richard Schorr served as Program Manager for
this study. Dr. Dale A. Denny served as EPA Project Monitor. The study
was completed by IERL-RTP. Project responsibility was transferred to the
Industrial Pollution Control Division of lERL-Cincinnati on October 15, 1975.
iii
-------
TABLE OF CONTENTS
Page
INTRODUCTION. .' ... 1
SUMMARY 4
DESCRIPTION OF THE PRESSED AND BLOWN GLASSWARE INDUSTRY 11
General Process Description 11
Plants and Location 13
Shipment Value and Volume 13
Glass Compositional Types 13
Soda/Lime Silica Glasses 15
Borosilicate Glasses 16
Lead Glasses. ... ......... 16
Opal Glasses 16
Process Details 17
Batch Handling 19
Batch Composition 21
Melting and Fining 21
Forming 25
Post-Forming and Finishing 28
EMISSIONS 32
Raw-Materials Preparation and Handling 32
Glass Melting 38
Nitrogen Oxides 38
Sulfur Oxides 42
Fluorides 44
Carbon Monoxide . . 44
Preceding page blank
-------
TABLE OF CONTENTS
(Continued)
Page
45
Hydrocarbons
Particulates
AO
Selenium
Other Emissions 50
Forming and Finishing 50
Forming 50
Treatment 52
Annealing 53
Decorating 53
Frosting of Light Bulbs 55
Acid Cleaning 55
Emission Characteristics 55
Raw Materials Preparation 55
Glass Melting 56
Forming and Finishing 56
Ground-Level Concentrations 57
Affected Population 68
CONTROL TECHNOLOGY 71
Raw-Materials Preparation 72
Emissions ......' 7'2
Raw-Materials-Control Technology 73
Glass-Melting Operation 76
Emissions 76
Glass-Melting-Control Technology 78
Efficiency of Equipment 85
VI
-------
TABLE OF CONTENTS
(Continued)
Page
Forming and Finishing 86
Forming Emissions 87
Forming and Finishing Control Technology 87
Decorating 87
Frosting of Electric Light Bulbs 88
Acid Cleaning 88
Surface Treatment 88
FUTURE PRODUCTION OF PRESSED AND BLOWN GLASSWARE 90
REFERENCES 91
APPENDIX A
Geographical Listing of Pressed-and-Blown Glass Plants A-l
APPENDIX B
Emissions Data B-l
APPENDIX.C
Stack Heights from the Various Segments of Glassmaking Process. . C-l
APPENDIX D
State Listing of Total Emissions as of 1972 C-l
LIST OF TABLES
TABLE 1. GLASS INDUSTRY STATISTICS 3
TABLE 2. AVERAGE UNCONTROLLED EMISSIONS OF MAJOR SPECIES FROM
PRESSED AND BLOWN GLASSWARE PLANTS 6
TABLE 3. SOURCE SEVERITY FOR PRESSED AND BLOWN GLASS EMISSIONS. 9
vii
-------
LIST OF TABLES
(Continued)
TABLE 4. PROPORTION OF INDUSTRY OUTPUT ACCOUNTED FOR BY THE
CONSUMER, SCIENTIFIC, TECHNICAL, AND INDUSTRIAL
GLASSWARE SEGMENTS OF SIC 3229 14
TABLE 5. PARTICULATE EMISSIONS DURING RAW-MATERIAL PREPARATION
AND HANDLING FOR PRESSED-AND-BLOWN GLASS 36
TABLE 6. SPECIFICATION LIMITS FOR SEVERAL RAW MATERIALS USED
IN PRESSED-AND-BLOWN GLASS MANUFACTURE t14' 37
TABLE 7. EMISSIONS FROM PRESSED AND BLOWN GLASS MELTING
FURNACE OPERATIONS 39
TABLE 8. EMISSIONS FROM THE FORMING AND FINISHING OPERATIONS
FOR ALL PRESSED AND BLOWN GLASS 51
TABLE 9. EMISSIONS FROM THE ANNEALING OF PRESSED AND BLOWN
GLASSWARE 54
TABLE 10. PARAMETERS OF, A SODA/LIME GLASS-MELTING FURNACE (8.1
Gg ANNUAL PRODUCTION) REPRESENTATIVE OF THE PRESSED
AND BLOWN INDUSTRY AS USED IN ATMOSPHERIC-DISPERSION
CALCULATIONS $8
TABLE 11. PARAMETERS OF A SODA/LIME GLASS-MELTING FURNACE (29.9
Gg ANNUAL PRODUCTION) REPRESENTATIVE OF THE PRESSED
AND BLOWN INDUSTRY AS USED IN ATMOSPHERIC-DISPERSION
CALCULATIONS 59
TABLE 12. PARAMETERS OF A LEAD GLASS-MELTING FURNACE REPRESENT-
ATIVE OF THE PRESSED AND BLOWN INDUSTRY AS USED IN
ATMOSPHERIC-DISPERSION CALCULATIONS 60
TABLE 13. RELATIVE FREQUENCY OF ATMOSPHERIC STABILITIES^). . . 64
\
TABLE 14. MAXIMUM POLLUTANT CONCENTRATIONS AND SOURCE SEVERITY
FOR EMISSIONS FROM REPRESENTATIVE PRESSED AND BLOWN
MELTING FURNACES 65
TABLE 15. MAXIMUM AVERAGE GROUND-LEVEL CONCENTRATION (Xmax) OF
SELECTED AIR POLLUTANTS FROM REPRESENTATIVE MATERIALS
HANDLING AND TREATMENT OPERATIONS 67
TABLE 16. GLASS-GRADE PARTICLE-SIZE SPECIFICATIONS FOR SAND,
LIMESTONE, AND 10- AND 20-MESH DOLOMITE '. 74
TABLE 17. MAXIMUM USE TEMPERATURE FOR VARIOUS FABRIC-FILTER
MATERIALS 82
vili
-------
LIST OF TABLES
(Continued)
Page
TABLE A-l. GEOGRAPHICAL LISTING OF THE 176 PRESSED AND BLOWN
GLASS PLANTS A-l
TABLE B-l. SUMMARY OF SOURCE TEST DATA FOR MATERIALS PREPARATION
AND HANDLING(a) B-2
TABLE B-2. NOx EMISSIONS FROM PRESSED AND BLOWN GLASSWARE
FURNACES B-4
TABLE B-3. SOx EMISSIONS FROM PRESSED AND BLOWN GLASSWARE
FURNACES B-6
TABLE B-4. PARTICULATE MEISSIONS FROM PRESSED AND BLOW GLASSWARE
FURNACES B-8
TABLE B-5. CO EMISSIONS FROM PRESSED AND BLOW GLASSWARE B-9
TABLE B-6. HC EMISSIONS FROM PRESSED AND BLOWN GLASSWARE .... B-ll
TABLE B-7. HYDROCARBON EMISSIONS FROM FORMING PRESSED AND BLOWN
GLASSWARE B-14
TABLE C-l. TYPICAL STACK HEIGHTS OF BATCH-HANDLING OPERATIONS
FOR SODA/LIME GLASS C-l
TABLE C-2. TYPICAL STACK HEIGHTS FOR MELTING OPERATIONS OF GLASS
FURNACES C-2
TABLE C-3. TYPICAL STACK HEIGHTS FOR FORMING OPERATIONS C-3
TABLE C-4. TYPICAL STACK HEIGHTS FOR ANNEALING OPERATIONS OF
BOROSILICATE GLASS C-5
TABLE C-5. TYPICAL STACK HEIGHTS FOR DECORATING OPERATIONS OF
SODA/LIME GLASS C-6
TABLE C-6. TYPICAL STACK HEIGHTS FOR TREATMENT OPERATIONS OF
SODA/LIME GLASS C-7
TABLE D-l. STATE LISTING OF TOTAL EMISSIONS AS OF 1972 D-l
TABLE D-2. (CONTINUED) D-2
TABLE D-3. (CONTINUED) D-3
TABLE E-l. CONVERSION FACTORS E-l
ix
-------
LIST OF FIGURES
FTRTTOF 1
FIGURE 2.
FIGURE 3.
FIGURE 4.
FIGURE 5.
FIGURE 6.
FIGURE 7.
FIGURE 8.
PROCESS FLOW DIAGRAM: SIC 3229, PRESSED-AND-BLOWN
(Q)
PROCESS FLOW DIAGRAM OF A TYPICAL BATCH PLANT v ...
ILLUSTRATIVE SKETCH OF A CONTINUOUS TANK-TYPE OF
SIDE-PORT REFENERATIVE-MELTING FURNACE '3'
TYPICAL POINTS OF PARTICULATE EMISSION FROM RAW-
MATERIALS HANDLING
NOX EMISSIONS FROM FLINT GLASS-CONTAINER FURNACE^18^ .
PARTICULATE EMISSIONS SHOWN ARE LINEAR WITH THE
RECIPROCAL OF BRIDGEWALL TEMPERATURE
Pag
12
18
20
23
26
33
41
47
FIGURE 9. PARTICULATE EMISSIONS FOR GLASSES AS A FUNCTION OF
PULL RATE 48
FIGURE 10. ILLUSTRATION DEPICTING CALCULATION OF AREA WHICH CON-
TAINS THE AFFECTED POPULATION 69
x
-------
LIST OF SYMBOLS
Symbol Definition
AAQS Ambient air quality standard
A, B, C, D, E, F Atmospheric stability classes
a, b, c, d, e, £ Constants in dispersion equations
A^ The ratio Q/aciru
B_ The ratio -H2/2c2
Jx
CI Confidence Interval
D. Inside stack diameter
e Natural logarithm base
H Effective stack height
h Physical stack height
AH Plume rise
k "Student t" test variable
m Number of samples
N. Sample value
p Atmospheric pressure
Q Mass emission rate
R Downwind dispersion distance from source
of emission release
s Sample standard deviation
S Source severity, ratio X /AAQS
UlClaK
T Ambient temperature
a
T Stack gas temperature
S
t Instantaneous averaging time of 3 minutes
t Averaging time
TLV Threshold limit value
u National average wind speed
v Stack gas exit velocity
S
R Horizontal distance from centerline of
dispersion
xi
-------
LIST OF SYMBOLS (Continued)
Definition
Sample mean
3.14
Standard deviation
Downwind ground level concentration at
reference coordinate x and y with emission
height of H
X Time average ground level concentration of
an emission
X Instantaneous maximum ground level
concentration
X Time average maximum ground level
concentration
xii
-------
SECTION I
INTRODUCTION
Air emissions released in the manufacture of pressed and blown
glassware have been examined in this study. This report describes the
nature of the pressed and blown glass industry, the nature of air emissions
from this industry and their environmental impact, the control technology
employed, and the future growth of this industry segment. Pressed and blown
glassware is one of three segments of the glass industry (glass containers,
flat glass, and pressed-and-blown glassware). Each segment is defined by a
Standard Industrial Classification (SIC) number, as used by the Department of
Commerce. Pressed-and-blown glassware is designated by SIC 3229 and includes
all glassware not classified under SIC 3221 or SIC 3211. This industry
segment is very diversified and includes products such as:
Table, kitchen, art, and novelty glassware
Lighting and electronic glassware
Scientific, technical, and other glassware
Textile glass fibers.
Industry shipments in 1973 had a value of $1.12 billion, which was about
25 percent of the glass industry total. This has increased to 1.3 billion
by 1974. '.
i
Glass containers are designated by SIC 3221 and include the
manufacture of glass containers for food, beverages, medicines, toiletries,
and cosmetics. It includes both narrow-neck and wide-mouth containers.
Shipments in this segment have grown at an average rate of about 3.5 percent
since 1971. Industry shipments in 1973 had a value of $2.3 billion, or about
51 percent of the glass industry total.
Flat glass is designated by SIC 3211. This includes both the manu-
facture of flat glass and some fabrication of flat glass into a tempered or
laminated-glass product. Flat glass products include: window glass, plate
glass, wire glass, tempered glass, and laminated glass. These products are
consumed primarily by the automotive and construction industries. Value of
shipments in 1973 was $1.1 billion, which was 24 percent of the glass
industry total.
-------
Table 1 lists some 1973 statistics on the three segments of the
glass industry. It shows that over 154,000 people produced merchandise
valued at over $4.5 billion.* Only 9 percent of the glass product shipments
are produced by the pressed and blown segment of the glass industry.
Separate Source Assessment Documents have been prepared for the
flat glass and container glass segments. This report deals only with pressed
and blown glassware exclusive of fibrous glass; however, many of the same
emissions and control technology are also found in the other glass industry
segments. The report delineates the various emission points, identifies the
type and quantity of emissions, and describes the characteristics of the air
pollutants found. Mass emissions for criteria pollutants (particulates,
NO , SO , and CO and hydrocarbons) from pressed and blown glassware plants
X X
are compared with national emissions from all stationary sources. The
theoretical maximum time average, ground-level concentrations due to emissions
from a pressed and blown glass plant are compared to the corresponding ambient
air-quality standards. Control technology which is being used or could be
applicable to the manufacture of glass products is also discussed. The
manufacturing operations for glass production are grouped into three categories:
Preparation of raw materials
Glass melting
Forming and finishing
Emissions and control technology for each of these three areas are presented.
* References are listed on Page 92.
-------
TABLE 1. GLASS-INDUSTRY STATISTICS
(a)
SIC Industry Segment
3221 Glass Containers
3211 Flat Glass
3229 Pressed-and-Blown Glass
-------
SECTION II
SUMMARY
This document describes a study of air emissions released during
the production of pressed and blown glassware. The industry is defined by
Standard Industrial Classification (SIC) No. 3229, except for glass fiber
production. The study encompasses the preparation of raw materials at the
plant site, the production of molten glass in a furnace, the forming of glass
articles, and certain post-forming operations necessary to manufacture the
products of this diverse industry.
The pressed and blown glass industry in the United States produced
an estimated 1.46 Tg* (1.614 million tons) of salable product in 1974. Of
that total, about 77 percent (1.12 Tg) was soda-lime glass; 11 percent
(0.15 Tg) was borosilicate glass; 5 percent (0.01 Tg) was lead glass; and
about 7 percent (0.10 Tg) was opal glass. In 1974, the industry segment
consisted of 110 manufacturers operating 176 plants. Geographically, these
plants were concentrated in or about the North-Central region of the country,
primarily in New York, Illinois, Indiana, Ohio, Pennsylvania, New Jersey, and
West Virginia. Pressed and blown glassware was produced in 28 states. The
average county population density at a plant site was estimated to be 356
people/km .
Manufacturing Technology
In a glass-manufacturing process, raw materials (e.g.s sand,
soda ash, limestone) are uniformly mixed and these loose materials are
transported to a furnace where they are melted at elevated temperatures
(> 1500 C) into a homogeneous mass. More than 90 percent of the glass is
made in fossil fuel-fired furnaces where energy is predominately transferred
to the glass by radiation from a flame or reradiation from the refractory
chamber containing molten glass. Molten glass is kept at. elevated temperatures
until it is of a quality (bubble-free) sufficient for making the desired
12
* Tg = 10 gram. Metric prefixes and other conversion
factors are given in Appendix E, Page E-l.
-------
product. The glass is then cooled to approximately 1300 C, and removed from
the furnace, either continuously or cut into "gobs". The molten glass is
fed to machines, where it is formed into desired shapes which then
undergo additional finishing operations. The type of finishing operation
depends upon the type of product being manufactured. Essentially,
all glass products go through an annealing furnace for removal of residual
stresses. The temperatures in annealing range from about 590 to 650 C.
Emissions
Emissions were examined from three areas within the glass
manufacturing plant: (1) raw-materials preparation and handling, (2) glass
melting and (3) forming and finishing. The largest emissions occur from
the glass-melting operation.
Manufacturing Plant
Table 2 summarizes the average emissions of the major species
determined for this study. It should be noted that no one glass plant
will have all of these emissions, because they are dependent upon both
the type of glass produced and the type of process equipment employed. As
can be seen, 98 percent of the plant emissions come from the glass melting
furnace. The major species (over 93%) are NO , SO , and particulates.
i X X
Furnace stack heights average 19 m when ejection air is used and 44 m for
natural draft. Stack heights are summarized in Appendix C.
Total Industry
Nitrogen oxides constitute the second largest emission source
(4.25 g/kg + 43%). Total annual emissions are estimated to be 12.6 Gg which
amounts to approximately 0.109 percent of the 1972 NO- National emissions
from all stationary sources.
Sulfur oxide emissions can result from either sulfur compounds
added in the batch or from sulfur in the fossil fuel. They are the third
largest emission source (2.80 g/kg 4- 62%). Total annual emissions are
estimated to be 8.3 Gg which amounts to approximately 0.130 percent of
the 1972 National emissions of SO from all stationary sources.
-------
TABLE 2. AVERAGE UNCONTROLLED EMISSIONS OF MAJOR SPECIES FROM
PRESSED AND BLOWN GLASSWARE PLANTS
EMISSION SPECIES
EMISSION FACTOR. K/kg
(1) Raw Materials
(2) Glass Melting
(3) Form/Finish
Forming
Treatment
Annealing
Decorating
Frosting
Acid Clean
TOTAL ANNUAL
EMISSIONS. Mp
(1) Raw Materials
(2) Glass Melting
(3) Form/Finish
Forming
Treatment
Annealing
Decorating
Frosting
Acid Clean
NOX
0
4.25+43%
0
0
0.02+1002
0 ~
0
0
0
12,622
48
0
0
48
0
0
0
s°x
0
2.80+62%
0
0
0
0
0
0
0
8,316
0
0
0
0
0
0
0
Particulates
1.91+100%
8.7+60%
0
0.05+100%
Traced)
0
0
0
6
25,839
21
0
17
4
0
0
0
CO
0
0.1+100%
0
0
Trace*c>
0
0
0
0
297
7
0
0
7
0
0
0
Hydrocarbons
0
0.15+53%
0.06+100%
~0
Trace (c)
4.5+100%
" 0
0
0
445
638
178
0
4
456
0
0
Fluorides
!
0
10+100%
0
0
0
0
0.96+100%
0.18+100%
0
3000
103
0
0
0
0
87
16
Selenium
0
0.002+100%
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
HC1
0
0
0
0.02+100%
0
0
0
0
0
0
7
0
7
0
0
0
0
NH3
0
0
0
0
0
0
0.44+100%
0
0
0
20
0
0
0
0
20
0
(a) Based on 3.37 Tg of raw materials processed to melt 2.97 Tg of glass.
(b) Table 8 gives a breakdown of emission factors for various glass types.
(c) Trace is less than 0.01 g/kg.
-------
Particulate emissions from the melting furnace are the highest
emission source (8.70 g/kg ± 60%) of the three major species. Total annual
emissions are estimated to be 25.8 Gg, which represents approximately 0.020%
of National emissions from all stationary sources. Fluorides can be emitted
as both gases and particulates at a rate taken as 10 g/kg ± 100%. Total
annual emissions are estimated to be 3.0 Gg.
Carbon monoxide has an emission factor of 0.1 g/kg ± 100% with total
annual emissions of 0.3 Gg. The emission factor for hydrocarbons is 0.15
g/kg ± 53% with total annual emissions of 0.4 Gg. These contribute 0.002
and 0.001 percent, respectively, to the National emissions from all
stationary sources. Finally, the emissions factor for selenium is 2 mg/kg ±
100% with total annual emissions of 0.002 Gg.
Emissions from raw materials preparation and handling can give
rise to some particulate emissions, primarily from dust generated during
discharging, conveying, crushing, and mixing operations. The composition of
these emissions is the same as that of the raw materials used. The average
emission factor is 1.91 g/kg ± 100%. Total annual emissions are estimated
to be 0.006 Gg or 0.0004 percent of the National particulate emissions from
all stationary sources. A common practice for the industry is the employment
of controls (primarily filter bags) in dust generating areas.
Many different processes are used in the forming and finishing
operations, depending on the type of product being manufactured. Emissions
from the more common operations are identified in Table 2. These emissions
consist of hydrocarbons emitted from the forming operation (0.06 g/kg);
tin oxide, hydrated tin chloride particulates (0.05 g/kg), and HC1
(0.02 g/kg) emitted from surface treatment operations; combustion products
emitted from gas-fired annealing lehrs; and hydrocarbon (4.5 g/kg) emitted
from decorating operations that are used by about 10% of the industry.
Additionally, HF and NH are emitted during frosting of light bulbs and HF
is emitted during acid cleaning of some glass products. The emissions from
all these areas constitute less than 2% of the total emissions by the
industry.
-------
Control Technology
Emissions from furnaces melting soda lime glass are generally
not controlled by add-on equipment. More often, the particulate emissions
from furnaces melting other glass types will be controlled. This practice
varies with geographical location. Frequently, emission standards can be
met without the use of control equipment (e.g., baghouses and electrostatic
precipitators).
Source Severity
Impacts of these emissions are directly related to the ambient
concentrations the emissions create at ground level. Atmospheric dispersion
calculations have been made to calculate the theoretical maximum average
ground-level concentrations (X ) due to emissions from melting,
UicLJv
materials-handling, and surface treatment. The results of these calculations
are presented in Table 3.
The source severity factor, S, has also been used to describe the
impact of the emissions. For those pollutants which have an ambient air-
quality standard (AAQS), S is the ratio of X to the primary AAQS. In
cases where no AAQS has been established, S is based upon the Threshold
Limit Value (TLV) through the following £qua,tion which includes a factor
for correcting the TL'V.to a 24-hour day (8/24) and a safety factor (1/100).
X
s max
TLV (8/24) (1/100)
Results of the Source severity factor calculations also appear in Table 3.
The highest severity factor (0.89) is produced by fluoride emissions from
lead glass melting operations.
A third measure of the impact of the plant emissions is given by
the Affected Population. This measure is defined as the population
around the plant who are exposed to a source severity factor greater than
1.0. Computations of the affected population showed that all sources were
less than 1.0.
-------
TABLE 3. SOURCE SEVERITY FOR PRESSED AND BLOWN GLASS EMISSIONS
(a)
Primary Ambient
Air-Quality Standard
Source Pollutant
Melting Furnace (Glass)
Annual Production (Gg)
NOX
-sox
Particulates
CO
Hydrocarbons
Selenium
Fluoride
Materials Handling
Particulates
Surface Treatment
Hydrogen Chloride
Titanium Chloride
ug/m
Averaging
Time, hr
Theoretical
Maximum Average
Ground-Level-
Concentration, Xmax, Mg/nr
Soda/Lime Lead
100
365
260
4x10*
160
0.67
260
ft \
23
33
24
24
24
8
3
24
24
24
24
24
8.1
7.3
4.5
8.8
0.23
0.36
3.4xlO-3
29.9 4.6
17.7
11.1
21.9 10.7
0.48
0.88
8.4xlO~3 4. 1x10" 3
7.4
1.96
4.11
4.11
Severity Factor
Soda/Lime Lead
8.1
0.073
0.012
0.034
5.7xlO~6
2.3x10-3
5.0x10-3
29.9 4.6
0.18
0.030
0.084 0.041
1.2x10";?
5. 5x10" J
0.013 2.2x10-3
0.89
0.0076
0.176
0.123
(a) Other severity factors including those for tin particulates and ammonia were quite low.
(b) These concentrations are not air quality standards but were developed from threshold limit values and were used
in testing source severity.
-------
10
Future Growth
The pressed and blown glass industry produces a diverse and
always changing spectrum of glass products. Shipments are expected to
increase at a rate of 3 to 4 percent per year, and emissions would be
expected to increase proportionally. All-electric furnaces have become
economically attractive for melting some glasses and these furnaces are
virtually pollution free. A continued trend in this direction could offset
emissions due to growth of the industry. On the other hand, the general
>
unavailability of natural gas has resulted in increased use of oil. The
actual effect of conversion from gas to oil firing on emission rates is
not known, but would be expected to increase emissions, all other operating
parameters being the same.
-------
11
SECTION III
DESCRIPTION OF THE PRESSED AND
BLOWN GLASSWARE INDUSTRY
This section describes in general terms the process steps used in
the manufacture of pressed and blown glassware and presents certain statistical
information pertinent to the pressed and blown glassware industry described
by the Department of Commerce Standard Industrial Classification (SIC) 3229.
Textile glass fibers, which are a part of this classification, have been excluded
from this report.
General Process Description
The pressed and blown glassware industry, as represented by SIC
3229, essentially comprises all industrial establishments primarily engaged
in manufacturing glassware which is pressed, blown or shaped from glass
produced in the same establishment. It consists of every type of glass or
glassware except flat glass (SIC 3211) and glass containers (SIC 3221).
Establishments include those manufacturing: textile glass fibers; lighting,
electronic, and technical ware; machine made and handmade table, kitchen and
art-ware glass products. Textile fibers which are part of the Department of
Commerce classification SIC 3229 are excluded from discussion in this report.
Figure 1 is a process-flow diagram which generally depicts the flow
of materials through the glass manufacturing process. It can generally be
categorized into four steps: batch handling, melting and fining, forming,
and postforming. These steps are discussed in detail later in this section.
Basically, the manufacture of glass products entails the melting of
a mixture of raw materials which has been prepared in the batch handling step,
so as to minimize segregation and impurities in the batch. Gullet (scrap glass)
is also added in this step. In the glass melter, materials are melted,
molten glass fined (residual trapped gases removed), and the temperature of
the glass lowered so that it can be handled in the forming operation. Glass
passes from the melter to the forming equipment via the forehearth, a relatively
shallow, narrow refractory channel having a refractory roof and individual
heating and cooling systems for controlling glass temperature. Glass is removed
from the melter either in small lots or continuously, after which the molten glass is
pressed, blown, drawn, or cast into shape, depending upon the product. These formed
-------
12
Glass sand
Si02 S 99-S
to yield Si02
crushed, washed
and screened
to ~ 20-100
mesh
'
Soda as /i
Na2CO,
to yield Na20
-20- 120 mesh
cr granular
Limestone
or Burnt lime
to yield CdO
Usually some
MgO also resists
-20-120 mesh
Fvldnpar
R20 AljOj 6SiGj
to yield
AijOj.&O,
Na,0 and K20
pulverized or
granular
Other additions
for K/>. MgO,
ZnO. Bad. PbO.
etc and those for
fining, oxidizing.
coloring, and-
decolorizing
Side-port
continuous tank.
looking down
through top
Submerged
throat in
bridgewall
Cooling
Temperature 1300 "C
Distributing
Packing, warehousing.
and shipping
FIGURE 1. PROCESS FLOW DIAGRAM (2)
-------
13
glass products may then go through a variety of post-forming steps, one of
which is annealing (the removal of residual stresses through a controlled,
uniform cooling cycle). Other post-forming steps include surface treatment,
decoration, firepolishing, etching, cutting, and sealing.
Plants and Locations
According to information gathered from the Department of Commerce
/2\
1972 Census of Manufacturers, there were 158 establishments manufactured
$100,000 or more of pressed and blown glassware shipments in 1972. By 1975,
the number of establishments had increased by 163. These 163 plants were
operated by 110 manufacturers. The top 50 companies produced nearly 98 percent
of the total of the pressed and blown glassware shipped. Approximately 44 plants
produced handmade pressed and blown glassware almost exclusively.
The industry is concentrated in or about the North Central region of
the United States, primarily New York, Pennsylvania, West Virginia, Ohio,
Indiana, and Illinois. Additionally, plants are located in 22 other states,
as shown in Appendix A.
Shipment Value and Volume
Table 4 gives estimated 1974 output data for the pressed and blown
glass industry. As can be seen, each of the three product types listed comprises
a significant portion of this industry. Handmade glassware utilizes significantly
different manufacturing methods and is listed separately.
The value of shipments for 1974 was estimated to be $1307 million as
compared with $1108 million in 1971, or an increase of 18 percent. Shipments
from the many industry segments were difficult to estimate because of the many
different product types and methods of reporting. The estimated 1974 output,
approximately 2.7 Tg (2197 million tons) of glass, was melted to produce industry
shipments of 1.3 Tg (1.4 million tons) of glass. Thus, about 1.4 Tg of glass was
recycled as cullet.
Glass Compositional Types
Most commercial glasses (> 90%) are composed of SiO- as the major in-
gredient (> 55%) with other inorganic oxides added to achieve specific modifications
-------
14
TABLE 4. PROPORTION OF INDUSTRY OUTPUT ACCOUNTED FOR
BY THE CONSUMER, SCIENTIFIC, TECHNICAL, AND
INDUSTRIAL GLASSWARE SEGMENTS OF SIC 3229
Process and
Major Products
SIC
Percent of 1974
Shipments Value
Shipments Volume
Table, Kitchen,
and Art Ware
Machine Made (32291)
Hand Made
Lighting and (32292)
Electronic
Scientific and (32294)
Industrial
Total Industry (Percent)
(Actual)
35.0
6.5
30.4
28.1
100
$1,306,529,000
57
3
23
17
100
1.46 Tg (1,613,600 Tons)'
(a) Shipments volume estimated on the basis of shipment
values of $500; $1,800; $1,075; and $1300; respectively,
for the three SIC categories.
Source: Current Industrial Reports, MA-32E for
shipment value in SIC 32291, 32292, and 32294.
-------
in glass properties. The addition of alkali oxides (especially Na2
-------
16
Borosilicate Glasses
Borosilicate glass batches are also predominantly sand (silica) but
boric oxide replaces much of the alkali content which is characteristic of soda
lime-silica batches. The calcium plus magnesium oxide content is very low. A
few percent of alumina and zero to a few percent of barium oxide are included.
A higher melting temperature is required for borosilicate glasses than for soda-
lime and volatility from the melt for borosilicate glasses is much higher. The
compositional ranges of typical commercial borosilicates are: 70-82 percent silica
(3)
2-7.5 percent magnesia plus calcia, and 0-2.5 percent baria.
The borosilicate glasses have excellent chemical durability and electrical
properties and their low thermal expansion yields a product having high resistance
to thermal shock. These combined properties make them ideal for demanding industrii
and domestic use such as chemical laboratory ware, cookware, pharmaceutical ware,
and for some lens reflectors and lamp envelopes. Pyrex , produced by Corning Glass
Works, and Kimax , produced by the Kimble Division of Owens Illinois, Inc., are
examples of products made from borosilicate glasses.
Lead Glasses
These glasses are composed basically of silica and lead oxide. In additi
most contain significant amounts of alkali oxide. The compositional range of typic
commercial lead glasses is: 35-70 percent silica, 12-60 percent lead oxide, 4-8
percent sodium oxide, 5-10 percent potassium oxide, and 0.5-2.0 percent alumina.
The lead glasses are characterized by high electrical resistivity, high
refractive index and slow rate of increase in viscosity with decrease in temperatur
This viscosity characteristic makes them particularly well suited to hand fabricati
Lead glasses are used in high-quality art glass and tableware; for special
electrical applications; optical glasses; fluorescent lamp envelopes; and X-ray,
gamma-ray, and neutron radiation shielding windows.
Opal Glasses
Opal glasses are translucent and may be colored. Commercial products of
opal glass include lighting globes, ointment jars, dinner ware, and wall paneling.
The batch composition of common commercial opal glasses is basically similar to
soda-lime glass but with modifications and additions. The alumina content is higher
lime lower, and opacifiers are added such as fluorides or phosphates plus other ntt
ingredients.
-------
17
The translucency or opacity of opal glasses is produced by multiple
scattering of light inside the glass. This scattering is achieved by the
precipitation of crystals (or an immiscible amorphous phase) with an index of
refraction different from that of the base glass. The degree of opacity is
determined by the difference in refractive index between the base glass and
dispersed crystal and by the number and size of the crystals. The amount of
the opacifying phase is a minor percent of the total glass. Time-temperature
relations for the forming, cooling, or heat-treating of opal glasses are critical
because they determine the number and size of the dispersed phase and the resulting
degree of ©pacification.
Commercial opal glasses commonly employ fluorine additions to yeild
opacifying crystals of sodium or calcium fluoride. Typical commercial compositions
(4)
are:
Glass Jar Illumination Glass
Ingredient Weight Percent Weight Percent
Si02 71.2 59.0,
A1203 7.3 8.9
CaO 4.8 4.6
MgO 2.0
Na20 12.2 7.5
K20 2.0
F2 4.2 5.0
ZnO 12.0
PbO 3.0
Process Details
Figure 2 is a process flow diagram which generally depicts the flow
of materials through plants producing glassware in the pressed and blown glass
industry. All products produced in this industry undergo generally similar
batch handling and melting and fining procedures, but the forming and post-forming
operations differ widely in the typical operations listed for each five-digit
SIC code indicate. Glass produced within SIC's 32292 and 32294 (Lighting,
Electronic and Technical Ware) typically undergoes a variety of post-forming
operations. Additional information about the processes shown in Figure 2 is
given in subsequent discussions of the processing steps.
-------
W Batch Handling-
>+4-Melting & Fining fr-f^-Forming-
Raw Materials
Weighing Mixing
{Crushed Gullet }
Batch
Charging
Melting
Fining
Feeding
Conditioning
Post Formii
SIC 32291, Table Kitchen & Art Ware
SIC 32292 & SIC 32294
Lighting, Electronic,
& Technical Ware
Press
Blow
Press/blow
Draw
Surface Treatment
Cast
Anneal
Cut
Grind/Polish
Seal
Fire Polish
Tech/Leach
Product Handling-l-*
Inspect
Test
Pack
Store
Ship
1
00
FIGURE 2. PROCESS FLOW DIAGRAM: SIC 3229, PRESSED-AND-BLOWN GLASS (EXCLUDING TEXTILE FIBERS)
-------
19
Batch Handling
The function of the batch-handling operation is to prepare and feed
to the melting furnace a batch which is both chemically and physically uniform
in composition. Control of the composition, impurity level, size, and moisture
content of the raw materials is important. Gullet (scrap glass) collected from
the plant or sometimes purchased, is added in varying amounts, usually between
10 and 50 percent, to the batch. ' The quantity added depends primarily on
its availability. In some processes, large amounts of cullet are produced
(e.g., manufacture of lamps). Cullet is crushed and either mixed with the raw
materials or added later. Each of the raw materials is carefully weighed,
mixed together, and conveyed to the batch chargers. Care must be taken so that
segregation of a uniformly mixed batch does not occur.
A large plant operating a continuous machine forming process, may utilize
a highly automated process for raw material mixing and conveying housed in a
structure termed a "batch house". A flow diagram of a typical batch house is
shown in Figure 3. In most (> 80%) batch houses, the storage bins are located
on top, with the weigh hoppers and mixers located below them to make use of
gravity flow. Major raw materials and cullet (broken scrap glass) are conveyed
from railroad hopper cars or hopper trucks by a combination of screw conveyors,
belt conveyors, and bucket elevators, or by pneumatic conveyors to the elevated
storage bins. Minor ingredients are usually delivered to the plant in paper bags
or cardboard drums and transferred by hand to small bins. Materials are gravity
fed from the storage bins into weigh hoppers and then transported by transverse
belts or bucket elevators into a mixer. Cullet is crushed to a desired size
(usually between 0.5 and 2.0 cm).
After mixing, the glass batch is transferred no a charging bin located
next to the glass-melting furnace or into a batch-storage bin, depending upon
the design of the batch-handling system. Positive displacement or vibratory feeders
at the bottom of the bins feed the materials to the glass-melting furnace chargers.
Cullet may be added to the batch in the mixer, while the batch is being transferred,
(4)
or charged separately to the melting furnace. Batch is fed into the melter in
either a dry or moist state. Many companies add two to four percent water to the
dry batch to help prevent segregation during transport of the batch, to minimize
dust problems in the melter, and to avoid carryover of dust into the regenerators.
The various handling and mixing operations are a source of particulate
emissions which are similar (same materials, same processes) are those in other
industries. Because of environmental and economic incentives, essentially
-------
20
CUUET
RAW MATERIALS
RECEIVING
HOPPER
V
SCREW
CONVEYOR
FILTER
VENTS
STORAGE BINS
MAJOR RAW MATERIALS
MINOR
INGREDIENT
STORAGE
BINS
BELT CONVEYOR
(I
BATCH
STORAGE
BIN
r
FURNACE
FEEDER
GU
Kl
FIGURE 3. PROCESS FLOW DIAGRAM OF A TYPICAL BATCH PLANT
-------
21
all large manufacturers practice dust control, usually by means of cloth
filters and baghouses. '
In the case of batch mixing and charging of clay pots or clay tanks
used in manufacture of handmade glassware, the batch handling may be manual
rather than automated.
Batch Composition
For the soda-lime-silica glasses, which represent the majority of
pressed and blown tonnage, the basic batch ingredients are sand (silica), soda
ash Csodium carbonate), limestone (calcium carbonate) and feldspar (a silicous
mineral used as a source of alumina and alkali). Additionally, the batch will
contain minor ingredients which promote fining, act as decolorizer or a colorant
or impart other specific properties.
Borosilicate glass batches are also predominantly sand (silica) but
boric oxide replaces much of the alkali content which is characteristic of soda
lime-silica batches. The calcium plus magnesium oxide content is very low. A
few percent of alumina and zero to a few percent of barium oxide are included.
A higher melting temperature is required for borosilicate glasses than for
soda-lime and volatility from the melt for borosilicate glasses is much higher.
In lead-alkali silicate glass batches, the lead oxide essentially re-
places the lime of soda-lime glass; for the higher than soda-lime, but less than
for borosilicate glasses.
The batch composition of common commercial opal glasses is basically
similar to soda-lime glass but with modifications and additions. The alumina con-
tent is higher, lime lower and opacifiers are added such as fluorides or
phosphates plus other minor ingredients.
Gullet (scrap glass) collected from the plant or sometimes purchased,
is added in varying amounts, usually between 10 and 50 percent, to the batch. '
The quantity added depends primarily on its availability. In some processes,
large amounts of cullet are produced (e.g., manufacture of lamps).
Melting and Fining
The melting, fining, and conditioning of the molten glass is done in
three separate ways according to the amount of glass required. Continuous furnaces
are standard for the machine-pressed and blown, tubing, television tube, and
-------
22
incandescent lamp glass subcategories. Clay pots and day tanks are used in
the manufacture of hand-made ware. Continuous furnaces range in holding capacity
(14)
from 1 to 500 tons, and outputs may be as high as 300 tons/day. In general,
more than 80 percent of the glass is melted in continuous regenerative furnaces
which use preheated combustion air. Additionally, there are a number of fossil-
fuel fired furnaces where the combustion air is not preheated, and some all-
electric melters.
All furnaces which preheat the combustion air burn fossil fuels and
some utilize additional energy input from electric "boosting". The furnaces
in which the combustion air is preheated are generally classified as end-port
or side-port regenerative. In the pressed and blown segment of the glass in-
dustry, practically every type of furnace is used. These are:
Side-port regenerative
End-port regenerative
Unit melters
All-Electric
Electrically boosted
Recuperative
Day tanks/pot melters.
The type of furnace installed in each plant is dictated by such factors as local
fuel cost and availability (fossil fuel versus electric), market size, plant floor
space, or product volume desired.
Side-Port Regenerative. These furnaces utilize a design similar to that
shown in Figure 4 which illustrates common features. Regenerators (refractory
\
brick checkerworks) are attached to the furnace ports and used to preheat the in-
coming air, which is mixed with natural gas or oil as it enters the melting chamber
The regenerators are about two stories high and positioned on both sides of the
furnace. The number of ports on a side varies from 3 to 7, depending upon furnace
size. Batch enters the furnace where it is melted, fined (entrapped bubbles re-
moved), and homogenized as it moves to the refractory-lined throat, where it
passes into a conditioning chamber, popularly called the refiner.
-------
23
RAW f.U7£P'A.S
J (W«ighirtSC
-------
24
End-Port Regenerative. The end-port furnace is also common furnace
design. It has only two ports, located at the feed-end of the furnace. A flame
is formed as the fossil-fuel/air mixture leaves one port and the combustion
products travel in a horseshoe path over the molten glass until they exit
through the second port. End-ports furnaces are usually smaller in size than
sideport furnaces. However, considerable overlap in size does occur. While an
exact estimate is not available, the combination of side-port and end-port re-
generative furnaces account for the production of 75 percent or more of the glass
melted by this industry.
Unit Melters. The unit melter is a non-regenerative, fossil-fuel fired
melter. They are normally long and narrow and have a relatively low output (less
than 100 tons per day). Their length to width ratio varies from 5:1 to 4:1 and
they normally have 40 percent more surface area per ton of glass than a regenerative
furnace.
All-Electric Melters. In all-electric melters, the glass is heated by its
own self-resistance as an electric current passes through it. All-electric melters
currently melt less than 10 percent of the glass in the United States. Because
the energy is supplied internally to the glass, a higher percentage of the electrical
energy can be converted into usable heat to melt the glass than with fossil-fuel
fired melters. The melter is virtually free from any pollutants. Experience with
larger melting furnaces (> 150 ton/day) does not exist.
Electrically Boosted Regenerative Melters. Many fossil-fuel fired furnaces
are electrically boosted in order to obtain increased production or to reduce
particulate emissions. Usually, 5 to 10 percent of the total energy input to the
melter is supplied via electric boosting, although amounts up to 40 percent may be
added when emission control is of primary concern. Since boosting can be added
while the melter is operating and used only when needed, it is the most popular
way of increasing the output of an existing furnace.
Day Tanks/Pot Furnaces. Pots and clay tanks are employed for the variable
composition and small quantities of glass required in plants manufacturing handmade
glass. The multi-pot furnace is the primary method of melting in these plants.
Eight or more pots may be grouped in a circular arrangement as part of one furnace.
Temperatures as high as 1600 C may be achieved. Pot capacities range from 9 kg to
1800 kg. A day tank is a single furnace and is somewhat larger than a pot, generally
having a capacity of 900 to 3600 kg. Both pots and day tanks are batch fed at the
end of the working day and allowed to melt overnight.
-------
25
Forming
Several methods are used to form glass products in the pressed and
blown industry. They include blowing, pressing, drawing, and casting.
Blowing. The individual section (I.S.) forming machine is used for
making certain types of table and kitchenware. The molten glass is cut into gobs
by a set of shear blades as the glass leaves the forehearth of the melting tank.
Chutes direct the gobs into blank molds. The shear blade and chutes are lub-
ricated and cooled with a spray of emulsified oil or a silicone-based solution.
The molten glass gob is settled with compressed air, and preformed with a
counter blow of compressed air. The preformed gob (parison) is then inverted
and transferred into a blow mold where the glass product is finished by final
blowing.
Incandescent lamp glass envelopes are formed using a ribbon machine.
The ribbon machine employs modified blowing techniques to form the envelopes.
The molten glass is discharged from the melting tank in a continuous stream
and passes between two water cooled rollers. One roller is smooth while the
other has a circular depression. The ribbon produced by the rollers is then
directed horizontally onto a plate belt which runs at the same speed as the forming
rollers. Each plate on the plate belt has an opening and the pill-shaped glass
portion of the ribbon sags through the openings from the action of gravity. The
glass ribbon is met by a continuous belt of blow heads that aid the sag of the
glass by properly timed compressed air impulses premolding the glass. After
the glass has been premolded, it is enclosed by blow molds which are brought
up under the premolded glass on a continuous belt. The blow molds rotate about
their own axis to produce a seamless smooth surfaces. Both the blow heads and
molds are lubricated with a spray of emulsified oil. The formed envelopes (bulbs)
are separated from the ribbon by scribing the neck of the bulb and tapping the bulb
against a metal bar. Residual glass is collected as cullet. Figure 5 illustrates
the ribbon forming machine.
Hand-blown glassware is made using a blowpipe. Molten glass is gathered
on the end of the blowpipe and, using lung power or compressed air, is blown into
its final shape. After the main section is formed, additional parts such as
handles and stems can be added by gathering a piece of molten glass, joining it
to the molded piece and then forming the joined pieces with special glassworking
tools.
-------
26
Water-cooled
> rotten
)-, Mold closing cam
^ i ^' i ..Stripper
IT t7 tJ "w? ~ C w"'~ "Z* ~ ~ ~ ^ ' ii y
S'S ± S ~ r ^^S S S.'S S &*£=
T
Mold opening cam Glass bulbs' Ware conveyor
FIGURE 5. THE CORNING RIBBON MACHINE
(2)
-------
27
Pressing. Over a fifth of the glassware is manufactured using presses.
A press mold consists of three sections: the mold bottom, the plunger, and an
enclosing ring that seals the mold between the mold bottom and the plunger.
Pressing is done manually in the handmade subcategory or by machine in the
remainder of the industry.
In manual pressing of glassware, molten glass is collected on a steel
rod and allowed to drop into the mold bottom. When the proper amount of glass
is in the mold, the glass remaining on the rod is separated from that in the mold
by cutting with a pair of shears. The plunger is then forced into the mold with
sufficient pressure to fill the mold cavities. The glass is allowed to set up
before the plunger is withdrawn and the pressed glass removed from the mold.
Machine pressing is done on a circular steel table. The glass is fed
to the presses in gobs from a refractory bowl at the end of the forehearth of the
melting tank. The molten glass is cut into gobs by oil-lubricated shear cutters
beneath the orifice of the refractory bowl. The motions of the shear blades and
the press table are synchronized such that the gobs fall into molds on the press
table. After the gob is received in the mold, it moves to the next station on
the press table, where it is pressed by a plunger. In the remaining stations,
the pressed glass is allowed to cool before it is removed from the press and
conveyed to the annealing lehr. The mold bottoms are usually cooled by air jets
and the plunger sections are cooled with internally circulated water. The mold
temperature is critical and dependent upon the type of glass being made. If the
mold is too hot, the molded piece will stick to the mold and if it is too cold,
the piece may have an uneven surface. In some cases, the mold is sprayed with
water and lubricants prior to receiving the glass. The steam formed when the
molten glass is introduced helps prevent sticking. Machine pressed glass products
include tableware, lenses, reflectors, and television picture tube faceplates.
Drawing. Glass tubing may be formed using one of three different
processes. In the Danner process, a regulated amount of glass falls upon the
surface of a rotating mandrel which is inclined to the horizontal. Air is blown con-
tinuously blown through the center of the mandrel to maintain the bore and the
diameter of the tubing as it is drawn away from the mandrel. The tubing is pulled
away from the mandrel on rollers by the gripping action of an endless chain.
Tubing dimensions are controlled by the drawing speed and the quantity of air
blown through the center of the mandrel. The tubing is scribed by a cutting stone
that is accelerated to the drawing speed and pressed vertically against the
tubing and which is then cut by bending against a spring controlled roller.
-------
28
In the Velio process, the molten glass passes downward through the annular
space between a vertical mandrel and a refractory or steel ring set in the bottom
of a special forehearth section of the melting tank. The tubing is drawn away
from the melting tank by a Velio machine and cut in a manner similar to that used
for the Banner process.
The updraw process is used to make large diameter tubing and glass pipe.
In the Updraw process, the tubing is drawn upward from a refractory cone. Air is
blown up through the cone to control dimensions and cool the tubing. The tubing
is cut into lengths at the top of the draw.
Casting. Television picture tube envelope funnels are normally formed
by centrifugal casting, although occasionally they are pressed. Molten glass is
cut into gobs by oil-lubricated shear blades, and the glass gob dropped into the
mold. The mold is spun rapidly so that the centrifugal force cause& the glass
to flow up ;the sides of the mold to form a wall of uniform thickness. Newer
processes seal against a ring, much like pressed glass, while older methods entail
cutting off the upper edge of ,the funnel.
Post-Forming and Finishing
This part of production can consist of many operations depending upon
the particular product be .ng manufactured. Thy include: surface treatments,
annealing, decorating or engraving, cutting, scalihg, polishing, etching, aid c.oatisj
,'
Prior to annealing and sometimes after, glassware may receive surface tveatjien s
to improve its chemical resistance or improve abrasion resis';anca. Man/ ptodui.its
are fire-polished, which involves passing newly, formed ware :hrough a line of
oxygen-gas burners that are directed onto the ware to smooth out ric.ges or edg.is.
After fire-polishing, the ware goes into a lehr for a normal annealing cycle.
After the glass is formed, annealing is usually required to relieve
strains that might weaken the glass. The entire piece of glassware is brought to
a uniform, elevate temperature (590 to 650 C) to permit the release of internal
stresses and then it is cooled at a uniform rate to prevent new stresses from
developing. Annealing is done in long continuous ovens called lehrs. Heat
treating to allow a portion of the glass to crystallize may also be done in lehrs.
After heat treatment, some ware may be decorated with enamel colors applied to
annealed articles and then refired in annealing lehrs.
-------
29
Television picture tube envelopes are manufactured in two pieces,
referred to as the screen and funnel. Both pieces require the addition of
components prior to annealing and several finishing steps follow annealing.
After forming and prior to annealing, the seam (shearmark) on the screen is
fire polished and mounting pins are installed using heat. The counting pins
are required for proper alignment when the electronic components are placed
into the picture tube. The stem portion and an anode to be used as a high
voltage source are added to the funnel prior to annealing by fusion into the
funnel. Following annealing, screens and funnels are visually inspected for
gross defects such 'as large stones, blisters, and entrapped gas bubbles. The
screen dimensions and mounting pin locations are then gaged to check for
exactness of assembly- The funnel portion is not gaged until all finishing
steps are completed.
Screens and funnels are finished separately using different equip-
ment. The first finishing step applied to the television screen section is
abrasive polishing. Polishing is required to assure a flawless and parallel
surface alignment so that an undistorted picture will be produced when the
tube is assembled. The edge of both the screen and funnel must be perfectly
smooth so that a seal will be formed when the two sections are glued together.
The seal must be sufficiently tight to hold a vacuum. Abrasive polishing is
accomplished in four steps using rough and smooth garnet, pumice, and rouge or
cerium oxide. The abrasive compounds are in a slurry form and are applied to
the screen surface by circular polishing wheels of varying texture. Between each
polishing step the screen is rinsed with water. The slurry solutions are generally
recycled through hydroclones or settling tanks and only fine material too small
to be useful for grinding or polishing is wasted. Following abrasive polishing,
the screen edge is ground, bevealed, and rinsed with water. This edge is then
dipped in a hydroflouric acid solution to polish and remove surface irregularities.
This step is commonly referred to as fortification. Following rinsing, to remove
residual acid, and drying, the screen receives a final inspection. The front edge
of the funnel is polished with a diamond wheel and fortified as previously
described. The polishing surface is bathed in oil and, therefore, the funnel must
be rinsed with water and dried before final gaging and inspection.
-------
30
Incandescent lamp envelopes are generally frosted. After annealing,
the envelope interior is sprayed successively with several frosting solutions.
The specific formulation of these solutions is proprietary, but primary con-
stituents include hydroflouric acid and other fluoride compounds, ammonia, water,
and soda ash. Residual frosting solution is removed in several rinse stages.
The manufacture of hand pressed and blown glass also involves several
finishing steps including: crack-off, washing, grinding and polishing, cutting,
acid polishing, and acid etching. The extent to which these methods are employed
varies substantially from plant to plant. Many plants use only a few of the
finishing methods, of which washing, grinding, and polishing are the most
prevalent.
Crack-off is required to remove excess glass that is left over from
the forming of hand-blown glassware. Crack-off can be done manually or by machine.
When a machine is used for stemware, for example, the stemware is inserted into
the crack-off machine in an inverted position. The bowl of the stemware is scribed
by a sharp edge, the scribed edge passes by several gas flames and excess glass is
broken off. The scribed surface is then beveled on a circular grinding medium
similar to sandpaper. Carborundum sheets are used in most cases. The grinding
surface is sprayed with water for lubrication and to flush away glass and
abrasive particles. Hydrofluoric acid polishing of the beveled edge may follow
crack-off and is considered part of the crack-off operation in this study. This
operation involves rinsing the glassware in dilute hydroflouric acid and city
water, and in some cases, a final deionized water rinse.
Cutting as applied to handmade glassware manufacturing may be defined
as the grinding of designs into the glassware or as the removal of excess glass
left over from forming. Designs may be placed onto the glassware manually or by
machine. In mechanical design cutting, the ware is placed on a cutting machine
and is rotated in a circular motion. Designs are cut into the surface at the
desired points using a cutting edge. In the other form of cutting a saw may be
used to remove excess glass from some handmade products. Water is used in both
machine design cutting and sawing to lubricate the cutting surface and to remove
cutting residue.
Acid polishing may be employed to improve the appearance or to remove the
rough edges from glassware. Automatic machines or manually dipped racks may be
employed. In the manual operation, the glassware is placed in racks and treated
with one or more hydrofluoric acid dips followed by rinsing. Tha complexity and
number of steps is determined by the product.
-------
31
Abrasive polishing is used to polish the glass surfaces and edges on
some types of handmade glassware. The glassware is placed in a bath of abrasive
slurry and brushed by circular mechanical brushes or polishing belts. After
polishing, the ware is rinsed with water in a sink and dried.
Complicated designs may be etched onto handmade stemware with hydro-
fluoric acid. The design is first made on a metal template and is transferred
from the template to a piece of tissue paper by placing a combination of beeswax
and lampblack in the design and then pressing the tissue paper against the design.
The tissue paper is placed on the stemware and then removed leaving the pattern in
wax. All parts of the ware,except for the pattern are then coated with wax. The
wax-coated stemware is placed in racks and immersed in a tank of hydroflouric acid
where the exposed surfaces are etched. Following a rinse to remove residual acid,
the ware is placed in a hot water tank where the wax melts and floats to the sur-
face for skimming and recycling. Several additional washes and rinses are required
to clean the ware and to remove salt deposits from the etched surfaces. In some
cases, a nitric acid bath may be used to dissolve these deposits.- Deionized water
may be used for the final rinse to prevent spotting.
-------
32
SECTION IV
EMISSIONS
Emissions from pressed and blown plants are categorized according
to three operations within the manufacturing process
Particulate emissions from the raw-materials handling,
preparation, and transfer.
Gaseous and particulate emissions from the glass-melting
furnace. These emissions are primarily SOV, NO , submicron
x x
condensates, hydrocarbons, CO, fluorides, borates, and
lead oxides.
Gaseous and particulate emissions from a variety of
forming and postforming operations. These result from
annealing, decorating, surface treatment, and coating
operations and can include particulates, hydrocarbons,
NO , and SO .
X X
This section describes the various emissions, their characteristics, their
levels, total quantities, and environmental effects. The information is
organized according to the three sources within the manufacturing process
Raw-materials preparation and handling
Glass melting
Forming and finishing.
Raw-Materials Preparation and Handling
*
Typical points of particulate emissions during raw-materials preparation
and handling are shown in Figure 6. These points include
Unloading and conveying
Crushing of cullet (scrap glass)
-------
33
Conveying to storage
bins by
Acme
Minerals
Screw conveyors
Belt conveyors
* Bucket elevators
Pneumatic conveyors
Gravity-fed into
weight hoppers
Transported by
Transverse belts
Bucket elevators
Transferred to
charging bin
Batch charging
Transferred
to charging
bin
Gullet is crushed
Unloading of cutlet
t
Transferred to storage bin
f
Storage bin
i
Transferred to grinder
FIGURE 6. TYPICAL POINTS OF PARTICIPATE EMISSION
FROM RAW-MATERIALS HANDLING
-------
34
Filling and emptying of storage bins
Weighing and mixing of batches
Feeding of batch to glass furnace (batch charging).
However, those uncontrolled particulates which remain in the manufacturing plant
may constitute an OSHA health and safety problem distinct from plant emissions.
For the purposes of this study, fugitive-dust emission has been defined
as particulate emissions that result from industrial-related operations, and
which escape to the atmosphere through windows, doors, vents, etc., but not through
a primary exhaust system, such as a stack flue, or control system. This definition
is derived in part from a paper presented by Lillis and Young of the U. S. En-
vironmental Protection Agency.' Information obtained from ambient sampling
up-wind and down-wind of the manufacturing facility is the preferred source of
data. If actual data from high-volume samples are not available, engineering
estimates based on the particle size of raw material which can contribute to
dust emissions are desired. For the purposes of this study, the particle size
range of the raw materials which will be considered as contributing to fugitive
dust emissions is 100 microns or less in diameter. This definition of dust is
not as broad as the technical definition given by Stern,(°' but this particle-
size distribution seems reasonable for glass-manufacturing processes. The
settling velocity of a 100-micron-diameter sphere, with a specific gravity of
2.0 g/cc, is approximately 50 cm/sec in still air at 25 C and 1 atm. Such a
settling rate is sufficiently slow that the emission of dust from a tall source,
such as the raw materials storage bins, would probably contribute to the total
air emissions.
Dust is usually emitted during unloading and conveying operations.
To minimize dust emissions, these operations are generally enclosed and the
vents on storage bins and mixers exhausted through fabric filters. Batch
wetting, the addition of water to the batch during the mixing operation, is
another practice commonly used to minimize particulate emissions. As a result,
limited data on particulate emissions from primary exhaust systems are available,
and no data are available on fugitive dust. Particulate-emission data from
point-source measurements have been reported in NEDS^ ' (National Emission Data
System) and are given in Appendix B. Although these particulate emission data
-------
35
/*
were not listed by specific ingredients, the data enable the calculation of overall
average emission factors for raw-materials handling and preparation. This over-
all emission rate is determined to be 1910 mg/kg ±100 percent. Total annual
particulate emissions for raw-materials handling and manufacturing are 6.44 mg
±100 percent, based on 3.37 Tg of raw materials being processed to produce 2.97
Tg of glass, and is equivalent to 0.0004 percent of the national particulate
emissions from stationary sources. Table 5 shows a breakdown of raw-materials
handling for the various points of emission. This listing has been determined
/Q\
primarily by NEDS datav ' and confirmed by observations made during plant visits.
The ingredients contained in these particulate emissions reflect the raw material
used in the manufacturing process (soda ash, limestone, feldspar, quartz, PbO,
borates, and fluorides), since no chemical reaction have taken place. Quantitative
data on the amount of each ingredient emitted are not available.
Pressed and blown glass manufacturers minimize the dusting problem by
limiting the amounts of fine particles (<100 micron) in the batch material.
Manufacturers generally specify particulate sizes ranging from 820 to 44 micron
(-20, +325). Table 6 shows the specification limits for several of the raw
materials used in the manufacture of pressed and blown glassware. Since most
of the materials have specified particle size limits greater than 100 microns
(150 mesh), the amount of material emitted from the plant site due to inertial
forces is minimal. Note also that quartz particles in the mix are generally
larger than 100 microns in diameter and, as such, would not be expected to be
emitted as respirable quartz either through the stacks or as fugitive dust.
Based on information available from raw-materials suppliers, a reasonable
assumption is that <1 percent of the materials used have a particle size
less than 100 microns in diameter (150 mesh). Assuming at least 90 percent
of dust emissions are captured in fabric filters (fabric-filter efficiency
is >98 percent), then approximately 4 mg of fugitive dust would be emitted
annually (based on 3.37 Tg of raw materials produced annually). These data
appear consistent with both observations from plant visits and information
obtained from pressed and blown glass manufacturers.
-------
36
TABLE 5. PARTICULATE EMISSIONS DURING RAW-MATERIAL PREPA-
RATION AND HANDLING FOR PRESSED-AND-BLOWN GLASS
Total Annual
Emission Factor EmissionsW
Process Step mg/kg Mg
Handling (unloading, 1500 + 100% 5.06
conveying
Glass crushing Negligible
Storage bins 100+100% 0.34
Mixing and weighing 310 + 100% 1.04
Batch charging Negligible
Total 1910 + 100% 6.44
fa)
v Based on 3.37 Tg of raw materials processed to melt 2.97 of glass
(b) <0.1
-------
37
TABLE 6. SPECIFICATION LIMITS FOR SEVERAL RAW MATERIALS
USED IN PRESSED-AND-BLOWN GLASS MANUFACTURE(14>
Material Specifications Range
Mineral
Arsenic Trioxide
Cerium Oxide
Dolomite
Feldspar
Limestone
Sand
Soda Ash
Sodium Nitrate
Chemical
Formula
AS203
CeO,
(Ca,Mg)C03
-
CaC03
Si02
Na2C03
NaN03
Minimum
Amount (%)
9
-
0.5
2
0.2
3
0
1
Mesh(a)
+20
-
+16
+40
+20
+30
+20
+6
Maximum
Amount (%)
14
100
50
10
0.5
6.6
4.2
1.5
Mesh(a)
-325
-60
-100
-200
-300
-100
-120
-100
(a)
See Table 16 for micron equivalents.
-------
38
Glass Melting
In the United States, pressed and blown glass is predominantly
melted in fossil-fuel-fired furnaces. Emissions from these furnaces are
by far the largest source of pollutant from a glass plant. The type of
pollutant emitted depends on the glass composition and the furnace operating
conditions. These emissions will include NO , SO , particulates, CO, hydro-
X X
carbons, selenium, fluorides, borates, and lead compounds.
The overall emission rates and total emissions for furnaces melting
soda lime, borosilicate, opal, and lead glasses are given in Table 7. The
emission factors are based upon data reported in NEDS and derived from
various other sources. Data referred to as source measurements, as reported in
NEDS, have been obtained by actual point source test measurements. The emission
rates are highly dependent upon the operating conditions of the glass-melting
furnace. For instance, emissions are reported to range from trace amounts
to a high of 10 g/kg of glass melted for NO , 5.44 g/kg for SO , and frOffi 0.49
X X
g/kg to 12.57 g/kg for particulates. Each typ'e of emission is discussed in
greater detail for the various types of glasses. As has been shown previously,
(page 15) soda-lime, borosilicate, opal, and lead glass are estimated to com-
prise approximately 77, 11, 7, and 5 percent, respectively, of the glass pro-
duced in the United States.
Nitrogen Oxides
In a fossil-fuel-fired furnace, nitrogen oxides (e.g. NO and NO-) are
formed by a combination of atmospheric nitrogen and oxygen at the elevated
temperatures (> 1500 C) required for making glass. Because of the high tempera-
ture, NO would be expected to be the primary oxide of nitrogen formed. For
purposes of this analysis, nitrogen oxides are designated as NO . In -this study,
X
NO is compared against the &0- air quality standard. The assumption that the
X £-
NO, emission factor is equal to the NO emission factor is believed valid be-
fm X
cause once the plume has been diluted sufficiently with air (dispersion calcu-
lations show that the plume is diluted approximately 1000 to 1 at the point
Trace < 0.001 g/kg of glass melted
-------
TABLE 7. EMISSIONS FROM PRESSED AND BLOWN GLASS MELTING FURNACE OPERATIONS
Species
::o
X
SO
X
Parttcu-
lates
CO
HC
Fluorides
ScK.,iu,
Exis^ion Factor, g/kg
Sods/Liaa Eorasllicate Opal
4.25 + W.
^
2.68 + 50* 2.99 + 100Z
"" *
5.22 + 43Z 25 + 100Z 5 + 100Z
0.10 + 100X
0.15 + 53Z
0 10 + 100% 10 + 100Z
0 0
Loi-d Total Soda/Line
4
2.
15 + 100Z 8.
0.
0.
10 ± 100Z
0.
25 -1-
~
80 +
70 +
10 't-
is +
10 +
002
43% 10.0
62Z 6.3
60% 12.3
10CZ 0.2
53% 0.3
100Z 0
Total Annual Emissions W,
Cg
Bcrosllicate Opal Lead Total Soda/Lima
__ 12.
1.1 8.
9.0 1.1 0.4 25.
. o.
0.
0,7 2.2 0.1(c) 3.
0 0 9
6(t>) 0.086
3 0.010
8 0.009
3 0.002
4 0.001
0
Percent of National Emission*
from all Stationary Sources
Borosilicate Opal Lead Total
0.109
0.002 0.013
0.007 0.001 0.020
0.002
0.001
(a) 3ued on 2360, 360, 220, and 30 Gg of gliss produced annually for soda/lime, boroslllcate, opal, and
(b) Assuicas equivalent emission factor for other glass types.
(c) Not found in all glass melted.
(d) Standard for NO,
lead glass.
-------
40
where it touches the ground), the photochemical conversion of NO to N02 is
quite rapid. This report does not attempt to determine the relative pro-
portions of each gas.
Nitrogen oxides.represent the second largest mass fraction (^21%)
of emissions from the glass-melting furnace. As seen in Figure 7, the formation
of NO in a glass-melting furnace is extremely temperature sensitive. In one
JV
case, NO concentration has been observed to increase some six times (from
x
100 ppm to ^600 ppm) as the furnace temperature (measured at the bridgewall)
increased from 1460* to 1550 C and the production rate of a soda lime glass
doubled(1°). The rate of NO formation depends upon factors such as peak
A
flame temperature, percent excess oxygen, and post-time/temperature history
of the flame. Consequently, considerable variation in the rate of NO emissions
X
can and does occur.
Soda/Lime Glass. Source measurements reported in NEDS'*' and
taken from the open literature give an average emission rate of 4.25 g of
NO per kg of soda lime glass produced. This average is based on 14 measure-
X
ments (see Appendix B) and is calculated to be accurate to within ±43 percent
at a 95 percent confidence level. Individual values range from 0.41 to 10.0
g/kg. The average NOX emission amounts to approximately 10.0 Gg og NOX emitted
annually from furnaces melting soda lime glass, which is equivalent to ^approxi-
mately 0.086 percent of 1972 National NO. emissions from all stationary sources
(Appendix D).
(Q\
Other Glasses. No point source measurements were reported in NEDS*1
nor available from the open literature for NO emissions from furnaces melting
jt
borosilicate, opal or lead glasses. The emissions from furnaces melting boro-
silicate glass are expected to be higher than that reported for soda lime
because of the higher melting temperatures required. Furnaces melting opal
and lead glasses would be expected to be equal to or lower than those found
for soda lime, since processing temperatures are similar or lower.
The maximum NOX emissions expected for borosilicate glass is estimated
to be three times that observed for soda lime glass or about 13 g/kg. This
emission rate will produce total annual emissions of 4Gg or 0.035 percent of the
1972 National NO- emissions from all stationary sources .
-------
600
I500h
w 400
m
c
0)
oo
o
1-.
300
o
-------
42
Sulfur Oxides
Sulfur oxide emissions can occur through both the decomposition of
sulfates (e.g., Na2SO,) added in the glass batch and from the oxidation of
sulfur in the fossil fuel." When oil is used as the fossil fuel, a large
majority (> 80 percent) of sulfur oxide emissions are derived from sulfur in
(13)
the oil . Sulfur oxides from the batch generally combine with alkali
volatiles and exit as a participate, while sulfur in the fossil fuel exits
predominately at SO . While .some glasses contain no sulfur (e.g. borosilicates),
it is present in all soda lime glasses. Soda lime glass generally contains
about 0.15 weight percent sulfate (added usually as salt cake, gypsum, or
blast-furnace slag) which is added for melting and finishing purposes and is
a necessary ingredient for making container glass. The range of values for sul-
fate (SO^*) in glass as reported in 1973 varies from 0.03 to 0.32 percent,
with 82 percent of some 106 analyzed glasses falling between 0.10 and 0.20
(13)
percent . The amount of mineral sulfate added in the batch will, of course,
be higher and usually falls within the 0.5 to 1.0 percent range. Sulfur oxide
emissions from the batch materials do occur ' and these depend primarily
upon the quantity of glass melted.
Sulfur oxide emissions will be greatly influenced by any switch from
natural gas (the primary fossil fuel), which is essentially sulfur-free to
fuel oils or powdered coal containing sulfur. Such a trend does exist primarily
because of the reduced availability of natural gas in most sections of the
country. Sulfur in fossil fuels readily oxidizes in the glass-melting furnace
and appears as SO in the exhaust gases. For instance, a fuel oil containing
^& * ^^
one weight percent sulfur emits approximately 600 ppm (calculated as S09) in
the flue gas(13).
An approximate materials balance, which illustrates the dependence
of SO emissions on the fossil fuel and batch materials as follows for a
X
furnace producing approximately 136 Mg/day (150 tons/day) of a soda lime glass
used for light bulbs'
-------
43
Natural Gas No. 5 Fuel Oil
Sulfur Input
Batch 2.5 kg/hr (5.6 Ib/hr) 2.9 kg/hr (6.4 Ib/hr)
Fuel 0.0 kg/hr 5.3 kg/hr (11.6 Ib/hr)
Total 2.5 kg hr (5.6 Ib/hr) 8.2 kg/hr (18.0 Ib/hr)
Sulfur Output
Glass 2.3 kg/hr (5.0 Ib/hr) 2.6 kg/hr (5.7 Ib/hr)
Particulates 0.3 kg/hr (0.6 Ib/hr) , 0.3 kg/hr (0.7 Ib/hr)
SO <0.1 kg/hr (<0.1 Ib/hr) 5.0 kg/hr (11.0 Ib/hr)
Total 2.6 kg/hr (5.6 Ib/hr) 7.9 kg/hr (17.4 Ib/hr)
This example illustrates that the sulfur oxides predominately result from the
fuel used.
Source measurements reported in NEDS or in the literature are essen-
tially for natural gas-fired furnaces. Hence, the emissions are not fully
representative of an industry which is gradually switching to fuels containing
sulfur. However, such future emissions will essentially correspond directly
to the sulfur found in the fuel oil or powdered coal.
Soda/T.tme mass. Only three source measurements are available for SOy
emissions from furnaces producing soda lime glass. These sources gave an average
SOX emission rate of 2.68 g/kg of glass melted. A reasonable way to assign an
accuracy to these values in nonexistant. Individual SO emissions from furnaces
X
melting soda lime glass in the pressed and blown industry is 6.3 Gg, which is
equivalent to 0.086 percent of 1972 National SO emissions from all stationary
sources.
(9)
Borosilicate Glass. Two point source measurements were reported
for SO emissions from furnaces producing borosilicate glass. Since sulfates
were not used as batch materials in borosilicates, these emissions must have been
taken from oil fired furnaces. The individual rates were 0.54 and 5.44 g/kg
for an average of 2.99. If all borosilicate glass melted had this emission
rate, total annual SO emissions would be 1.1 Gg.
Other Glasses. Sulfates are also not used as batch materials in opal
and lead glasses. Emissions of SO from furnaces melting these glasses would
A
only occur from sulfur in the fossil fuel. No information was available for
-------
44
furnaces using fuel oil to produce these glasses. Some aluminosilicate and
opthalmic glasses used sulfate-containing raw materials; however, no infor-
mation was available.
Fluorides
Fluoride emissions can occur from furnaces melting opal, corosilicate,
and lead glasses. These emissions come from batch materials such as fluorspar
(CaF2), K2SiFg, NaSiFg, Lepidolite, and Cryolite (Na3AlFg) . Fluorine acts as
as a flux and the fluoride can remain as a separate phase when the glass
cools, imparting a milky white color to the finished product, (e.g., opal
glass). During melting, a portion of the fluorides in the batch volatizes and
escape as gaseous compounds. Some of these compounds are retained in the
glass and some can also be emitted as particulates. The gaseous compounds
include HF, BF
Total fluoride emissions, either as a gas or particulate, were cal-
culated on a worst-case basis from data reported in the literature ' '
Assuming the worst-case emission rate as 10 g/kg (as F ) of glass melted and
a total annual production of glass containing fluorides as 0.30 Tg, total annual
emissions would be 3.0 Gg of F .
Carbon Monoxide
Carbon monoxide can be emitted through incomplete combustion of the
fossil fuel through the use of a luminous flame, or by reaction of a powdered
coal added to the glass batch to reduce sulfate compounds. Only three emissions
on soda lime glass were available, and they gave an emission rate of 1.10 g/kg,
which is considered to be a worst-case situation. This emission rate is believed
accurate to within ±100 percent at a 95 percent confidence level and represents
an annual emission of 0.29 Gg of CO, or 0.002 percent of 1972 National CO
emissions from all stationary sources. CO emissions would be expected to be
independent of glass type.
-------
45
Hydrocarbons
Hydrocarbon emissions form in glass-melting furnaces primarily through
the incomplete combustion of a fossil fuel. Hydrocarbon source test measurements
are limited. An estimated average emission rate for soda lime glass is 0.15
g/kg, based on seven measurements calculated to be accurate to within ±53 percent.
Such an emission rate represents an annual emission of 0.44 Gg of hydrocarbons
or 0.0012 percent of 1972 National emissions for hydrocarbons from stationary
sources. Hydrocarbon emissions would be expected to be independent of glass
type.
Particulates
Particulates from glass-melting furnaces can originate both from
physical entrainment of batch materials being charged to the melting furnace and
from condensation of compounds, such as sodium sulfate (which forms from sulfur
oxides and volatized sodium) . Particulates exiting with exhaust gases are essen-
tially all (> 95 percent) condensates, as indicated by the solubility of the col-
lected residues in water. Studies do show that batch materials are carried out
of the furnace by the combustion products; however, such materials usually do not
show up in the stack-gas samplings; therefore, these coarser batch materials are
assumed to be retained in the furnace-flue system^,19) t
Considerable opinion exists as to the exact mechanisms by which con-
densate particles are formed. For soda-lime glasses, analyses have shown the
particulates to consist predominately (> 75 percent) of submicron sodium sulfate
(5,10,18)^ The particulates from borosilicate glasses are made up of boric oxide,
alkali borates (e.g., Na-B.O, and NaCl). With lead glass, the particulates con-
sist of lead oxide, sulfate and anhydrite. When the furnace is fired with oil,
the particulates change color from yellow to white, because PbSO^ is emitted.
The particulates can also contain NaF, Na^O^, and Sb203. The particulates from
opal glass contain B-O-, NaF, and Na2SiFg.
Uncontrolled particulate emissions are least for soda lime glasses,
intermediate for lead glasses, and highest for borosilicate glasses. For one
manufacturer' ', uncontrolled particulate emissions for glass melting furnaces
producing 75 to 100 tons/day will normally be approximately 2.3 kg/hr for soda-lime,
-------
46
9.5 kg/hr for lead, and 15 kg/hr for borosilicate glass. Another manufacturer '
reported uncontrolled emissions for lead glasses to be about 15 g/kg and about
25 g/kg for borosilicate glasses.
The formation of particulates depend upon batch composition (type
of glass), temperatures in the melting furnace, production rate, surface area
of molten glass, and cullet ratio. Of these factors, glass composition,
production rate, and temperature of the molten glass are the more important
factors affecting particulate emissions. Since these variables are inter-
related, determination of the relative influence of each variable is difficult
although, for a given composition, temperature appears to be the most sig-
nificant variable. Data from one furnace melting a soda lime glass shows
that at zero production rate (tank soaking), the particulate emissions are
approximately 20 percent of that measured at its normal furnace capacity W)'.
While temperature is maintained at a constant value (1450 C), emissions range
from 1.814 kg/hr (4 Ib/hr) at zero pull to 7.711 kg/hr (17 Ib/hr) at normal
pull of 211 Mg/day. Other data' ' collected on soda lime glass during this
study, indicate that particulate emissions follow an Arrhenius curve when
plotted against the reciprocal of temperature; that is, a linear relationship
with the logarithm of the emission rate. This relationship is shown in
Figure 8. Similar results are found for a borosilicate glass(1*). Figure 9
shows the emission rate as a function of pull rate for soda lime and boro-
silicate glasses. Note that particulate emissions occur even at a zero pull
rate, so long as the temperature of the furnace is maintained.
Particulate emissions taken from glass melting furnaces have been
found to be generally submicron in size. In one study^ , particulate emis-
sions from furnaces melting soda lime glasses averaged 0.13 micron.
Soda Lime Glass. Source measurements for particulate emissions from
soda lime glass-melting furnaces give an average emission rate of 5.22 g/kg.
This emission rate varies from 0.49 g/kg to 12.57 g/kg. Source measurements
are from 19 points and are calculated to be accurate within ±43 percent at a
95 percent confidence level. These emissions represent an estimated total annual
particulate emission of 12.3 Gg or 0.009 percent of the 1972 National particulate
emissions from all stationary sources^ .
-------
100
50
00
.id
M 10
47
§
1-i
CO
CO
H
I
01
2
o.i
I
6.2
6.3
6.4
6.5
6.6
Bridgewall Temperature (1/T (C) x 10"4)
6.7
FIGURE 8. PARIICU1ATE EMISSIONS SHOWN ARE LINEAR WITH
THE RECIPROCAL OF BRIDGEWALL TEMPERATURE
-------
48
1000 -
JS
ft
cu
4J
o
f-l
a]
CO
60
,3
100 _
50 -
10 _
5 -
Borosilicate
Tsridgewall
Soda/Lime
TBridgewall = 143° C
2 3
Feed/Pull Rate (tons/hr)
FIGURE 9. PARTICULATE EMISSIONS FOR GLASSES AS A FUNCTION OF FULL RATE
-------
49
Borosilicate Glass. Source measurements for particulate emissions
from borosilicate glass-melting furnaces were not available. Using a worst-
case, uncontrolled emission rate of 25 g/kg, the total annual particulate emis-
sion was estimated to be 9.0 Gg or 0.007 percent of the 1972 National particulate
emissions from all stationary sources
Opal Glass. Source measurements for particulate emissions from
opal glass-melting furnaces were not available. Using a worst-case, uncontrolled
emission rate of 5 g/kg, the total annual particulate emission would be 1.1 Gg,
or 0.001 percent of the 1972 National particulate emissions from all stationary
sources .
Lead Glass . Only one source measurement was available for particulates
from furnaces melting lead glass (4.52 g/kg). Using a worst-case uncontrolled
emission rate of 15 g/kg, the total annual particulate emissions would be 0.30
Gg, or 0.0002 percent of the 1972 National particulate emissions from all sta-
tionary sources .
Selenium
Selenium is used by glass manufacturers as a decolorizer to neutralize
i
the tint from transition metal oxide contaminants such as iron, and is usually
used in amounts of less than two weight percent ' ' . Source measurements on
selenium emissions are unavailable. Selenium volatizes at rather low temperatures
(315 C for SeO, 685 C for Se) ' , therefore, it can be expected to be present
in the waste gases. If the temperature of the waste gases is below 200 C,
selenium condensates are likely to be found. Approximately 0.359 Gg of
(23)
selenium are consumed annually in the United States , of which an estimated
5 percent or 0.018 Gg is used by the pressed and blown glass industry. Under the worst
case, approximately half of the selenium will be emitted, representing an emission
rate for selenium of 0.002 g/kg. Using these estimates, total annual emissions will
be 0.002 Gg.
-------
50
Other Emissions
Other minor emissions can include antimoney and arsenic, which are
sometimes added as fining and decolorizing agents. The use of both of these
materials has been steadily declining in recent years. Similarly, chlorine
can be emitted because of its association with soda ash produced by the Solvay
process. In recent years, most of the glass industry has switched from syn-
thetically produced soda ash to that manufactured from a naturally occurring
ore which does not contain chlorine. By 1977, more than 90 percent of the
(5 22)
industry will be using natural soda ash '
Forming and Finishing
A wide variety of forming and finishing operations are used within
the pressed and blown glass industry. Molten glass, properly conditioned, leaves
the forehearth of the melting furnace as a single stream or is cut into individual
"gobs" which are then transferred to a forming machine. Glass may be blown,
pressed, rolled, or cast into a shape suitable for additional processing. »After
forming, glass may be surface treated, and sometimes fire-polished, after which
it is passed through an annealing lehr. Once annealed, the glass article may
undergo a variety of decorating, surface treatment, or coating operations.
Little data is available on emissions during the forming and finishing
operations. However, compared to the melting operation thay are considered to
be minor. These emissions can include: hydrocarbons emitted during forming
operations; HC1 and metal oxides emitted during surface-treatment operations;
emissions associated with combustion gases produced during annealing; and
hydrocarbons, lead oxide, HF, and NH3 emitted during the finishing operations.
Estimates that have been made of emissions from some of these operations are
given in Table 8.
Forming
Gob shears, delivery chutes, and the forming molds for pressed and
blown glass are lubricated with various solutions. These solutions can contain
-------
TABLE 8. EMISSIONS FROM THE FORMING AND FINISHING
OPERATIONS FOR ALL PRESSED AND BLOWN GLASS
Emission Factor >
0 0
0 0
(a) &ust!
-------
52
grease, oils, graphite, and silicone-based emulsions. In the past decade, there
has been a transition from grease and oil lubricants to the use of silicone-
emulsions and water-soluble oils (1 part silicone or oil to 90-150 parts water)
on gob shears and gob-delivery systems/ Grease and oils are still utilized
on molds. During forming operations, a visible puff of white smoke is.
formed when the molds are swabbed with a lubricating solution. Although this
smoke dissipates in a few seconds, hydrocarbon vapors are probably released. The
resultant emissions are probably drawn through the large ventilators on the roof
of the plant.
Hydrocarbon emissions from the forming operation are estimated to be 0.06
g/kg, based on three point source measurements for soda lime glass. Data for
other manufacturing lines are not available. Considering this rate as a worst
case for the whole industry, total annual emissions would be 0.18 Gg, or 0.0006
percent of 1972 National emissions from stationary sources.
Treatment
Pressed and blown glassware will occasionally receive a metal oxide
(titanium or tin) surface treatment to improve resistance to scratching. In
addition, this transparent treatment acts as a lubricant which can facilitate
handling and shipping operations. The oxide treatment is obtained by subjecting
the hot article (coming from the forming machine) to a vapor of metal chloride.
This treatment is done within a hood. The metal chloride pyrolyzes to the metal oxide
on the glass surface, leaving a metal-oxide film and releasing hydrogen chloride.
Emissions from the surface-treatment operation will consist of HC1,
metal oxides, and hydrated-metal chlorides. Anhydrous tin chlorides which do
not react with the glass will decompose by the action of heat and moisture
within the exhaust ductwork to form metal oxides, hydrated metal chlorides,
(23 9)
and HC1. Estimations based upon available data * indicate that approxi-
mately 60 percent of the total weight of the metal chloride input is released
into the atmosphere. Using tin tetrachloride as the input material, these
estimations indicate that of the total weight input, 14 percent is released
into the atmosphere as a metal oxide, 27 percent as hydrated tin chloride,
and 21 percent as HC1.
-------
53
Emissions from the surface-treatment operation for glass articles
were determined by engineering calculation to be 0.02 g/kg of tin or titanium
oxide, 0.03 g/kg of hydrated tin or titanium chloride and 0.02 g/kg of HC1.
Total annual emissions were estimated to be 6.9 Mg of metal oxide, 10.4 Mg of
hydrated metal chloride, and 6.9 Mg of HC1. This worst-case estimate was
based on 25 percent of the total melting output.
Annealing
Essentially all pressed and blown glassware undergoes an annealing
operation, during which the glass is brought to a temperature (approximately 550 to
650 C) necessary to remove residual stresses and is subsequently cooled uniformly
(to about 150 C) before the glass is removed from the annealing lehr (oven). Most
lehrs are heated by natural gas.
The only emissions from annealing lehrs are combustion products. Since
natural gas is used almost exclusively (some lehrs are electric) and the
temperatures are relatively low, emissions are low. Measurement data are not
available and emission rates are estimated on the basis of emission factors
for the combustion of natural gas. These factors are given in Table 9. Total
emissions are calculated on the basis of all product being annealed in gas-
fired lehrs.
Decorating
\
Tableware, artware and novelties are often decorated with vitrifiable
glass enamels or organic materials. A wide variety of decorating techniques
are employed. Decorations are applied by brush, with stencils, banding machines,
stamps, offset processes, electrostatically, and silk-screen priniting. Metallic
decorating materials, such as gold, platinum and silver may also be applied.
Emissions occur predominately from organic solvents and binders used in these
decorative coating which are released during the curing of the compounds.
Approximately 30 percent of tableware and are glass are estimated to
have decorative coating, amounting to 100 Gg of glassware decorated annually.
Only one point source measurement is available. Considering a worst-case
situation of 4.5 g/kg for HC emissions for decorating, the total HC annual
-------
54
TABLE 9 . EMISSIONS FROM THE ANNEALING OF
PRESSED AND BLOWN GLASSWARE
Species
NOX
S0x
Farticulates
CO
Hydrocarbons
Emission
g/kg
0.016
0
0.0012
0.0022
0.0014
Factor
(Ib/ton)
(0.032)
0
(0.0024)
(0.0044)
(0.0028)
Total Annual
Emissions (a),
Gg (ton)
0.048 (43.5)
0
0.004 (3.2)
0.007 (5.9)
0.004 (3.6)
Percent of
National
Emissions
From all
Stationary
Sources
0.0003
0
Trace 0>)
Trace (b)
Trace ^
(a) Based on 3.0 Tg of glass processed.
(b) Trace <0.0001.
-------
55
emissions would amount to 0.45 Gg of HC emitted which is 0.003 percent of
national HC emissions from all stationary sources.(11)
Frosting of Light Bulbs
Electric light bulbs are frosted with a hydrofluoric acid-ammonia
solution. Because of the corrosive nature of the fumes, this operation is
carried out in hoods or fume chambers equipped with scrubbers. Emissions
of HF and NH3 vapors are always controlled by scrubbing. The controlled
emission factor is estimated to be 0.96 g/kg for HF and 0.22 g/kg for NH,.
Total annual emissions are estimated to be 87.3 Mg of HF and 20.0 Mg of NH,,
even though no sampling data is available. This worst-case estimate is made
(Appendix B) by assuming that the scrubber in a frosting operation is per-
forming at an 80 percent efficiency, and that the amount of frosted light
bulbs are 90.7 Gg annually. This data is extracted from that available on
water pollution from the frosting operation.' '
Acid Cleaning
The funnel and screen of television picture tubes are cleaned with
a sulfuric acid/hydrofluoric acid solution before being joined together. The
process generates HF fumes which are controlled by scrubbing. The controlled
emission factor is estimated to be 0.18 g/kg and total HF emissions are esti-
mated to be 16 Mg. Although no sampling data on air emissions is available,
a worst-case estimate can be made using water pollution data,( *' as shown
in Appendix B.
Emission Characteristics
Raw Materials Preparation
Emissions from this part of the manufacturing process will reflect
the raw materials used (that is soda ash, limestone, feldspar, silica sand,
borax, and the like) since no chemical reactions take place. Softer materials
like limestone and soda ash will be more easily crushed to dust. Manufacturers
-------
56
usually specify particulate sizes ranging from 44 to 830 micron (+325 to -20
mesh). The primary ambient standard for particulate is 260 pg/m .
Glass Melting
At a glass plant, the majority of atmospheric emissions come from
the melting furnaces. Calculations to portray the effect which a glass plant
has on its neighboring air environments have been made in the following sections.
Principal attention has been given to the pollutants issuing from the melting-
furnace stacks. Emissions from the melting furnace consist of criteria pollutants
such as NO , SO , particulates, CO and hydrocarbons, as well as borates, fluorides,
X X
lead compounds, selenium and some minor pollutants. These emissions contribute
to photochemical atmoshperic reactions to produce smog and can be irritating to
the lungs.
Particulates can vary considerably depending upon the glass composition
being melted. For soda lime glasses, which comprise more than 75 percent of the
glass produced by pressed and blown glass manufacturers, the particulates consist
predominately (> 85 percent) of sodium sulfate. A clear guideline as to whether
these sulfate emissions pose a health hazard is unavailable.
Forming and Finishing
Emissions from the forming and finishing operations consist of:
0-) N0x, particulates, CO, and hydrocarbons emitted from gas-
fired annealing lehrs.
(2) Hydrocarbons produced by flash vaporization of lubricants
used in the forming region and those emitted from decorating
operations.
(3) HC1, tin or titanium oxide, and hydrated metal chlorides
exhausted from fume chambers during surface treatment
operations.
(4) Fluoride or ammonia fumes from etching and acid cleaning
operations.
(5) Other minor gaseous or particulate emissions associated
with the wide variety of finishing operations described
earlier.
-------
57
Total nationwide emissions of the criteria pollutants produced in
the different stages of the glass manufacturing process were listed previously
in Table 2.
Ground-Level Concentrations
Ground-level ambient concentrations of pollutants were used in deter-
mining the environmental effects of the atmospheric emissions. They were
calculated for representative operations used in the manufacture of pressed
and blown glassware. The diverse nature of the pressed and blown glass in-
dustry precluded selection of process equipment which was representative of the
entire industry, therefore several examples were calculated. Two soda lime glass
furnaces, one having an annual production rate of 9.1 Gg and the other with an
annual production rate of 29.9 Gg, were used in the calculations, along with
a lead glass furnace having an annual production rate of 4.6 Gg. The furnace
stack emissions were derived from the emission factors given in Table 7 and
were applied to the annual production rates. Stack heights were 24.4 meters
and 36.7 meters for the two soda-lime furnaces and 45.7 meters for the lead
furnace. Tables 10, 11, and 12 list all the parameters for the melting furnaces,
their stacks, and the ambient meteorology as used in calculating the ground-
level pollutant concentrations. Stack heights for pressed and blown glass furnaces
their stacks, and the ambient meteorology as used in calculating the ground-level
pollutant concentrations. Stack heights for pressed and blown glass furnaces
were prepared as Appendix C. They ranged from 8 to 53 meters, with the predominant
height being about 20 meters.
The nMMftmu"n ground-level concentration is used to determine information
for the environmental effect criteria. This maximum concentration can be obtained
from actual measurement or from a nomagraph for substitution into an equation.
The equation is
- -22- ^
-------
58
TABLE 10. PARAMETERS OF A SODA/LIME GIASS-MELTING FURNACE (8.1 Gg
ANNUAL PRODUCTION) REPRESENTATIVE OF THE PRESSED AND BLOWN
INDUSTRY AS USED IN ATMOSPHERIC-DISPERSION CALCULATIONS
Stack Parameters
Glass produced: 8.1 Gg/yr (9,000 T/yr)
Stack height: 24.4 m (80 ft)
Stack diameter: 0.85 m (2.8 ft)
Exit temperature: 204 C (400 F)
Gas flow rate: 710 m3/min (23,000 ACFM)
Exit velocity: 21.0 m/sec (68.5 ft/sec)
Meteorological Conditions
Wind speed: at 10 meters ~ 4.1 m/sec ^ (9.2 mph)
at top of stack 7.3 m/sec (b) (16.3 mph)
Ambient temperature at top of stack: 15 C (59 F)
Atmospheric pressure: 1000 millibars
Atmospheric stability: D'C'
Calculated Parameters
Plume rise: 8.8 m (28.9 ft)
Effective stack height: 33.2 m (109 ft)
Estimated Parameter
Mean wind speed affecting the plume between the effective
stack height and the surface : 6 m/sec
Emissions. (Q)
NOX: 1.09 g/sec (37.9 T/yr)
SOX: 0.69 g/set (23.9 T/yr)
Particulates : 1.34 g/sec (46.6 T/yr)
CO: 0.026 g/sec (0.89, T/yr)
Hydrocarbons: (0.039 g/sec (1.34 T/yr)
Selenium: 5.1 x 10~3 g/sec (0.018
(a) Average of annual mean wind speeds measured at city
airports near 30 glass-plant locations.
(b) Increase of wind with height in suburbs and level
country as given in Figures 1-3 of ASME Recommended
Guide for the Prediction of the Dispersion of Air-
borne Effluents. 1968.
(c) D stability is the predominant stability as determined
from a cross section of Star Program results (see
Table 14).
(d) Plume rise was calculated from the Holland equation
for neutral stability.
(e) Worst case.
-------
59
TABLE 11. PARAMETERS OF A SODA/LIME GIASS-MELTING FURNACE (29.9 Gg
ANNUAL PRODUCTION) REPRESENTATIVE OF THE PRESSED AND BLOWN
INDUSTRY AS USED IN ATMOSPHERIC-DISPERSION CALCULATIONS
Stack Parameters
Glass produced: 29.9 Gg/yr (33,000 T/yr)
Stack height: 36.7 m (120 ft)
Stack diameter: 1.8 m (6.0 ft)
Exit temperature: 399 C (750 F)
Gas flow rate: 341 m3/min (12,000 ACFM)
Exit velocity: 2.24 m/sec (7.29 ft/sec)
Meteorological Conditions
Wind speed: at 10 meters -- 4.1 m/sec(a) (9.2 mph)
at top of stack 8.2 m/sec(b) (18.4 mph)
Ambient temperature at top of stack: 15 C (59 F)
Atmospheric pressure: 1000 millibars
Atmospheric stability: D^c^
Calculated Parameters
Plume rise: 2.99 m(d) (9.8 ft)
Effective stack height: 39.7 m (130 ft)
Estimated Parameter
Mean wind speed affecting the plume between the effective
stack height and the surface: 6 m/sec
Emissions.Co)
NCv: 4.03 g/sec (140.1 T/yr) -
SO: 2.54 g/sec (88.3 T/yr)
Particulates: 4.95 g/sec (172.0 T/yr)
CO: 0.095 g/sec (3.30 T/yr)
Hydrocarbons: 0.142 g/sec (4.94 T/yr)
Selenium: 0.93 x 10-3 g/sec (0.067 T/yr)(*'
(a) Average of annual mean wind speeds measured at city
airports near 30 glass-plant locations.
(b) Increase of wind with height in suburbs and level
country as given in Figures 1-3 of ASME Recommended
Guide for the Prediction of the Dispersion of Air-
borne Effluents. 1968.
(c) D stability is the predominant stability as determined
from a cross section of Star Program results (see
Table 14).
(d) Plume rise was calculated from the Holland equation
for neutral stability.
(e) Worst case.
-------
60
TABLE 12. PARAMETERS OF A LEAD GLASS-MELTING FURNACE REPRE-
SENTATIVE OF THE PRESSED AND BLOWN INDUSTRY AS
USED IN ATMOSPHERIC-DISPERSION CALCULATIONS
Stack Parameters
Glass Produced: 4.62 Gg/yr (5,100 T/yr)
Stack height: 45.7 m (150 ft)
Stack diameter: 1.5 m (5.0 ft)
Exit temperature: 466 C (870 F)
Gas flow rate: 654 m3/min (23,000 ACFM)
Exit velocity: 6.17 m/sec (20.1 ft/sec)
Meteorological Conditions
Wind speed: at 10 meters -- 4.1 m/sec (a' (9.2 mph)
at top of stack 8.8 m/sec0>) (19.7 mph)
Ambient temperature at top of stack: 15 C (59 F)
Atmospheric pressure: 1000 millibars
Atmospheric stability: D'C'
Calculated Parameters
Plume rise: 5.72 m (18.7 ft)
Effective stack height: 51.4 m (169 ft)
Estimated Parameter
Mean wind speed affecting the plume between the effective
stack height and the surface: 6 m/sec
Emissions,( )
Particulates: 2.20 g/sec (76.4 T/yr)
Fluorides: 1.47 g/sec (50.9 T/yr)(8'
Selenium: 2.93 x 10~4 g/sec (0.010 T/yr)
(a) Average of annual mean wind speeds measured at city
airports near 30 glass-plant locations.
(b) Increase of wind with height in suburbs and level
country as given in Figures 1-3 of ASME Recommended
Guide for the Prediction of the Dispersion of Air-
borne Effluents. 1968.
(c) D stability is the predominant stability as determined
from a cross section of Star Program results (see
Table 14).
(d) Plume rise was calculated from the Holland equation
for neutral stability.
(e) Worst case.
-------
where
Xmax * maximum concentration (gm/m3)
Q - pollutant emission rate (gm/sec)
u = mean wind speed (m/sec) at the height of the stack
H - effective stack height (m), the physical height
of the stack plus the plume rise
o^ - vertical plume standard deviation (m)
a » horizontal plume standard deviation (m)
e - base of natural logarithms, 2.718
TT - 3.14.
a
For stability Type D, the ratio -2- is on the order of 0.5 varying from
(251
0.57 to 0.24 between 0.1 km and 10 km downwind from a source . The ratio
is approximately 1.0 for stability Type C. The maximum concentration occurs at a
distance where a = H//2. Turner has presented a nomagraph from which
z
X u/Q and the distance to the point of maximum concentration can be deter-
max
mined for any stability and effective stack height. When emission rate and wind
speed are known, the value of X can be calculated.
r max
The environmental effects criteria are developed for 24 hr average
concentrations, while the dispersion predictions discussed above are for short
periods (3 to 10 min). For longer periods, one must consider that variations
in wind direction and wind speed will cause the average concentration at a
downwind monitor to be less than the concentration calculated for a short-term
wind blowing constantly from the source to the monitor. Turner has given an
equation by which the long-term average concentration can be estimated when
the short-term concentration is known:
-------
62
where Xn = concentration for the long period (t.)
* *
X » concentration for the short period (t )
s s
t. « long-time period, min.
X*
t = 'short-time period, min.
The value of the dimensionless exponent, b,
is between 0.17 and 0.2
While this equation is most applicable for X « 2 hr or less, it can be applied
to a 24-hr period. Turner gives the conversion coefficient of 0.35 for trans-
forming a 3-min average into a 24-hr average. Other conversion coefficients
are 1 hr, 0.61, and 3-hr, 0.51.
Before calculating ambient pollutant concentrations, representative
meteorological parameters for the area need to be chosen. These parameters,
along with stack parameters, are required for determining plume rise and
dispersion. Plume rise is calculated from the Holland equation:
/vd\ r ri r T - T
AH -V.-J-; (1.5 +[2.68 x 10"jp[_-^ -
S
where
AH = rise of the plume above the stack, m
v » stack gas exit velocity, m/sec
S
d m inside diameter of stack, m
u - wind speed at top of stack, m/sec
p - atmospheric pressure, millibars
T = stack gas temperature, K
T = air temperature, K.
Ci
-------
63
Choices of the meteorological parameters are made after a review of
climatology in some of the areas of the country where glass plants are found(26).
Account is also taken of the variations of meteorology between the surface and
the top of the furnace stack. The values selected for the melting-furnace calcu-
lations are listed in Tables 10, 11, and 12. Stability Type D (neutral class)
is the most frequently occurring stability throughout the United States as calcu-
(25)
lated by the Turner method which considers the surface wind speed and the net
radiation (Table 13). A surface wind speed of 4.1 m/sec is chosen as representative
of the conditions at the glass plants based on a survey of the average annual
wind speeds listed for the National Weather Service meteorological stations
located at 30 cities which have glass plants. The reader should note that the
4 meter/sec wind speed in Turner's scheme"for determining stabilities can accompany
stabilities verying from Type B to Type E, depending on the solar radiation.
Type D is chosen for the dispersion calculations on the basis of its predominant
frequency. Wind speeds increase with altitude and this effect is taken into
account for the effective stack heights of the representative furnaces. Wind
speed in the layer in which the downward dispersion of the plume should take
place, 0-33.2, 0-39.7, and 0-51.4 meters for the three furnaces, is estimated
to be 6 meters/sec. This is an extrapolation from the standard wind-measurement
height of 10 meters over suburban and level rural areas. For stack heights of 30
to 50 meters, the wind speed is expected to be 1.5 (level terrain) to 3 (urban
areas) times stronger at the top of the stack than at 10 meters.
Table 14 (in its second column) presents the theoretical maximum pollu-
tant concentration predicted for ground level in the vicinity of the glass-melting
furnaces. These concentrations are the contributions from only the furnace and do
not take into account other glass-plant emissions or emissions from sources other
than the glass plant.
Emissions from two other sources representative of air emissions from
a manufacturing operation in the pressed and blown glass industry were also con-
sidered in relation to their effect on ambient-air quality. These were:
(1) Particulates from a baghouse collecting the emissions
from materials handling
-------
64
TABLE 13. RELATIVE FREQUENCY OF ATMOSPHERIC STABILITIES
(a)
Stability Class
Station
Milwaukee
St . Louis
Peoria
Pittsburgh
Columbus, 0.
Mobile
Los Angeles
Dallas
" A
0.001
0.005
0.003
0.001
0.010
0.008
0.001
0.004
B
0.031
0.047
0.042
0.022
0.058
0.052
0.041
0.042
C
0.094
0.103
0.102
0.083
0.100
0.115
0.148
0.107
D
0.636
0.555
0.577
0.567
0.500
0.453
0.482
0.586
E and F
0.238
0.289
0.276
0.306
0.331
0.371
0.329
0.262
(a) Based on Output from U.S. Department of Commerce
National Climatic Center Star Program for Five
Years of Data.
-------
65
TABLE 14. MAXIMUM POLLUTANT CONCENTRATIONS AND SOURCE SEVERITY FOR EMIS-
SIONS FROM REPRESENTATIVE PRESSED AND BLOWN MELTING FURNACES
Ambient Air- 3-Minute Ad justed (a)
Quality Standard, x,^ x^,
Pollutant (|J.g/m3) (Hg/m3) Qj-g/m3)
Soda/Lime Furnace 8.1 Gz Annual Production
N0x 100(b) ' 19.9 7.3
S0x 365(c) 12.6 4.5
Particulates 260^c' 24.6 8.8
CO 40,000^ 0.52 0.23
Hydrocarbons 160 ^ 0.72 0.36
Selenium 0.67(f) 9.4xlO~3 3.4xlQ-3
Soda/Lime Furnace 29.9 Gg Annual Production
NO 100 (b) 49.3 17.7
x
SOV 365^c) 31.0 11.1
X
Particulates 260^) 60.6 21.9
CO 40,000 1.13 0.48
Hydrocarbons 160 ^' 1.74 0.88
Selenium 0.67 30.9 10.7
Fluoride 8.33 4.1xlO"3 l.SxlO'3
(a) 3-minute X^x adjusted to match sampling time of the
standard using the following conversion factors from
Turner^5^ :
0.36 for 24 hours, 0.42 for 8 hours
and 0.51 for 3 hours.
(b) Annual arithmetic mean assumed here as 24-hr standard.
(c) 24-hr standard.
(d) 8-hr, standard.
(e) 3-hr standard. 3
(f) Obtained from TLV x 8/24 x 1/100 where TLV = 2.5 mg/m
for fluoride and 0.2 mg/m3 for selenium.
Severity,
S
0.073
0.012
0.034
5.7xlO~6
2.3xlO"3
5.0x10-3
0.18
0.030
0.084
1.2xlO-5
5.5xlO~3
9.013
0.041
0.89
2.2xlO'3
-------
66
(2) Hydrogen chloride and titanium chloride from surface
treatment operations.
To make the ambient-concentration estimates for these sources, emissions and
(9)
stack parameters were adapted from data given in the NEDS listing . Meteorolgical
conditions similar to those used in the glass-furnace emission-dispersion calcula-
tions were used for these other sources with adjustments for differing stack heights.
Information regarding these calculations was prepared as Table 15.
For each of the maximum ambient concentrations that have been calculated,
a source severity, S, is also determined. Source severity for criteria pollutants
(particulated, sulfur oxides, nitrogen oxides, carbon monoxide, and hydrocarbons)
is determined from the following equation:
X
max
AAQS
where
X = maximum average ground-level concentration of the
H13.2C A
pollutant for the time period of the standard (pg/m )
AAQS = ambient air-quality standard (yg/m ),
For noncriteria pollutants, the source-severity equation uses the threshold limit
value instead of the ambient air-quality standard with a correction for a 24-hour
period and a safety factor:
X
max
TLV (8/24) (1/100)
where
TLV = Threshold Limit Values for each species
8/24 - Correction factor for the 8-hr work day which
is the basis for the TLV
1/100 = Safety factor.
-------
67
TABLE 15. MAXIMUM AVERAGE GROUND-LEVEL CONCENTRATION (X_ ) OF SELECTED
AIR POLLUTANTS FROM REPRESENTATIVE MATERIALS HANDLING AND
TREATMENT OPERATIONS
Source 1. Baghouse Controlling Materials-Handling Emissions (98% Efficiency)
Materials Handled: 0.3 Gg/yr
Emission Point: Stack
height: 9.1 m; diameter: 0.5 m
exit temperature: 21 C; exit velocity: 10.3 m/s
Emission Factor: 1.91 g/kg
Emissions: 0.57 Mg/yr
(0<63
Species
Particulates
(3 min),
Ug/m3
Xmax» V8/m3 Ambient Severity
(specified time) Std yg/m3 Factor
5.46
1.96
260
0.0076
Source 2. Surface Treatment of Glass
Production: 9 Gg/yr
Point of Emission: 13 m stack (no plume rise assumed)
Pollutants Considered
Hydrogen Chloride
Titatanium Chloride
Emission Factor (g/kg)
0.02
0.02
Emissions
5.71xlO"3 g/sec
(0.20 ton/yr)
5.71xlO-3 g/sec
(0.20 ton/yr)
Pollutant
Hydrogen Chloride
Titanium Chloride
yg/m3
11.43
11.43
(Specified Time)
4.11
4.11
Severity Factor Determination
Pollutant
Hydrogen Chloride
Titanium Chloride
TLV. mg/m3
7
10
Severity Factor
0.176
0.123
-------
68
A review of the source-severity factors in Tables 14 and 15 shows the
highest value to be that produced by emissions of fluorides from a lead glass
furnace, S » 0.89. The next highest source-severity factors is 0.18 for both
nitrogen oxides emitted from a soda-lime furnace and hudrogen chloride from a
surface treatment operation.
Affected Population
As a consequence of the dispersion of pollutants, the severity starts
at zero near the stack, increases downwind, reaches a maximum, and then decreases
to zero again (see Figure 10). The affected population is defined as the popu-
(28)
lation around plant exposed to a severity greater than 1.0 . To determine the
downwind distances enclosing the affected population, the standard dispersion
equation for the centerline concentration from an elevated source is used.
[-If
*-
where
_3
X = pollutant concentration at surface (gm )
u = average wind speed through the dispersion
layer (m sec" ) . The winds from all directions
are assumed to be euqally likely.
Other parameters are the same as in the earlier
dispersion equation (page 58).
By rearranging, this equation becomes
-------
69
c
o
c
4>
O
u
c
o
u
.2
0)
U
1.0
R
max.
Distance Downwind, X
R
R z Outer radius
R i Inner radius
FIGURE 10. ILLUSTRATION DEPICTING CALCULATION OF AREA
WHICH CONTAINS THE AFFECTED POPULATION
-------
70
The value of X is specified by the requirement for S * 1.0 and then it is cor-
rected to the three-minute average concentration which the dispersion equation
gives. Substituting values of o and a from Turner's graphs of dispersion
y z
coeffiency as a function of. distance downwind into the righthand side of the
equation versus downwind distance. These values are plotted in a fashion similar
to Figure 10 and the values of R- and R- are determined. These values form the
inner and outer radii of an annulus enclosing the affected population.
Since no source severity factor for the pressed and blown glass industry
was found to be greater than 1.0, no affected population calculation was made.
-------
71
SECTION V
CONTROL TECHNOLOGY
Control of air emissions in the glass industry varies considerably,
depending on the type, source, and amount of emission. Control technology
has evolved for both economic and environmental reasons, and various methods
are utilized to reduce air emissions from the different portions of the
glass-manufacturing process. These methods include: (a) development of
process modifications, (b) new furnace designs, and (c) application of
control equipment. For example:
(1) Use of arsenic has been reduced for use as a fining
agent.
(2) Many fossil-fuel-fired furnaces are equipped with
electric boosting which can increase output, thus
reducing the amount of effluent per unit of output.
Some manufacturers have switched entirely to all-
' ' ' i
electric melting.
(3) Fabric filters, electrostatic precipitators, and
scrubbers, are being used or have been examined
for removal of particulates. In addition, several
commercial equipment manufacturers are attempting
to develop methods for removal of SO and NO
X X
emissions at the same time particulates are removed.
This section discusses the control technology currently being used or being
considered for use by manufacturers of pressed and blown glassware. The
study does not consider the economics or verify the control technology it-
self. Rather, this section identifies control technology reportedly
applicable to the glass industry. The discussion is organized in a manner
similar to the emission section.
-------
72
Raw-Materials Preparation
The handling and mixing of raw materials is a source of particulate
emissions from any glass plant melting such materials. Raw materials are
normally conveyed (by screw conveyors, belt conveyors, bucket elevators, or
pneumatic conveyors) from hopper railroad cars or trucks to elevated storage
bins, as has been shown in Figure 3. Other glass-batch ingredients are
delivered to the plant in paper bags or cardboard drums, and are later trans-
ferred manually to smaller storage bins or fed directly from the storage device.
Materials are gravity fed from the storage bins into weigh hoppers
and then transported by transverse belts or bucket elevators into a mixer.
Gullet is crushed to a desired size. After mixing, the glass batch is trans-
ferred to a charging bin located next to the glass-melting furnace or into
a batch storage bin, depending upon the design of the batch-handling system.
Positive displacement or vibratory feeders at the bottom of the bins feed
the materials to the chargers, where it is fed into the glass-melting furnace.
Gullet may be added to the batch in the mixer, while the batch is being
(29)
transferred, or charged separately to the melting furnace
Emissions
Little information is available regarding plant emissions due to
dusting during the raw-materials handling stages of the process. As discussed
in the previous section, the fraction of the dust generated which leaves the
plant site will consist of particles smaller than 100 microns in diameter.
Also, as described in the previous section on emissions, the particulate
emission rate for raw-materials handling is estimated to be 1.5 g/kg. Based
upon the total glass batch handled by the pressed and blown glass industry,
annual particulated exhausted are estimated to average 6.26 Gg.
Information on the composition of these particulate emissions is
not available, but they will consist essentially of the same raw materials
being handled (soda ash, silica sand, limestone, etc.), since no chemical
reactions occur during this portion of the manufacturing process. Softer
materials (e.g., ash) can be expected to predominate. Glass manufacturers
-------
73
will generally use raw materials which are coarser than 44 micron, as shown
in Table 16. Pressed and blown glass manufacturers use a greater percentage
of raw materials finer than 100 microns. Even still, uncontrolled emissions
should not exceed 5 percent of the total materials handled. The amount of
raw material emitted from the plant site due to inertial forces alone would
be relatively small, as reported measurements indicate^.
Raw-Materials-Control Technology
Process Modifications or Materials Selection. Manufacturers of
pressed and blown glassware will generally minimize dusting problems in batch-
handling operations by limiting the amount of fine particles (<100 microns) in
the batch material. Specifications for glass-grade raw materials will generally
require removal of the finer sizes of material, especially with softer materials
that tend to be crushed to dust easier than sand.
Another batch-preparation method that is used to control dusting
during handling is the addition of water to the raw batch (batch wetting).
Trials have also been conducted during which the batch is wet with a liquid
(4)
caustic-soda solution that is substituted for soda ash . Water is presently
added in amounts up to 4 percent to the mixed batch materials. The substi-
tution of a caustic-soda solution for a soda ash is not generally practiced
by the glass industry ^4>3°\
At those points in the raw materials handling and preparation stage
where dust may be generated, control is accomplished through the use of
collection equipment. This is almost always done with fabric filters (e.g.,
baghouses).
Efficiency of Control Equipment. Transport of raw materials in rail-
road hopper cars and hopper-bottom trucks (dump trucks) is still practiced.
During unloading of these trucks or railroad cars, the dumping of materials
onto conveyor belts can result in some dust being dispersed into the air.
Generally, the hopper cars or trucks are connected to sealed receiving hoppers
with fabric sleeves and the dust generated during the unloading operation is
(31 32)
filtered through the sleeves or exhausted through a baghouse ' . Enclosing
the loading area with a suitable fabric structure and sealing all covers and
access opening with gaskets is effective in reducing dust during this operation.
-------
TABLE 16. GLASS-GRADE PARTICLE-SIZE SPECIFICATIONS FOR
SAND, LIMESTONE, AND 10- AM) 20-MESH DOLOMITE
Approximate
Particle
Size
2 .3 mm
1.3 mm
820 p.
410 M,
150 n
105 M,
74 M.
44 p.
Glass- Glass -Grade
U.S. Standard Grade Glass-Grade Dolomite. %
Mesh Size Sand, % Limestone, % 10-Mesh
Cum retained on 8 - 0.0 0.0
Cum retained on 16 - 2.0 max 15.0 max
Cum retained on 20 0.0 10.0 max
Cum retained on 40 12.0 max
Cum retained on 100 - - 90.0 min
Cum retained on 140 92.0 min 85.0 min
Cum retained on 200 99.5 min 94.0 min 97.0 min
Cum retained on 325 100.0 min
20-Mesh
0.0
-
2.0 max
-
80.0 min
95.0 min
96.0 min
-
-------
75
This results in an inward-air velocity across the open mouth of the bag that
prevents an eruption of dust into the atmosphere ^32\ Trapped air and fine
dust can then be filtered by a conventional fabric filter and the cleaned air
exhausted into the atmosphere.
Weigh hoppers and mixers require ventilation because of surges in
material from the large air flows. In older mixers, polyvinylchloride seals
are generally installed between the rotating body of the mixer and its frame
to reduce air leaks. In newer mixers, the body does not rotate. The exhaust
gases are usually filtered of particulates greater than submicron size by the
use of fabric filters.
The use of fabric filters for separation of particulates from air
has been practiced for a number of years in the glass industry. The earliest
fabric filters were known as "baghouses", since they were large free-standing
units for exposed fiber bags. By passing the exhaust air through layers of a
woven fabric, the particulates were collected. Unfortunately, as the thickness
of the collected layer of particulates increases the pressure differential
required for continued air flow also increases. Thus, the collected dust must
be periodically removed by manual or mechanical shaking. Almost all container
(9 33)
glass plants use fabric filters to remove entrained dust particles ' . The
fabric filters used today are totally enclosed, and most have a continuous
removal operation for the trapped particulates. The traditional woven and
synthetic fabrics are used. Today, fabric filters are generally made of
low-temperature materials such as Nomex, nylon, terylene, or Orion'3*' .
Fabric filters are used to collect particulates from the raw-materials
and handling operations for several reasons. First, they have an efficiency of
greater than 99 percent and they can be used to collect fine particulates. In
addition, the trapped particulates can sometimes be recovered for reuse or re-
cycle <5«31'32^. One manufacturer has from 2 to 6 baghouses with a stack height
less than 50 feet at a plant manufacturing 72.6 Gg (80,000 tons) of container
glass per year^9'33^. They used nylon-fabric filters operating at 98 percent
efficiency and collecting about 36.3 kg (80 Ib) of dust per year.
-------
76
Glass-Melting Operation
In a glass-melting furnace, raw materials and cullet are heated until
a homogeneous, viscous liquid, free of gas inclusions, is formed. Temperatures
in the melter will generally be in the range of 1500-1600 C (2730 to 2913 F)(3A).
Natural gas and fuel oil are the principal types of fuel, with natural gas pre-
dominating (60-65 percent) . Over 80 percent of glass-melting furnaces
have regenerative firing systems for purposes of heat recovery and fuel con-
/Q\
servation . To increase melting capacity, many furnaces now have electric-
boosting systems. These systems consist of several water-cooled electrodes
euqally spaced along the sides or bottom of the melter, below the surface of the
glass.
Additionally, all-electric melting furnaces are utilized by portions
of the pressed and blown glassware industry. With all-electric melting, the
glass is heated by its own self-resistance by passing an electric current
through it. Electric melters currently melt less than 10 percent of the
(37)
glass in the United States . This type of melter contains a. blanket of
glass batch which covers the entire surface of the molten glass. Any volatiles
are almost entirely trapped by the glass batch as they percolate up through
the batch blanket especially when borosilicate and opal glasses are being
melted. Electric melting offers somewhat less pf an abatement advantage for
the melting of soda-lime glass.
i
Emissions
Major criteria air emissions from a glass-melting furnace consist
of NO , SO , HF, and particulates. Other emissions include CO, hydrocarbons,
X X
and selenium.
Nitrogen oxides represent the second largest fraction by mass,
about 21 percent of glass-furnace emissions ' . As described earlier, the
source test measurements of NO emission rates vary from 0.41 to 10.0 g/kg
X
of glass produced. Based on an average emission rate of 4.25 g/kg, glass-
melting furnaces with a total production rate of 2.8 Tg would emit 9.5 Gg
of NO yearly.
x ' J
-------
77
S0x emission, on the other hand, is dependent primarily upon the sulfur
content of the fuel and, to a lesser extent, on the sulfur content of the batch
material. Sulfur present in the fuel oil will oxidize and appear as SO in the
exhaust gas. A fuel oil containing 1 percent sulfur by weight emits * 600 pom
(13)
S02 in the flue gas . Sulfur can also be present in the batch materials,
usually as Na^O^. During heatup, the sulfate decomposes and sulfur dioxide
forms, some of which is chemically incorporated into the glass (as SO") and some
of which is released within the furnace. An average emission rate of SO for soda
x
lime glass is 2.68 g/kg. Thus, plants melting 2.24 Tg of glass annually would
emit approximately 6.0 Gg of SO yearly.
X
HF is emitted from opal and certain lead and borosilicate glasses. The
emissions result from the decomposition of fluoride bearing batch materials. A
portion of the fluoride (^40 percent) remains in the glass, the remaining being
emitted as HF gas or as a fluoride compound. The quantity of HF emitted depends
on the glass batch composition and the furnace operating parameters. The uncon-
trolled average total fluoride emission rate is estimated to be 10.0 g/kg, with
annual emissions of 2.5 Gg of F~.
Farticulate emissions from a glass-melting furnace result primarily
from volitization of materials in the melt that combine with gases such as S03
or HF to form condensates in the flue system. Farticulate emissions from soda
lime glass consist of approximately 80 percent sodium sulfate » » » . These
particulates form from the condensed vapors in the melt and are submicron sized
(6,18-21)^ with the nedian particle diameter being about 0.13 u ' Larger
sized particles are generally retained in the regenerative system ' . Far-
ticulates from other glass types are somewhat less defined. They can include
NaF, B,03, PbO, PbS04, and Na2SiFg, depending upon the glass type. Particle size
distributions are not clearly defined, but the average size is generally less than
2 micron (2°'21\ Source-test measurements for particulate emission rates vary
from 0.49 to 12.57 g/kg of glass produced, which average to a particulate emission
rate of 5.22 g/kg.
Carbon monoxide is exhausted from the glass-melting furnace, primarily
as a result of incomplete fuel combustion. Source-test measurements have reported
-------
78
emission rates from 0.09 to 0.15 g/kg. An estimated average emission rate is
0.10 g/kg.
Hydrocarbons are also formed in the glass-melting furnace as a result
of incomplete fuel combustion. Source-test measurements have reported emission
rates from 0.02 to 0.27 g/kg. The calculated average emission rate is 0.15 g/kg.
Actual emission rates are a function of firing conditions (extent of fuel/air
mixing, excess air, firing temperature).
Selenium is generally used in amounts of 2 weight percent or less in
the soda-lime glass batches as a decolorizer to neutralize the green tint in
glasses caused by iron impurities. No test measurements on actual selenium
emissions have been reported, but it likely leaves the stack as selenium
vapor because of its low vaporization temperature (315 C for SeO and 685 C for
Se) ' . Based on an average production of 1.8 Tg (2 million tons) of glass,
an average emission rate for selenium has been calculated to be 0.002 g/kg.
Glass-Melting-Control Technology
Control of emissions from the glass-melting furnace has occurred for
both economic and environmental reasons. Five general approaches have been
employed:
(1) Modification of feed material
(2) Modification of furnace design, including electric melting
(3) Increase of checker volume
(4) Adoption of commercial-control apparatus
(5) Modification of furnace operation.
Modification of feed material, furnace design, and furnace operation have been
used primarily to control gaseous emissions, while the other methods have fo-
cused on control of particulate emissions.
Modification of Feed Material. Some of the glass batch raw materials
have a tendency to vaporize or decompose in the glass-melting furnace (e.g.,
fluorides, nitrates, and selenium). By minimizing the amount of these or other
ingredients used or by substitutions of other materials, the volume of gaseous
-------
79
emissions exhausted from the glass-melting furnace can be reduced. For example,
the use of arsenic as a fining agent has been reduced and changes have been made
to produce fluoride-free glass batches. Cerium is used to partially replace
selenium as a decolorizer. In addition to reducing the selenium in the batch
(by about 40 percent), this modification leads to the elimination of arsenic
in the batch, since cerium and arsenic are not compatible. Cerium is especially
appealing because it tends to form high-melting compounds which do not readily
vaporize.
Modification of Furnace Design. Increasing the fuel efficiency of
the glass-melting furnace can in turn lead to a decrease in combustion products,
a decrease in dust entrainment by hot combustion gases passing over the melting
glass batch, and possibly a decrease in furnace temperature. In addition,
emissions from low melting and easily vaporized fluxing or fining agents can be
lowered. Several methods currently in practice to improve furnace efficiency
are:
(1) Setter instrumentation for regulating air/fuel mixtures
and monitoring furnace temperature and stack gas composition.
(2) Combustion control to produce long luminous flames that
eliminate spurious hot spots in the furnace and provide
better heat transfer to the melt.
(3) Improved refractories to increase corrosion resistance,
which permits furnaces to be more fully insulated.
(4) Use of electric boosting to increase furnace capacity,
increase furnace efficiency, and lower temperatures
above the molten glass.
All of these methods have been employed to control gaseous emissions. Sulfur
oxides that form can be controlled by both limiting the sulphate in the feed
material and by the improvement of furnace efficiency. Ryder and McMackin have
found that the SO emission rate increases directly with, an increase in production
rate on a sideport furnace melting soda lime glass. This increase is attributed
to the higher temperatures needed (1552 C versus 1460 C) (2825 F versus 2660 F)
when the daily production rate is doubled to 181 Mg (200 tons).
-------
80
NO emission can be also lowered when the furnace efficiency is
x
increased if the furnace temperature also drops. A 10 percent decrease in
fuel consumption can cause a 10 percent decrease in NO emissions *
X
Electric boosting is commonly used on fossil fuel-fired furnaces in
the container glass industry, primarily to increase output. Boosting can result
in a reduction in emissions per unit of output.
Electric Melting. These furnaces are used to essentially eliminate
both particulate and gaseous emissions from the glass-melting operation. As
discussed previously, the cold batch covering the glass traps the majority of
these emissions. In a fossil-fuel fired melter, volatilization occurs at the
interface between the hot glass and the combustion gases. This condition does
not exist in the all-electric melter, and consequently this source of emissions
is elimated also.
Electric melting is utilized to a much greater extent in the manufac-
ture of pressed and blown glassware than with container glass because higher
quality glass can be produced at virtually zero emission rates. It is not
used to make flat glass because furnace sizes are more incompatible. Electric
melting does have certain operational and control problems, and experience
with large melting units (> 120 Mg) is essentially nonexistent. Because of
capital considerations and the higher cost of electricity, electric melting is
often not judged to be economical. In recent years, the need to control
emissions has made the use of electric melting more economically appealing
for non-soda-lime glasses. However, in 1975, less than 5 percent of the glass
manufactured was made by electric melting
Adoption of Commercial-Control Apparatus. Particulates can be cleaned
from the glass furnace exhaust by scrubbers, fabric filters, or electrostatic
precipitators (ESP). Scrubbers can also be used to collect SO emissions, while
fabric filters and ESP's only remove particulates. Teller '' suggests
spraying the stack gas with an alkaline solution, causing the acidic gases
(SO , HF, of HC1) to n
X
by the control device.
(SO , HF, of HC1) to react and form particulates that can then be collected
X
Scrubbers. One type of particulate scrubbing is a two-step process.
Initially, particles in the exhaust gases are "contacted" or wet by a scrubbing
-------
81
fluid that draws the particles into agglomerates. These agglomerates are then
separated from the gas stream by an inertial mist-elimination process.
A low-pressure (< 10-in. water) centrifugal scrubber used by the
Thatcher Glass Company in Saugus, CaliforniaC5) had two separate contacting
sections within a single casing. Separate 50.7 metric horsepower (50 horsepower)
circulating fans forced dirty gas through each section containing two to three
impingement elements similar to fixed blades of a turbine.
(39)
One reference mentions a scrubber that uses a packed-bed pre-
conditioning chamber. Hot gases (538 C) containing volatized sodium compounds
enter the chamber, while the vapors condense out onto the packing material. This
material is wet by a scrubbing solution and provides a large surface area for
condensation. A standard Venturi-type scrubber completes the system. This
scrubber is presently installed on a 0.181 Gg/day (200 ton/day) container (soda-
lime) glass furnace and it reduces particulate loading from more than 0.23 to less
than 0.046 g/sdm (from more than 0.10 to less than 0.02 g/sdcf)'39'.
One soda-lime glass manufacturer installed a tower scrubber (2.9-meter
2 2
diameter) on a 44.8 meter (482 ft ) melter. Hot effluent from the furnace is
initially quenched and saturated with a caustic solution passing through the ex-
haust gas at 900 gal/min. The gas then passes into a 300 gal/min variable throat
Venturi operating at 30 in. of water. This scrubber has been plagued by mal-
functions and breakdowns. A highly visible steam plume is exhausted when it is
not working.
(34)
Fabric Filters. Fabric filters, also known as "baghouses"v , collect
particulates by filtering exhaust gas from glass-melting furnaces through closely
woven natural or synthetic fabric filters that are capable of trapping submicron
particulates. Unlike wet scrubbers, fabric filters are less affected by varia-
tions in the gas flow rate. Temperature control, however, is very critical for
proper functioning and the type of fabric filter selected is dependent upon the
temperature of the exhausted gases. Fabric filters are generally made of cotton
(32 34)
sateen, standard nylon, wool, dacron, orlon, NOMEX, teflon, and fiberglass
Maximum operating temperatures for these fabrics are given in Table 17. Since
stack gas from a glass melting furnace is at 316 to 645 C (9600 to 1200 F) , the
gas must be cooled to a temperature compatible with the fabric filter bag. This
cooling can be accomplished by using the following methods, either alone or in
combination:
-------
82
TABLE 17. MAXIMUM USE TEMPERATURE FOR VARIOUS
FABRIC-FILTER MATERIALS
Maximum Temperature
Fabric
Cotton Sateen
Standard Nylon
Wool
Dacron
Orion
Nomex
Teflon
Fiber Glass
F
190
200
225
275
275
400
450
550
C
88
93
107
135
135
204
232
288
-------
83
(1) Air dilution
(2) Radiation-cooling columns
(3) Air/gas heat exchangers
(4) Water-spray chambers.
Dilution of off gasses with air is the simplest and most trouble-free method for
reducing temperature, but requires the largest baghouse because of the increased
volume of gases. Air-to-gas heat exchangers, and radiation and convection duct-
work are subject to fouling from dust in the effluent. A water-spray increases
humidity and requires careful temperature control to avoid condensation, but it
does permit use of smaller baghouses. Care must be taken with all of these
methods to avoid cooling the gas to the temperature where SO and H.O will
j fc
combine and condense, fouling or reacting with the fabric filters. In addition
to being selected for their thermal compatibility, fabric-filter bags must also
be corrosion and abrasion resistant. Cotton, orIon, and dacron can deteriorate
(221
from the SO in the flue gasv '.
A fabric filter air-pollution control system was installed in 1974
on a 41.8 m (450 ft ) melter producing soda-lime glass( . The 482 C (900 F)
effluent from the furnace was initially cooled to 177 C (350 F). A fine powder
aluminate precoat was then introduced into the air stream at 18.1 kg/hr (40 Ib/hr)
along with ambient air, further reducing the gas temperature to 121 C (250 F).
2 2
The baghouse contained 1200 m (12,915 ft ) of dacron-filter cloth divided into
six compartments, each containing 900 filter bags. During normal operation, the
air-to-cloth ratio was 1.55, but this increased to 1.86 during the cleaning cycle.
The pressure drop across the bags ranged from 3.5 to 4.5 in. of water. An
exhaust blower had to develop 16 to 18 in. of water pressure to overcome the re-
sistance of the checkers, heat exchanger, baghouse, and about 46 meters of duct.
Initially, the heat exchanger required maintenance about 15 percent of the time
due to plugging with material condensing from the gas stream. By blocking off
about 40 percent of the tubes, a normal maintenance schedule was used, but the
temperature increased slightly. Discharge of particulates from the baghouse
outlet was typically 1.1 kg/hr (2-3 Ib/hr). Tests using a Brinks Impactor
showed these particulates to be < 0.75 micron.
-------
84
Electrostatic Precipitator (ESP). In an electrostatic precipitator
(ESP), a voltage source creates a negatively charged area, usually created by
hanging wires in the gas flow path. Grounded collecting plates composed the sides
of the ESP. A pwerful electric field is created by the high potential difference
between these grounding plates and the discharging wires. As the gas stream
passes through the field, the particles become electrically charged and are
drawn to the collecting plates. Periodically, accumulated particles are removed
from these plates by vibration, rapping, or rinsing. Thus, by applying the
collecting force only to the particles to be collected, a much lower power input
is required (i.e., 200 watts per 0.5 m /s) .
NAFCO Engineering, Ltd. (a Japanese firm) has developed a new type of
ESP. In contrast to the conventional units, the NAFCO ESP uses thousands of
stainless steel needles affixed to the leading and trailing edges of positively
charged electrode plates. Thirty-five of these systems are now in operation in
Japan, with nine of them being used on soda lime glass-melting furnaces and
the remaining on other pressed and blown type glasses. United McGill Corporation,
who is the licensed United States distributor for the NAFCO ESP, has installed
the unit on 20 pressed and blown glass furnace to date^^\ All of these
systems have an outlet particulate loading of < 0.046 g/std nr* (0.02 g/scfd)
or less.
2 2
An 84.4 m (908 ft ) melting furnace, used for producing soda lime glass,
had an ESP installed in early 1974 . It consisted of dual chambers, in which the
air flow could be directed to either chamber or divided between them. Each chamber
had three electrical fields connected in series. Designed for 12.9 sec. treatment
time at 0.67 m/s (2.2 fps) velocity through the treater, one chamber was found to
be as effective as two, the conclusion being that the system was over-designed.
Other Technology. Collector systems previously discussed are primarily
useful for collecting particulates and for decreasing opacity of gaseous emis-
sions. One company now offers dry and wet systems ' ' to control both parti-
culate and gaseous emissions. A nucleation scrubber is used on their wet system
to effect collection of submicron particulates and acidic gases (HF and SO ). A
solid absorbent, on the other hand, is injected into the gas stream to react with
the noxious gases in their dry system. The absorbant is then separated from the
gas along with particulates in a fabric filter.
-------
85
A patent (U.S. 3,789,628) was issued for a scrubber in which an aqueous
solution of sodium silicate is sprayed into the gases as they are exhausted in
the furnace stack. Water from the solution evaporates in the gas stream and
the sodium silicate forms a small sticky sphere which can react chemically with
N0x' S0x* and Physically with particulates. These spheres can then be collected
and recycled into the glass batch .
The quantity of N0x from a glass-melting tank was studied by Kitayama,
C ^31
et al. , to evaluate methods for reduing fuel consumption under photo-
chemical smog warnings. A glass-melting furnace, (of unknown glass composition)
with a 154.2 Gg/day (170 ton/day) capacity using preheated air at 1100C, emitted
850-1000 ppm of N0x< By varying the damper opening and reducing the excess air
by 10 percent, the N0x emissions were reduced to 480 ppm. When the excess air
was reduced 20 percent, the NO emissions were reduced to 45 ppm.
(44)
Takasaki developed a method for removing NO from flue gases by
wet oxidation and absorption. This technique appeared to eliminate more than 90
percent of the NO from the flue gas of a glass-melting furnace. By using acti-
3
vated carbon and chlorine acid soda, a pilot plant with 51 kg m /hr reduced its
NO emissions by 95 percent. This system consisted of a special liquid-gas con-
tact tower that utilizes a chlorine dioxide and chlorine oxidizing agent known
as Fujinon-Ox to convert No into N02 which was absorbed by a liquid and stabilized.
The exiting gas contained no NO. < 10 ppm NO,,, < 5 ppm S00, no chlorine
«» tm
oxide, chlorine, or hydrogen chloride, 13 percent C02, 3.5 percent 02, and
0.03 mg/kg m of dusts. Other details were not reported.
Kanematsu reports on scrubbers handling 377, 7.1, and 28.6 kg m /hr
of SO in the flue gas. By using a wet or dry desulfurization method whereby
X
the sulfur oxides are absorbed by NaOH solutions and oxidized in air, the SO
f / a.\
can be recovered as mirabilite. Kanematsu1 also suggests use of low sulfur
fuels, high stacks, and stack-gas desulfurization systems as method for con-
trolling SO emissions.
X
Efficiency of Equipment
Least effective of the air-pollution control devices is the wet scrub-
ber*405 . In addition to being subject to numerous malfunctions and breakdwons, they
have been found to exhibit particulate-collection efficiencies as low as 66 percent(40)
-------
86
to as high as 90 percent (if gain loadings are low). By fitting the column
with impingement plates, efficiency can range up to 95 percent with particles as
small as 5 microns . A major advantage of this system is its ability to remove
acidic gases.
Baghouses have a reputation for high efficiency and dependability.
Fabric filters are capable of > 99 percent efficiencies and can collect particu-
lated to below 0.75 micron ' . Major disadvantages are that exhaust
gases must be pretreated to remove gaseous emissions and must be cooled before
they contact the low-temperature fabrics.
Electrostatic-precipitator performance is highly sensitive to tempera-
ture and volume fluctuations. Electrical characteristics of particulates, which
affect collection efficiency, vary with temperature, humidity, SCL content, and
the type of particulate. Conventional ESP's have been shown to have efficiencies
up to 95 percent and collect particulates doewn to submicron size. The NAPCO ESP
3
on the other hand, has a reported outlet loading of less than 0.046 g/std m
(0.02 grains/scf) . For an uncontrolled emission rate of 1 kg of particulate/i
(2 Ib of particulate/ton) glass and an air flow of 3119 std m /Gg (100,000 scf/to:
the efficiency will be 85 percent. For an emission rate of 10 kg/Gg (20 Ib/ton),
the efficiency will be greater than 98 percent. This ESP is designed so addi-
tional sections can be added and efficiencies greater than 99 percent can be
, , ,(30,47)
obtained . (
Wet or dry desulfurization methods, presently in use by one glass
company in Japan, have shown respective efficiencies of better than 97 and
80 to 90 percent for the wet or dry SO removal^ .
A
Forming and Finishing
As the glass leaves the forehearth of the melter, it is normally
cut into "gobs" by a pair of mechanical shears. Chutes direct the gobs
from the feeder into blank molds where it is formed by one or two methods.
The glass can also be cats, drawn, or rolled after it exists from the fore-
hearth. The gob is usually pressed or blown into its final shape.
As discussed in an earlier section, a wide variety of forming and
finishing steps may be employed, depending upon the product desired. These
steps can include surface treatment with a metal chloride, fire-polishing with
-------
37
an oxygen-gas flame, decorating with enamels or organic base colors, and coating
with an organic material. All of the glass is heat-treated for purposes of
crystallizing the glass when appropriate and annealing thermally induced strains
from the glass.
Forming Emissions
Molds on forming machines, gob shears, and delivery chutes are lubri-
cated with solutions ranging from grease and oils to graphite and silicone-based
emulsions. During the past decade, silicone emulsions and water-soluble oils
have replaced some grease and oil lubricants on gob shears and gob-delivery
systems . Grease and oils are still used on molds and cause white smoke
emulsions during flash vaporization of the swab. Although the smoke dissipates
in a few seconds, hydrocarbon vapors are released. These emissions are released
inside the plant since hoods are not used to vent the hydrocarbons outside.
Source tests indicate the rate of emission for hydrocarbons is 0.06 g/kg. Total
annual emissions for the industry are calculated to be 0.23 Gg.
Forming and Finishing Control Technology
Efforts to control the hydrocarbons emissions have centered on finding
lubricants capable of withstanding high temperature (1100 C [2200 F]) without voli-
talizing. Use of silicone emulsions and water-soluble oils (90 to 150 parts of
water to 1 part oil or silicone) can eliminate these emissions. Unfortunately,
they have not performed well as mold-release compounds^ . Emissions from the
forming machinery are dispersed within the plant and exhausted by the room ventila-
ting systems. No manufacturers have been identified as using a control device for
these emissions.
Decorating
Emissions. Hydrocarbon emissions from organic solvents and binders
used in coatings on containers are released when decorative coatings are cured in
annealing lehrs. A worst case emission rate for these hydrocarbons is 4.5 g/kg.
-------
38
Control Technology. Process modifications are difficult difficult
to accomplish without harming the quality of the coating . In addition,
they do not completely eliminate hydrocarbon emissions. Several such changes
involve the substitution of solvents and a reduction of solvent concentration
in the coating. Hydrocarbon emissions can be controlled by incineration, absorption
(48)
(activated charcoal or silica gel), or condensation
Frosting of Electric Light Bulbs
Emissions. Hydrofluoric acid (HF) and ammonia (NH,) emissions occur
in the frosting of electric light bulbs. The controlled emission rate is esti-
mated to be 0.96 g/kg of HF and 0.22 g/kg of NH_, assuming an 80 percent efficiency
of control equipment.
Frosting Control Technology. Scrubbers are used to control emissions
from these operations. Efficiencies are reported to be on the order of 80 to 90
(24)
percent .
Acid Cleaning
Emissions. In certain segments of the pressed and blown glass industry,
acid cleaning (sulfuric acid and hydrofluoric acid) is done to prepare parts for
further processing and HF fumes are generated. Using available water pollution
data^ , the emission rate for HF is estimated to be 0.18 g/kg.
Control Technology. Scrubbers are utilized in this area and reportedly
(24)
operate at an efficiency of 80-90 percent .
Surface Treatment
Emissions. Emission from the coating of glass products with tin or
titanium tetrachloride include both particulates (tin chloride, tin oxide) and
gases (tin compounds, HCL, C^). Chlorine and unreacted metal chloride are released
-------
89
into the atmosphere. The emission rate is estimated to be 0.02 g/kg of metal
oxides, 0.03 g/kg of hydrated metal chlorides, and 0.02 g/kg of HC1. Exhausted
particulates are generally composed of submicron-sized tin chloride and tin
oxide.
(49)
Coating Control Technology. One patent (U.S. 3,789,109) has been
issued for an apparatus to be used for cleaning solid, liquid, and gaseous
pollutants from a hot-end coating station of a glass manufacturing plant. Glass
is coated with an external metallic coating to reduce breakage. Because most of
the anhydrous stannous chloride used does not adhere to the glass but discharges
through the air-exhaust system, a potential pollution problem is created. In this
apparatus, the air is heated until the metallic chlorides disassociate to metallic
oxides and hydrogen chloride gas. Exhaust gases are then sprayed with fresh water
to cool th- stream with the water reacting with the hydrogen chloride to form
hydrochloric acid. Exhaust air passes into a scrubber in which the pollutants
are removed.
-------
90
SECTION VI
FUTURE PRODUCTION OF PRESSED AND BLOWN GLASSWARE
The pressed and blown segment of the glass Industry produces a diverse
and always changing spectrum of glass products. Portions of the industry manu-
facture products for direct consumer use (e.g., tableware and artware) while
other portions manufacture products key to other industries (automotive, elec-
tronic, medical, etc.).
Future production is tied very much to the general growth of the
economy. For instance, recent downturns in the purchase of television sets has
resulted in severe curtailment in that portion of the industry which produces
lead glass. The projected growth rates for pressed and blown glass is estimated
to be between 3 and 4 percent through 1980
The shortage of natural gas and the allocation of petroleum products
have placed some constraints on production, which would have been more severe if
the economy were not in a somewhat depressed state. The industry has historically
been very dependent on the use of natural gas. Oil is the normal replacement
fuel, for which the industry does not have an historical use pattern.
The industry is research oriented and many new products exist today
which were still in the laboratory ten years ago. Fiber optics is one such
product, which potentially could replace all major communications lines within
the next decade.
-------
91
REFERENCES
(1) Schorr J. R. and Anderson, G. A., "Final Report on Industrial Energy
Study of the Glass Industry to FEA and DoC", Battelle Columbus Labora-
tories, Contract No. 14-01-0001-1667, pp 80-142 (1974).
(2) Current Industrial Reports, Series M32E (74)-B (May 1975).
(3) Chemical and Process Tech. Encyclopedia. Ed. Considine, 551-561 (1968).
(4) Tooley, F. V., "Raw Materials", Handbook nf p.iaM Man.,fapi-..r0] Vol. 1,
Books for Industry, New York (1974), Chap. 2.
(5) Danielson, J. A., Air Pollution Engineering Manual. 2nd Edition, EPA
Publication No. AP-40 (May 1970).
(6) Anon, "A Screening Study to Develop Background Information to Determine
the Significance of Glass Manufacturing", prepared by Research Triangle
Park Institute for EPA, Contract No. 68-02-0607-Task 3 (December 1972).
(7) Lillis, E. J., and Young, D. "EPA Looks at 'Fugitive Emissions'", J.
Air Pollution Control Assoc., 25_ (10), 1015-18 (1975).
(8) Air Pollution. Vol. 1, Edited by A. C. Stern, 2nd Edition, Academic Press,
N.Y. (1968), "Noriviable Particles in the Air", (M. Corn), 49-52.
(9) Anon, National Emission Data System, Environmental Protection Agency Re-
search Triangle Park, North Carolina (1974).
(10) Ryder, R. J. and McMackin, J. J., "Some Factors Affecting Stack Emissions
from a Glass Container Furnace", The Glass Industry, 50 , 307-11, 346-350,
(June 1969).
(11) Anon, State-by State Listing of SourceTypes That Exceed the Third Decision
Criteria, Special Project Report, Monsanto Research Corp., Contract
68-02-1874, 1-3 (1975).
(12) Arrandale, R. S., "Air Pollution Control in Glass Melting", Symposium Sur
La Fusion du Verre, Brussels (October 1968), 619-644.
(13) Reed, R. J., "Combustion Pollution in the Glass Industry", The Glass In-
dustry, 54 (4), 24-26, 36 (1973).
(14) Information supplied by large manufacturer of pressed and blown glassware
for this study.
(15) A D Little Inc., "Development of Methods for Sampling and Analysis of
Particulate and Gaseous Fluorides from Stationary Sources", EPA, NTIS:
PB 213313, November 1972.
-------
92
(16) Robinson, J. M., et a1.,'Engineering and Cost Effectiveness Study of
Fluoride Emissions Control1,' Vol. 1; NTIS: PB 207506, Office of Air
Programs, Environmental Protection Agency, January 1972.
(17) Anon, "Symposium on Pollution, Stratford-Upon-Avon, 30 Hay-1 June 1973",
Glass Technology, JL (6), 140-144.
(18) Arrandale, R. S., "Pollution Control in Fuel Fired Tanks", The Glass
Industry, J55 (12), 12 ff (August and November 1974).
(19) Stockham, John D., "The Composition of Glass Furnace Emissions",
Journal of the Air Pollution Control Assoc., 2JL (11), 713-715 (1971).
(20) Custer, W. W., "Electrostatic Cleaning of Emissions from Lead, Borosili-
cate, and Soda/Lime Glass Furnaces", presented at the 35th Annual Con-
ference on Glass Problems Ohio State University (Nov. 14-15, 1974).
(21) Teller, A. J., "Control of Emissions from Glass Furnaces", Ceramic Bul-
letin 51, 637-640 (1972).
(22) Frantz, C. J., Miser, D. L., Troy, H. N., and Stabbe, E. D., collected
papers from the 32nd Annual Conference on Glass Problems, Dept. of
Ceramic Engineering, University of Illinois, 25-38 (1971).
(23) Personal communication with Glass Containers Manufacturer Institute
(GCMI).
(24) Development Document for Proposed Effluent Limitations Guidelines and
New Source Performance Standards for the Pressed' and Blown Glass Seg-
ment of the Glass Manufacturing Point Source Category, U.S. Environ-
mental Protection Agency, EPA 440/1-74/034, August 1974.
(25) Turner, D. B., Workbook of Atmospheric Dispersion Estimates. EPA Pub-
lication No. AP-26 (1970), Figures 3-9.
(26) Recommended Guide for the Prediction of the Dispersion of Airborne Ef-
fluents . Edited by M. Smith, ASME (1968).
(27) Climatic Atlas of the United States. U.S. Dept. of Commerce (1968).
(28) Reznik, R. B., Source Assessment: Flat Glass Manufacturing Plants.
EPA Environmental Protection Technology Series, Monsanto Research
Corp., Dayton, Ohio (Oct. 1975).
(29) Bauer,W. C., Tooley, F. V., and Manring, W. H., "Batch Materials Hand-
ling and Preparation", The Handbook of Glass Manufacture, 1, 57-94
(1974).
(30) Rymarz, Ted M., and Lipstein, David H., "Removing Particulates from
Gases", Chemical Engineering Deskbook. 82 (21), The McGraw-Hill Pub-
lishing Co., New York 113-129 (October 1975).
-------
(31) Swift, P.^, "Dust Control Related to the Bulk Delivery of Particulate
Materials", The Chemical Engineer. 143-150 (March 1975).
(32) Edmundson, J. N., Rietz, L., Weise, R. L., and Fraas, J. Collected
papers from the 32nd Annual Conference on Glass Problems, Dept. of
Ceramic Engineering, University of Illinois, 39-54 (1971).
(33) Information supplied by large manufacturer of pressed and blown glassware
for this study.
(34) Arrandale, R. S., "Furnaces, Furnace Design, and Related Topics",
Handbook of Glass Manufacture. Vol. 1, Books for Industry, New York
(1974), Section 5, 249-387.
(35) Hibscher, William, Stertz, R., "The U.S. Glass Industry's Challenge
in These Energy Critical Times", presented at 35th Annual Conference
on Glass Problems, The Ohio State University, 85-101 (November 1974).
(36) Bartz, D. V., KVB Engineering, Inc., Control of Oxides of Nitrogen
from Stationary Sources in the South Coast Air Basin of California,
California State Air Resources Board (1974).
(37) Penberthy, L. , "Recent History of Electric Melting of Glass", Glass
Industry (5), 12-13 (1973).
(38) Teller, A. J., "Control of Glass Furnace Emissions", Glass Industry,
.57 (2), 15-19, 22 (February 1976).
(39) Keller, G. , "Scrubber System Lightens Load of Glass Furnace Emissions",
Chemical Processing, 3Q, 9 (Jan. 1975).
(40) Simon, Herbert, and Wiliamson, E., "Control of Fine Particulates from
Continuous Melting Regenerative Glass Furnaces", presented at the 68th
Annual Meeting of the APCA, Boston, Massachusetts (June 15-20, 1975).
(41) Private Communication, United McGill Corp., Columbus, Ohio.
(42) Mahoney, W. P., "Method for Controlling Furnace Emissions", U.S. Pat.
3,789,624 (1974).
(43) Kitayama, Hiroshi, Hideo, Hayashi, Sataro, Iwasaki, Tadashi, Fujimura,
Tomohiko, Mujano, Hideaki, Murayama, Tomihiro, Myuhata, "Effect of Com-
bustion Conditions on Nitrogen Oxides Formation of Furnaces", presented
at Japan, Soc. Air Pollution, 14th annual meeting, Fukushima, Japan,
(Nov. 1973).
(44) Takasaki, Shoichi, "Flue Gas Denitration by Wet Oxidation and Absorption",
Heat Management Pollution Control, 24 (1), 57-62 (Jan. 1974).
(45) Kanematsu, Jado, "Air Pollution Control in Glass Industry", Seramikk
(Ceramics), 9 (1), 49-55 (Jan. 1974)
-------
B
APPENDIX A
GEOGRAPHICAL LISTING OF
PRESSED-AND-BLOWN GLASS PLANTS
-------
TABLE A-l. GEOGRAPHICAL LISTING OF THE 176 PRESSED AND BLOWN GLASS PLANTS
State
Plant
City
County
AQCR
County
Population
Density,
persons/km^
(persons/mi^)
Furnaces
Arkansas
California
Colorado
Connecticut
De laware
Florida
Illinois
Scott Depot Glass Co.
Thomas Ind., Inc.
Arrowhead Puritas
Brock Glass Co.
Glass Works, Inc.
Owens Illinois
Pacific Glass Works
Potters Ind., Inc.
Ray-Lite Glass, Inc.
Pikes Peak Glass Co.
Rocky Mountain
Thermos Div., King
See ley Thermos Co.
»
Thermos Div. , King
See ley Thermos Co.
Kaufman Glass Co.
Big Pine Glass
Erie Glass Mfg. Co.
Johnson Glass
Owens Illinois
Peltier Glass Co.
Reha Glass Co.
Sellstrom Mfg.
Fort Smith
Fort Smith
Gardena
Santa Ana
Huntington Beach
City of Industry
Huntington Beach
Anaheim
South Gate
Colorado Springs
Durango
Norwich
South Windsor
Wilmington
Big Pine Key
Park Ridge
Chicago
Chicago Heights
Ottawa
Chicago
Palatine
Sebastian
Sebastian
Los Angeles
Orange
Orange
Los Angeles
Orange
Orange
Los Angeles
El Paso
LaPlata
New London
Hartford
New Castle
Monroe
Cook
Cook
Cook
LaSalle
Cook
Cook
17
17
24
24
24
24
24
24
24
38
14
41
42
45
50
67
67
67
71
67
67
57
57
662
696
696
662
696
696
662
41
4
127
427
340
19
2197
2197
2197
37
2197
2197
(148)
(148)
(1715)
(1803)
(1803)
(1715)
(1803)
(1803)
(1715)
(106)
(10)
(329)
(1105)
(881)
(49)
(5690)
(5690)
(5690)
(96)
(5690)
(5690)
4
18
1
2
6
6 5
t
2
2
2
6
8
-------
ID
TABLE A-l (Continued)
State
Indiana
Kentucky
Louisiana
Maryland
Massachusetts
Plant
Canton Glass
Corning Glass Works
Indiana Glass Corp.
Indiana Glass Corp.
Owens-Illinois
St. Clalr Glass Works
Sinclair Glass
Corning Glass Works
Corning Glass Work
GTE Sylvania, Inc.
General Electric
General Electric
Venesian Art Glass
Owens-Illinois
Owens-Illinois
American Optical
American. Optical
Atlantic Optical Moulding
Emerson & Curaing
GTE Sylvania, Inc.
GTE Sylvania, Inc.
City
Hartford City
Bluf fton
Dunkirk
Dunkirk
Warsaw
El wood
Hartford City
Danville
Harrodsburg
Versailles
Lexington
Sonuiierset
Catlcttsburg
Slireveport
Elkton
Southbridge
Southbridge
Dudley
Canton
Danvers
Ipswich
County
Blackford
Wells
Jay
Jay
Kosciuko
Madison
Blackford
Boyle
Mercer
West ford
Fayctte
Pulaski
Boyd
Caddo
Cecil
Worcliester
Worchester
Worchester
Norfolk
Essex
Essex
AQCR
76
87
76
76
82
76
76
102
102
102
102
105
103
22
114
118
113
118
119
119
119
County
Population
Density,
persons/km^
(persons/mi2)
34
25
23
23
34
117
34
43
24
28
239
20
123
97
56
162
162
162
594
493
493
(88)
(65)
(60)
(60)
(89)
(303)
(88)
(111)
(62)
(73)
(619)
(52)
(319)
(251)
(145)
(420)
(420)
(420)
(1538)
(1277)
(1277)
Furnaces
11
9
9
4
11
7
Michigan
B & J Optical
Lincoln Park
Wayne
123
1702 (4408)
-------
TABLE A-l (Continued)
State
Mississippi
Missouri
New Hampshire
New Jersey
New York
Plant
Cataphote
General Electrick '
Pittsburgh Corning Corp.
GTE Sylvania
Chemglass, Inc.
Fischer & Porter Co.
Friedrich & Dummock, Inc.
Masdcn, Inc.
National Glass & Plastics
Owens-Illinois
Owens-Illinois
Potters Ind., Inc.
Thermal American Fused Quartz
Westinghouse Electric Corp.
Wheaion Ind.
American Optical
Bausch & Lorab
Corning Glass Works
Eastman Kodak Co.
Gillinder Bros.
Kosslcr, Inc. (Warren)
Super Glass Corp.
Swift Glass Corp.
City
Jackson
Jackson
Sedalia
Greenland
Newficld
Vine land
Millville
N. Bergen
Newfleld
Pennsauken
Vine land
Carlstadt
Montville
Bloomf ield
Millville
Buffalo
Rochester
Corning
Rochester
Port Jervis
Long Island
Brooklyn
Elicira
County
Hinds
Hinds
Pettis
Hillsboro
Gloucester
Cumberland
Cumberland
Hudson
Gloucester
Caniden
Cumberland
Bergen
Morris
Essex
Cumberland
Erie
Monroe
Steuben
Mon roe
Orange
New York
Kings
Chemung
AQCR
5
5
138
121
45
149
149
43
45
45
149
43
43
43
149
161
159
163
159
160
43
43
163
County
Population
Density,
persons/km^
(persons/mi2) Furnaces
92
92
19
76
203
92
92
4907
203
797
92
1463
314
2672
92
406
404
27
404
101
25,337
14,132
94
(238) 6
(238)
(50) 3
(197)
(525)
(238)
(238)
(12,709)
(525)
(2065)
(238)
(3789)
(313)
(7154) 3
(238)
(1052)
(1046)
(70)
(1046)
(262) 5
(65,623) 2
(36.602) 15
(243)
North Carolina
Potters Ind., Inc.
Apex
Johnston
166
29
(75)
t"
-------
TABLE A-l (Continued)
State Plant
Ohio Anchor Hocking
Anchor Hocking
Anchor Hocking
Anchor Hocking
Brady Co., E.G.
Cambridge Glass
Corning Glass Works
Crystal Art Class
Federal Glass Co.
General Electric
General Electric
General Electric
General Electric
General Electric
Guernsey Glass Co.
Holophane
Imperial Glass Corp.
Interpace Corp.
I.abino Glass Labs.
Lancaster Glass Corp.
Owens-Illinois
Owens-Illinois
Potters Ind., Inc.
RCA Corp.
Rodefer-Gleason Glass Co.
Techniglass, Inc.
Variety Glass, Inc.
Oklahoma Corning Glass Works
Bartlett-Collins
Overmyer-Perram
City
Bremen
Canal Winchester
Lancaster
Lancaster
Cleveland
Cambridge
Greenville
Cambridge
Columbus
Bucyrus
Cleveland
Logan
Miles
Niles
Cambridge
Newark
Bcllaire
Tiffin
Grand Rapids
Lancaster
Columbus
Toledo
Cleveland
Circleville
Bellaire
Newark
Cambridge
Muskogee
Sapulpa
Tulsa
County
Fairfield
Fairfield
Fairfield
Fairfield
Cuyahoga
Guernsey
Darke
Guernsey
Franklin
Crawford
Cuyahoga
Hocking
Trumble
Trumble
Guernsey
Licking
Belmont
Scioto
Wood
Fairfield
Franklin
Lucas
Cuyahoga
Pickaway
Belmont
Licking
Guernsey
Creek
Tulsa
AQCR
176
176
176
176
174
183
173
183
176
175
174
182
178
178
183
176
181
103
124
176
176
124
174
176
181
176
183
186
186
County
Population
Density,
persons/km^
(persons/mi^)
55
55
55
55
1441
28
32
28
591
48
1441
19
144
144
28
60
58
49
55
55
591
539
1441
30
58
60
28
19
268
(142)
(142)
(142)
(142)
(3732)
(73)
(83)
(73)
(1531)
(124)
(3732)
(49)
(373)
(373)
(73)
(155)
(150)
(127)
(142)
(142)
(1531)
(1396)
(3732)
(78)
(150)
(155)
(73)
(48)
(964)
Furnaces
8
3
1
6
3
2
9
5
6
5
1
2
2
2
2
-------
TABLE A-l (Continued)
State Plant
West Virginia Beaumont Co.
Blenko Glass Co.
Brockway Glass
Brockway Class
Champion Agate Co.
Colonial Class Co.
Corning Class Works
Co I'u ing Class Works
Corning Class Works
Crescent Class Co.
n.ivis-I.yp.ch Glass Co.
Ill itc Co. , Inc.
Erj;!-: ine Ola:;:, & Mfg.
Funtori Art Class
Fostoria Class Co.
Gentile Glass Co.
Gentile- Glass Co.
Cl.uM ing-Vitio-/\gate Co.
llarton H.uv.icr.iCted Glass
Harvey 1ml.
Kunauha Glass Co.
l.fwis County Glass
Louie Glaus Co.
Marble King
Kid-Atlantic
Minners
Pennyboro Glass Co.
Pilgrim Glass Corp.
Rainbow Art Class
ScanJia Class Works
Seneca Class Co.
Earl Shelly Glass Co.
Viking Class Co.
Viking Class Co.
West Virginia Glass Specialty
West ingliouse Electric
The Paul Wissmach Class Co. Inc.
City
Morgantown
Milton
Clarksburg
Parkersburg
Penniiboro
West on
Paden City
Mart insburg
Parkersburg
Wei Isburg
Star City
Cameron
Wellsburg
Wi) 1 iamstown
Mounsville
Star City
Star City
Parkcrstmrg
Dunbar
Clarksburg
Dunbar
Jane Lew
West on
Paden City
Ellenboro
Salem
Puntiuboro
Caredo
Hunt ington
Kunova
Morgantown
Huntington
Hun ting ton
New Martainsville
Weston
Fa i rmon t
Paden City
County
Monongalia
Cabell
Harrison
Wood
Ritche
Lewis
Tyler
Wood
Hrooke
Monongalia
Marshall
Brooke
Miago
Marshall
Monongal la
Monongalia
Wood
Kanawha
Harrison
Kanawlia
Lewis
Lewis
Tyler
Ritchie
Harrison
Ri tchie
Cabel 1
Cabel
Wayne
Monongalia
Cabell
Cabell
Wetzel
Lewis
Rir ion
Tyler
AQCR
235
103
235
179
232
232
179
179
131
235
181
161
236
181
235
235
179
- 234
235
234
232
232
179
232
235
232
103
103
103
235
103
103
179
232
235
179
County
Population
Density,
persons/km^)
(persons/mi^)
66
144
66
88
8
17
15
88
129
66
47
129
29
47
66
66
88
95
66
95
17
17
15
8
60
8
144
144
28
66
144
144
21
17
75
15
(171)
(373)
(171)
(228)
(21)
(44)
(39)
(228)
(334)
(171)
(122)
(334)
(75)
(122)
(171)
(171)
(228)
(246)
(171)
(246)
(44)
(44)
(39)
(21)
(171)
(21)
(373)
(373)
(73)
(171)
(373)
(373)
(54)
(44)
(194)
(39)
Furnac
4
6
1
5
1
1
2
10
12
6
11 9
3 J
2
12
3
7
8
1
3
9
1
2
1
8
7
6
6
A
3
Wisconsin
Pope Scientific, Inc.
Menomonee Falls
Waukesha
238
161
(418)
-------
TABLE A-l (Continued)
State Plant
Pennsylvania Corning Class Works
Corning Class Works
Corning Class Works
Corning Glass Works
Fischer S, Porter Co.
General Electric
Haley Glass Co.
llouze Glass Corp.
Jeannette Corp.
Jeannette Shade & Novelty Co.
Kopp Class, Inc.
Lenox Crystal
Mayflower Glass
Hillstein, J. H.
Owens- I 1 1 inois
Pennsylvania Glass Products
Phoenix Class Co.
Pittsburgh Corning Corp.
Schott Optical Class Co.
L. E. Smith Glass Co.
Victory Glass Co.
Westmoreland Glass
Rhode Island Corning Glass Works
Texas EMC Glass Corp.
Multicolor Glass
Potter Ind. , Inc.
Virginia Corning Class Works
Washington Nuclear Pacific
Penbarthy Electromelt Int., Inc.
City
Bradford
State College
Wei Isboro
Charleroi
Warminster
Bridgeville
Creensburg
Point Marion
Jennnelte
Joanne tee
Pittsburgh
Mt. Pleasant
La t robe
Jeannette
Pittston
Pittsburgh
Monaca
Port Allegheny
Uuryea
Mt. Pleasant
Jeannette
Grapevillt;
Central Falls
Decatur
San Antonio
Brownwood
Danville
Seattle
Seattle
County
McCain
Centre
Tioga
Bucks
Allegheny
Westmoreland
Fayette
Westmoreland
Westmoreland
Allegheny
Westmoreland
Westmoreland
Westmoreland
Luzerne
Allegheny
Beaver
McKean
Luzerne
Westmoreland
Westmoreland
Westmoreland
Providence
Wise
Bexar
Brown
Pittsylvania
King
King
AQCR
195
151
45
397
197
197
197
197
197
197
197
197
151
197
197
178
151
197
197
197
120
. 215
217
210
222
229
229
County
Population
Density,
persons/km
(persons/mi^)
34
13
261
842
142
75
142
142
842
142
142
142
147
842
180
20
147
142
142
142
534
8
258
10
259
207
207
(88)
(34)
(676)'
(2181)
(369)
(194)
(369)
(369)
(2181)
(369)
(369)
(369)
(386)
(2181)
(466)
(52)
(386)
(369)
(369)
(369)
(1383)
(21)
(668)
(26)
(671)
(536)
(536)
Furnaces
1
2
2
4
1
4
5 i
<^
2
8
3
7
5
7
2
3
3
-------
TABLE A-l. (Continued)
State
Plant
City
County
AQCR
County
Population
Density
persons/km
(persons/mi^)
Furnaces
West Virginia
(Continued)
Pennsboro Glass Co.
Pilgrim Glass Corp.
Rainbow Art Glass
Gcandia Glass Works
Seneca Glass Co.
Earl Shelly Glass Co.
Viking Glass Co.
Viking Glass Co.
West Virginia Glass Specialty
Westinghouse Electric
The Paul Wissmach Glass Co. Inc.
Pennsboro
Caredo
-Hunt ing ton
Kenova
Morgantown
Hunt ing ton
Huntington
New MartainsvilLe
West on
Fairmont
Paden City
Ritchie
Cabell
Cabel
Wayne
Monongalia
Cabell
Cabel
Wetzel
Lewis
Marion
Tyler
232
103
103
103
235
103
103
179
232
235
179
8
144
144
28
66
144
144
21
17
75
15
(21)
(373)
(373)
(73)"
(171)
(373)
(373)
(54)
(44)
(194)
(39)
1
8
7
6
6
4
3
Wisconsin
Pope Scientific, Inc.
Menomonee Falls Waukesha
238
161 (418)
-------
APPENDIX B
EMISSIONS DATA
-------
B-l
APPENDIX B
EMISSIONS DATA
Raw Materials Preparation and Handling
Five typical points for particulate emissions have been considered for
the raw materials preparation and handling operations: (1) unloading and
conveying, (2) crushing of cullet (scrap glass), (3) filling and emptying of
storage bins, (4) weighing and mixing of batch, and (5) feeding of batch to
glass melting furnace (batch charging). Source test data are summarized in
Table B-l.
Emissions from the raw materials preparation and handling operation
consist entirely of particulates from raw materials. In practice, only
fugitive dust emissions should be considered, since particulate emissions
remaining within the plant may constitute an OSHA health and safety considera-
tion distinct from plant emissions. As discussed in the text, only particles
below 100 micron are considered as contributing to fugitive dust emissions.
Actual measurements of plant emissions from these operations are not avail-
able; however, personal observation indicates that there are no visible
emissions from the batch house. Measurements of particulate emissions within a
(9)
few plants are available from NEDS and have been used to determine particulate
emissions on a worst-case basis.
The average emission factors for the various raw material prepara-
tion and handling operations were taken to be the following, calculated on
a worst-case basis.
I
mg/kg
1. Handling 1500 + 100 %
2. Crushing <0.1
3. Storage 100 + 100%
4. Mixing 310 ±100%
5. Charging <0-1
1910 + 100%
-------
TABLE B-l. SUMMARY OF SOURCE TEST DATA FOR MATERIALS
PREPARATION AND HANDLING(a)
Particulate Emissions
Plant Mg/yr (Tons/yr)
1(C) 1.80 (2)
2 12.70 (14)
3(C) trace(b>
4 56.20 (62)
5 9.10 (10)
6 trace (b)
Production
Gg/yr
0.28
12.70
45.90
52.16
3.22
45.9
(Tons/yr)
(306)
(14,000)
(50,600)
(57,500)
(3,560)
(50,600)
Rate
rag/kg
(Ib/ton)
6.54
(13.07)
1.00
(2)
0
1.08
(2.16)
2.83
(5.62)
0
Control Equipment Operation
fabric filter raw material
handling
none raw material
handling
w
wet-scrubber mixing
none mixing
fabric filter mixing
fabric filter mixing
(a) Source NEDS^ - All for flint glass.
(b) Trace <1.0 T/yr.
(c) Data from source test measurements.
-------
B-3
Total annual emissions were based on 3.37 Tg of raw materials being
processed to melt 1.39 Tg of gia8S, assuming that 48 percent of glass
melted produced a saleable product. Furthermore, a 20 percent weight loss
during melting, such as in the decomposition of Na2C03, was also assumed.
Stack heights for these and other plant operations were listed
in Appendix C. They range from 2 m (6.5 feet) to 36 m (118 feet).
The accuracy is obtainable only for batch mixing where the
sample mean is 1.0 mg/kg and the sample standard deviation is 1.1 mg/kg.
The 95 percent confidence level is + 1.78 mg/kg. The accuracy of
engineering estimates are assumed to be + 100 percent.
Glass Melting
Nitrogen Oxides
(Q)
Measurements of NO emissions from NEDS are listed in Table B-2.
X
Emissions factors vary from 0.41 to 10.00 g/kg (0.81 to 20.00 Ib/ton), which
clearly reflect the wide range of operating conditions found in glass melt-
ing furnaces. The average emission factor of 4.25 g/kg (8.49 Ib/ton) is
based upon source test measurements from furnaces melting soda-lime glasses.
Since the type of glass is not expected to significantly affect the NO
emission rate, this emission factor is assumed to be representative of the
entire industry. The average emission factor for soda-lime glass was
determined by adding the average emission factors together and dividing
by the number of values. Because of the sparcity of data, various point
source measurements are used. Alternatively, the average is found by dividing
the total emissions by total production was 2.19 g/kg (4.74 Ib'/ton). The
difference is not significant because the standard deviation is 3.18 g/kg,
and the 95 percent confidence level is + 1.84 g/kg.
Standard deviations were determined by the following method.
1/2
-------
TABLE B-2. NO EMISSIONS FROM PRESSED AND BLOWN GLASSWARE FURNACES
x
Production
Soda-Lime
1(a)
2 (a)
3 (a)
4 (a)
5 (a)
6 (a)
7
-------
B-5
where: n = number of samples
X. = sample value
U = sample mean.
The confidence interval (CI) was determined by:
where: t = "Student's t" variable for n-1 degrees of freedom.
Sulfur Oxides
(Q)
Source test measuremnts of SO emissions from NEDS ' are
x
given in Table B-3. Only 5 measurements are available: 3 for soda-
lime glass and 2 for borosilicate glass. The values for borosilicate must
be due to oil firing since sulfate materials are not used to make
this glass. The emission factors vary from 0.54 to 5.44 k/kg
(1.09 to 10.87 Ib/ton). The average emission factor of 2.80 g/kg
(5.61 Ib/ton) is based upon all of the measurement data since, in
general, SO emissions will not be dependent upon the type of glass
Jv
being melted. For the values given the standard deviation is
calculated to be 1.41 g/kg, with the 95 percent confidence level
+1.75 g/kg. The dependence of SO emissions on fuel oil used
^^" *»
instead of natural gas is not clearly defined, but the increased
use of oil, or other sulfur bearing fossil fuel, is expected
to increase both the rate and amount of S0x emissions.
-------
TABLE B-3. SO EMISSIONS FROM PRESSED AND BLOWN GLASSWARE FURNACES
Production
Soda-Lime
l(a>
2
3
TOTAL
Borosilicate
l
2(a)
TOTAL
TOTAL FOR ALL
Gg/yr
126
65
8
199
3
3
6
205
(tons/yr)
(139,000)
(72
(8
(219
(3
(3
(7
(226
,000)
,000)
,000)
,680)
,680)
,360)
,360)
Emissions
Mg/yr
9
217
9
267
1
18
19
287
.98
.7
.98
.6
.81
.14
.96
.58
(tons/yr)
(11)
(240)
(11)
(295)
(2)
(20)
(22)
(317)
Emission Factor
g/Kg
2
3
1
2
0
5
2
2
.95
.33
.75
.68
.54
.44
.99
.80
(Ib/ton)
(5
(6
(3
(5
(1
(10
(5
(5
.90)
.67}
.50)
.36)
.09)
.87)
.98)
.61)
(average)
(average)
(average)
(a) Source test measurements.
-------
B-7
Particulates
Source test measurements obtained from NEDS^ are listed in
Table B-4. Emission data are only available from furnaces melting soda
lime and lead glasses. Only a single source test measurement was avail-
able for the lead glass. No data are available for borosilicate or opal
glasses. Emission factors for soda-lime varied from 0.49 to 12.57 g/kg
(0.97 to 25.14 Ib/ton). The average emission factor for soda lime is
5.22 g/kg (10.44 Ib/ton). The standard deviation is 4.7 g/kg and the
95 percent confidence interval is + 2.25 g/kg.
Based on data supplied by glass manufacturers, ' worst-case
engineering calculations were made for borosilicate, opal, and lead
glasses. The highest emission factor for borosilicate was taken as 25 g/kg
(50 Ib/ton), for opal glass 5 g/kg (10 Ib/ton), and for lead glass 15 g/kg
(30 Ib/ton). The accuracy was taken as + 100 percent.
An overall emission factor for particulates was taken to be
8.7 g/kg. It was determined as a weighted average of each of the above
emission factors and the percentage of each type of glass melted.
Carbon Monoxide
Measurements of carbon monoxide emissions from glass melting
furnaces are scarce, since this is not a major emission. High combustion
temperatures and the presence of excess air do not favor its formation.
It can form in glass melting furnaces because of incomplete combustion
within long diffusion flames used to effect uniform heat release over the
molten glass. When combustion is properly controlled, emissions are negligible.
The three-source test measurementsW available are for soda-lime glass.
They are listed in Table B-5. The emission rate is not expected to vary much for
other glass types. The average emission factor is 0.10 g/kg (0.19 Ib/ton).
The standard deviation is 0.045 g/kg with a 95 percent confidence level of 0.10
g/kg.
-------
TABLE B-4. PARTICULATE MEISSIONS FROM PRESSED AND BLOWN GLASSWARE FURNACES
Production
Soda-Lime
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
TOTAL
Lead
1
Gg/yr
38
59
78
8
8
10
8
16
4
30
15
29
13
18
5
37
48
65
7
496
5
(tons/yr)
(42,000)
(64,900)
(85,200)
(8,570)
(8,570)
(13,500)
(3,570)
(17.5JO)
(4,380)
(32,900)
(16,400)
(32,000)
(14,000)
(13, 4 JO)
(5,090)
(41,200)
(52,500)
(72,000)
(8,000)
(545,220)
(5,090)
Emissions
Mg/yr
38.1
145.2
38.1
49.9
99.8
74.4
92.5
31.8
49.9
62.6
8.2
44.5
158.8
14.5
20.9
194.1
71.7
127.0
26.3
1357.2
2.09
(tons/yr)
(42)
(160)
(42)
(55)
(110)
(82)
(102)
(35)
(55)
(69)
(9)
(49)
(175)
(16)
(23)
(214)
(79)
(140)
(29)
(1496)
(23)
Emission Factor
g/kg
1
2.4
0.49
6.29
12.57
7.84
11.66
2.00
12.56
2.10
0.55
1.53
12.50
0.82
12.56
5.19
1.50
1.94
3.62
5.22
4.52
(Ib/ton)
(2.00)
(4.93)
(0.97)
(12.57)
(25.14)
(15.69)
(23.31)
(4.00)
(25.11)
(4.19)
(1.10)
(3.06)
(25.0)
(1.65)
(25.11)
(10.39)
(3.01)
(3.89)
(7.25)
(10.44) (average)
(9.04)
00
-------
TABLE B-5. CO EMISSIONS FROM PRESSED AND BLOWN GLASSWARE
(a)
Production '
1
2
3
TOTAL
Gg/yr
78
12
37
127
(tons/yr)
(86,200)
(13,600)
(41,200)
(141,000)
Emissions
Mg/yr
4.5
1.8
2.7
9.0
(tons/yr)
(5)
(2)
(3)
(10)
Emission Factor
g/kg
0.06
0.15
0.09
0.10
(Ib/ton)
(0.12)
(0.29)
(0.15)
(0.19) (average)
W
VO
(a) Soda Lime Glass
-------
B-10
Hydrocarbons
Source test measurements of hydrocarbon emissions are also limited.
Formation occurs for the same reasons as cited for carbon monoxide. These
emissions are listed in Table B-6. Emission factors vary from 0.02 to
0.27 g/kg (0.05 to 0.55 Ib/ton). The average emission factor is 0.15 g/kg
(0.31 Ib/ton). The amount of hydrocarbon emission is not expected to be
significantly affected by the type of glass melted. The standard deviation
is 0.09 g/kg, and the 95 percent confidence interval is + 0.15 g/kg.
Fluorides
No source test data was available for fluoride emissions from
glass melting furnaces (opal, borosilicate and lead glasses). Based upon
(33)
information supplied by a glass manufacturer and the open literature
(15, 16, 17) a worst_case emission factor of 10 g/kg (20 Ib/ton) of fluoride
(as F~) was assumed. The accuracy was taken as + 100 percent.
Selenium
No source test measurements are available for selenium emissions
from soda lime glass furnaces. Selenium is used as a decolorizer to
neutralize the tint from transition metal oxide contaminants such as iron.
' /
Approximately 0.36 Gg (395 tons) of selenium are consummed annually in the
U.S., of which an estimated 5 percent (0.0186 g) is used by the pressed
and blown glass industry (0.06 Gg). Using a worst-case assumption, half of
the selenium used is volatilized and emitted from the glass melting furnace,
giving an emission rate of 0.002 g/kg. The accuracy of this calculation,
is taken as + 100 percent.
Forming and Finishing
Few point source test measurements are available on emissions from
the forming and finishing operations. Therefore, engineering calculations
considering worst-case situations are used to determine the severity of
emissions from these parts of the manufacturing process.
-------
TABLE B-6. HC EMISSIONS FROM PRESSED AND BLOWN GLASSWARE
(a)
Production v '
Gg/yr
1 78
2 30
3 17
4 25
5 17
6 37
7 13
TOTAL 217
(tons/yr)
(86,200)
(32,900)
(18,200)
(27,400)
(18,200)
(41,200)
(13,600)
(237,700)
Emissions
Mg/yr
8.2
4.5
4.5
4.5
4.5
0.9
0.9
28.1
(tons/yr)
(9)
(5)
(5)
(5)
(5)
(1)
(1)
(31)
Emission Factor
g/kg
0.10
0.15
0.27
0.18
0.27
0.02
0.07
0.15
(Ib/ton)
(0.21)
(0.30)
(0.55)
(0.37)
(0.55)
(0.05)
(0.15)
(0.31) (average)
(a) Soda Lime Glass
-------
B-12
Forming
During forming, an emulsion containing oil or silicone and water
is sprayed onto the molds, gob shears, and delivery chutes. From 1 to 3 g
of liquid are sprayed into a mold each time an article is formed. The
oil:water mixture is normally 1:125.
(a)
Three measurements were reported for hydrocarbon emissions
from forming operations, and were listed in Table B-7. The average emission
rate was 0.06 g/kg (0.11 Ib/ton, and is higher than that determined by
(52)
engineering calculation (0.035 g/kg) for forming glass containers . For
an emission factor of 0.06 g/kg, the standard deviation is calculated to
be 0.26 g/kg, with a 95 percent confidence interval of + 0.06 g/kg.
Treatment
Assuming that 25 percent of all pressed and blown glassware produced
receives a surface treatment to improve resistance to scratching and to
facilitate handling, by subjecting the glass to a tin or titanium chloride
vapor, emissions will consist of metal oxide, hydrated metal chloride particulates
and HC1. Approximately 60 weight percent of the total metal chloride input is
released (14% metal oxide, 27% hydrated metal chloride and 21% HC1). Emission
rates are estimated to be 0.02 g/kg (0.03 Ib/ton) of tin chloride, 0.03 g/kg
(0.06 Ib/ton) of hydrated tin chloride, and 0.02 g/kg (0.05 Ib/ton) of HC1.
Accuracy of the data is taken at + 100 percent.
Annealing
No reliable emission data are available for gas-fired annealing
lehrs; therefore, emission factors are estimated from other data on gas
3
combustion. A modern recirculating air-type lehr consumes 11 to 17 m /hr (400
cfh to 600 cfh) when annealing 91 Gg (100 tons of glass per day. Lehrs of older
design can consume 34 to 57 m /hr (1200 cfh to 2000 cfh) . On a worst-case
3 3
basis (57 m /hr would require 0.0062 m /kg of glass produced. For a plant
producing 319 Mg/day (352 ton/day) this would amount to 91 m /hr. With a heating
3
value of natural gas (1000 Btu/cf or 37.3 million joules/m ) this amounts to
-------
tf-iJ
0.93 million joules per second or ahont- n 01 ~,-ii- . , ^ ,
t- OC.-VUU or aoouc u.^j million joules per kg of glass
(200,000 Btu/ton).
Using tests on gas-fired burners(53), emission data was determined
as shown in Table B-8. Converting these on a basis of 0.24 million joules/kg
of glass gave the emission factors for annealing shown in Table B-9.
Decorating
Tableware, artware, and novelties are sometimes decorated with
vitrifiable glass enamels or organic materials. Emissions are derived from
organic solvents and binders used in the coatings. Estimating that 30 percent
of all pressed and blown tableware and art glass have decorative coatings, about
100 Gg (110,200 tons) of glassware are decorated annually. Only one source
measurement is available (0.02 g/kg or 0.05 Ib/ton). Using 4.5 g/kg (9.0 Ib/ton)
as a worst case, the total uncontrolled annual emissions will be 0.45 Gg (496
tons).
Frosting of Light Bulbs
While no information is available on atmospheric emissions from
(24)
frosting operations, data are availablev on water pollution. Waste-
water comes from both rinsing and scrubbing operations. The reported
pollutant level is 9.6 g/kg (1.92 Ib/ton) for fluorides (HF) and 2.2 g/kg
(4.4 Ib/ton) for ammonia (NH,). A worst-case calculation assumes that half
the effluent loading is from the scrubbing water and that the scrubber is
performing at 80 percent efficiency. The amount of frosted light bulbs
produced is estimated at 90.7 Gg (100,000 tons). The controlled air emission
factors are then 0.96 g/kg (1.92 Ib/ton) for HF and 0.22 g/kg (0.44 Ib/ton)
for NH3.
Acid Cleaning
(24)
While no air sampling data are available, information has been reported
on fluorides in wastewater from these operations. The rinse step and scrubber
generate wastewater containing fluorides at a level of 1.8 g/kg (3.6 Ib/ton).
Considering a worst case that half of the fluoride is from the scrubber and that
the scrubber operates at 80 percent efficiency, the controlled air emission rate
for HF will be 0.18 g/kg (0.36 Ib/ton). Assuming total product of picture tubes
is 90.7 Gg (100,000 tons) then the total annual emissions of HF would be 16 Mg (18 tons)
-------
B-14
TABLE B-7. HYDROCARBON EMISSIONS FROM FORMING
PRESSED AND BLOWN GLASSWARE
Production
Case No.
1
2
3
Total
Gg/yr
27
23
1
51
(tons/yr)
(30,000)
(25,000)
(1,600)
(54,600)
Emissions Emission Factor
Mg/yr
4.5
trace(a)
(a)
traceU;
4.5
(tons/yr) g/kg
(5) 0.17
0
0
(5) 0.06
(Ib/toti)
(0.33)
(0 . 11) (average
(a) <0.01 T/yr.
-------
APPENDIX C
STACK HEIGHTS FROM THE VARIOUS
SEGMENTS OF GLASSMAKING PROCESS
-------
C-l
TABLE C-l. TYPICAL STACK HEIGHTS OF BATCH-
HANDLING OPERATIONS FOR SODA/LIME GLASS
<40 m >40 m
No. of Stacks
1
2
1
1
1
1
2
1
Height, m No. of Stacks Height, m
2
5
9
13
14
23
24
36
Total No. of Stacks 10
Average 16
Median 14
Total No. of Stacks 0
Average 0
Median 0
-------
C-2
TABLE C-2. TYPICAL STACK HEIGHTS FOR MELTING OPER-
ATIONS OF GLASS FURNACES
<40 m >40 m
No. of Stacks
1
1
1
6
6
2
2
1
4
3
1
2
5
1
2
1
1
1
Height, m No. of Stacks Height, m
Soda Lime
8 8 41
9 1 44
12 3 46
13 1 ' 49
14 1 51
15 2 53
16
17
20
21
22
23
24
27
32
33
35
37
Total No. of Stacks 40
Average 19
Median 20
Total No. of Stacks 16
Average 45
Median 44
-------
C-3
TABLE C-2. (Continued)
<40 m
>40 m
No. of Stacks
Height, m
No. of Stacks
Height, m
2 11
1 21
2 27
L 35
Total No. of Stacks 6
Average 22
Median 23
Borosilicate
7 41
1 49
Total No. of Stacks 8
Average 42
Median 45
Lead
Total No. of Stacks
Average
Median
No. of Stacks
Average
Median
0
0
0
46
19
23
Total No. of Stacks 7
Average 42
Median 43
Total Industry
No. of Stacks
Average
Median
31
44
47
-------
C-4
TABLE C-3. TYPICAL STACK HEIGHTS FOR FORMING OPERATIONS
<40 m
>40 m
No. of Stacks
Height, m
No. of Stacks
Height, m
1
1
13
15
Soda Lime
41
Total No. of Stacks 2
Average 14
Median
Total No. of Stacks 1
Average 41
Median 41
Lead
2
1
1
1
15
21
28
38
Total No. of Stacks 5
Average 23
Median 21
Total No. of Stacks 0
Average 0
Median 0
-------
C-5
TABLE C-4. TYPICAL STACK HEIGHTS FOR ANNEALING
OPERATIONS OF BOROSILICATE GLASS
<40 m >40 m
No. of Stacks Height, m No. of Stacks Height, m
1 12
Total No. of Stacks 1 Total No. of Stacks 0
Average 12 Average 0
Median
12 Median
-------
C-6
TABLE C-5. TYPICAL STACK HEIGHTS FOR DECORATING
OPERATIONS OF SODA/LIME GLASS
<40 m
>40 m
No. of Stacks
Height, m
No. of Stacks
Height, m
1
1
11
12
Total No. of Stacks 2
Average 12
Median -
Total No. of Stacks 0
Average 0
Median 0
-------
C-7
TABLE C-6. TYPICAL STACK HEIGHTS FOR TREATMENT
OPERATIONS OF SODA/LIME GLASS
<40 m >40 m
No. of Stacks Height, m No. of Stacks Height, m
13
Total No. of Stacks 3 Total No. of Stacks
Average 13 AveraSe
Median 13 Median
-------
APPENDIX D
STATE LISTING OF TOTAL
EMISSIONS AS OF 1972
-------
D-l
TABLE D-l. STATE LISTING OF TOTAL EMISSIONS AS OF 1972
State
i ALASAPA
t ALASKA
» ARIZONA
» ARKANSAS
9 CALIFORNIA
4 COLORADO
T CONNECTICUT
DELAWARE:
* FLORIDA
10 6COR6IA
11' HAWAII
<2 IDAHO
1J ILLINOIS
1» INDIANA
IS IOWA
' It KANSAS
IT KENTUCxr
Mass of emissions, 1000 kg/yr (upper entry)
Percent of U.S. totals (lower entry)
Partic-
ulate
2002000. U
1.53000
163*0000.0
12.50000
3265000.0
2.«*OOU
1619000.0
1.21000
5675000.0
*. 33000
3156000.0
2.11000
365600.0
0.27900
130200.0
0.09930
2*30000.0
1.66000
2331000.0
1.76000
251200.0
0.19200
2*30000.0
1.65000
1581000. D
2.7*000
2202000.0
1.66000
2579000.0
1.97000
3356000.0
2.56000
1854000.1)
I.H2000
S02
1220UOO.O
i.nooo
22*600.0
0.30700
200200.0
0.31100
205*00.0
0,31900
2557000.0
3.9SOOO
*73300.0
0.73600
1227000.0
1.41000
*2070o.o
0.6SSOO
1755000.0
2,73000
163SUOO.O
2.5*000
232000.0
0.36100
511*0.0
O.U9200
3711UOO.O
5,76000
3036000.0
*. 72000
397*00.0
0.61600
225000.0
0.35000
1627000.0
2.53000
NO
X
26160U.O
2.27000
31990.0
0.27700
75100.0
0.6t>lUO
77310.0
0.67UOO
796800.0
6.91000
1166UO.O
1.010UO
152200.0
1.32000
15720.0
0.396UO
110300.0
3.S6JOO
29*200.0
2.SSUOO
*0790.0
0.3S*OC
33220.0
0.26600
66S10U.O
5.77000
*1**00.0
3.59UOO
137700.0
1.14000
109900.0
0.9b3UO
3020UU.O
2.623UO
Hydro-
carbons
3*2100.0
. 1.29000
1*0600.0
0.53200
171100.0
U. 6*700
261700.0
1.07000
191*000.0
7.2*000
29**00.0
1.110UO
259*00.0
0.96100
77510.0
0.29300
536200.0
2,03000
526700.0
1.99000
62720.0
0.23709
163600.0
0.619UO
13*3000.0
5.06000
675100.0
2.550UU
100600.0
1.52000
7*2600.0
2.81000
27*600. U
1.0100U
CO
372*30.0
2.0*000
*?22l>0.0
2.59000
179300.0
O.TfctJO
2?5800.0
1.2*000
19870UO.O
10.90000
105600.0
0.57900
92690.0
o.sa^oo
2*580.3
0.13503
3*02030.0
19.20000
705*00.0
3.66000
6*^50.0
0.16100
MS300.0
2.61000
«12500.0
2.2*000
162100.0
0.9-»700
90720.0
0,»97nfl
i;«6uo.o
0.95600
?193UO.O
1.20000
-------
D-2
TABLE D-l. (Continued)
State
18 LOUISIANA
19 MAINf.
*fl HARTLAND
HI- HASSACMUSE'TS
Hi MICHIGAN
M HINNtSOTA
ZH nississzppi
« MISSOURI
26 MONTANA
27 NEBRASKA
*8 NEVADA
n NEW HAMPSHIRE
40 ftCw JCRSET
»i MCW MEXICO
*2 NCM YORK
13 N CAROLINA
*» N DAKOTA
« OHIO
46 OKLAHOMA
47 ORC50N
Mass of emissions, 1000 kg/yr (upper entry)
Percent of U.S. totals (lower entry)
Partic-
ulate
1651000.0
1.260(10
1038000.0
0.79200
657300.0
0.90200
802700. 3
0.61300
2601000. 0
2.14000
3056000. 0
2.53000
1190000.0
1.1HOOO
2639000.0
2.17000
0975000. 0
3.60000
10*9000.0
2.33000
3155000.0
2.H1000
326500.0
0.24900
815600.0
0.62300
3548000.0
2.71000
2704000.0
2.0600U
2203000.0
1.68000
2851000.0
2.16000
3054000. U
2.33000
2276000.0
1.74000
2885000.0
2.2UOOO
SO2
sasaoo.o
c.niuo
771700. U
1.20000
1352UOO.O
7.10000
3640000.0
5. 97000
351SUOO.O
5.46000
846600.0
1.32000
280300.0
0.43600
1259000.0
1.96000
177000.0
0.27500
137100.0
0.21300
263100.0
0.40900
125800.0
0.50700
2922000.0
n.SSOOO
441400.0
0.66700
5137000.0
7,99000
2298000. 0
3.56000
379700.0
0.51100
4002000. 0
6.32000
163400.0
0. 25400
i7?SOO.O
11.S7-JOO
NO
X
21900U.O
1.900UO
54270.0
0.47UOO
2151UU.Q
1.66000
322300.0
2.79UOO
54800U.O
4.7&000
185000.0
1.60000
87010.0
0.75*00
267500.0
2.49000
34650.0
0.3UUOO
5094U.O
0.44200
56500.0
0.50700
3606U.O
0.31300
323400.0
2. 81)000
109800.0
0.9b^UO
721400.0
6.25000
336400.0
2.93000^
61110.0
0.53000
765800.0
6.010UO
1300UU.O
I.I3U08
(.2710.0
0.544UO
Hydro-
carbons
1741000.0 '
6.5DOOO
71970.0
0.27200
302300.0
1.14000
463100.0
1.75000
734000.0
2.760UO
386000.0
1.47000
350200.0
1.32000
586400.0
2.22000
174200.0
0.65600
255600.0
0.96600'
36140.0
0.13700
44430.0
0.16600
786600.0
2.97000
310200.0
1.170UO
1353000.0
5.11000
465100.0 .
1.76000
73930.0
U. 26000
1244000.0
4.700UO '
674700.0
2.55000
204800.0
0.774UO
CO
"399UO.O
M.bUOOO
61430.0
0.33600
1634110. 0
0.8*400
1904UO.O
1.04003
?99400.0
1.64000
150700.0
0,62500
?282UO,0
1,25000
268500.0
1.47000
23USCO.O
1.26000
59590.0
0.32600
26700.0
0.15700
302UO,0>
0.16500
281400.0
1.54000
494UO.O
0.2710.0
551600.0
3.02000
J71SOO.O
2.03090
22320.9
0.12200
462700.0
2.64000
200AOO.O
1.1UOOO
J04900.0
l.t'OOO
-------
D-3
TABLE D-l. (Continued)
State
it PENNSYLVANIA
. 99 KHOOl ISLAND
»fl SJTAHOLIM4
HI S DAKOTA
12 TENNLSSCt
IS TEXAS'
<* UTAH
IS VCRHONT
1* . VIRGINIA
17 WASHINGTON
16 M VIHGINIA
H WISCONSIN
90 WYOMING
US TOTALS
Mass of emissions, 1000 kg/yr (upper entry)
Percent of U.S. totals (lower entry)
Partic-
ulate
3132000. U
2.34000
113200.0
0.06600
1209000.0
0.923UO
2861000.0
2.18000
1789000.0
1.37000
9302000.0
7.10000
2161000.0
1.66000
292100.0
0.22300
1607000.0
1.23000
220*000. 0
1.66000
1261000.0
0.96200
2180000.0
1.66000
2851000.0
2.16000
131000000.0
S02
5603000.0
H. 72000
519900.0
0. 110900
1076UOO.O
1.67000
b9»i>0.0
0.10800
13071)00. 0
2.03000
1617000.0
2.63000
295*00.0
O.HHtOO
112bOO.O
3.17500
1366000.0
2.16000
626*00.0
0.97500
mssuoo.o
2.26000
1216UQO.O
1.69000
513000.0
0.79600
6*300000.0
NO
X
782200.0
f .78UUO
AA760.0
0.33600
1*6300.0
1.27000
10560.0
0.161UO
26*109.0
2.29000
(15500.0
6.03UUO
*6«10.0
O.U2UOQ
13710.0
0.11900
197800.0
1.71000
1263UO.O
_ 1.09000
306500,0
2.660UO
231300.0
2.00000
70570.0
0,61200
moooou.o
Hydro-
carbons
1331000.0
S.UJOOO
93730. U
U.35*UO
260500.0
0.965UO
$1110.0
0.3**00
3*0900.0
1.29QOO
*139000.U
15.60000
112800.0
O.*2600
25*60.0
0.0963U
15200.0
1.57000
361800.0
1.37000
172600.0
0.65300
362600.0
1.370UO
275200.0
1.0*000
26*00000.0
CO
527000.0 .
2,80000
29.190,1
1.16100
6J900.0
2.65000
21*80.0
0.129(30
?003UO.O
1.10000
»*013t.3,0
8. 2*000
*6e*o.o
0.25600
1*190.0
0.07770
233 J 00.1
1.29000
025500.0
2.33000
03S1UO.O
2.36000
161300.0
0.86300
20870.0
0,11*00
lasoocoo.o
-------
APPENDIX E
Conversion Factors
-------
E-l
TABLE E-l. CONVERSION FACTORS
To Convert From
Btu
degree Fahrenheit (F)
foot (ft)
3 3
foot (ft )
inch (in.)
mile (mi)
pound (mass, Ib)
ton (short)
Prefix Symbol
tera T
giga G
mega M
kilo k
mill! m
micro M
To
joule (J)
degree Celsium (C)
meter (m)
meter (m )
meter (m)
2 2
meter (m )
kilogram (kg)
gigagram (Gg)
PREFIXES
Multiplication
Factor
io12
109
io6
io3
ID'3
io-6
Multiply By
1.055 x IO3
t°c = (t°p - 32)/1.8
3.048 x 10"1
2.832 x 10~2
2.540 x 10"2
2.590 x IO6
4.536 x 10"1
9.072 x 10"4
Example
1 Tg = 1 x IO12 g
1 Gg = 1 x 10 g
1 Mg - 1 x IO6 g
1 km = 1 x IO3
1 mm = 1 x 10 m
1 ym = 1 x 10 m
------- |