United States Office of Air Quality EPA-450/5-82-005
Environmental Protection Planning and Standards August 1982
Agency Research Triangle Park NC 27711
Air
Preliminary Study
of Sources
of Inorganic
-------
-------
EPA-450/5-82-005
Preliminary Study of Sources
of Inorganic Arsenic
by
RADIAN Corporation
3024 Pickett Road
Durham, North Carolina 27705
EPA Contract Number: 68-02-3513
Task Number: 18
Project Officer. Warren D. Peters
U.S. EfTVfroBnwrtat Protection Agency,
Region V, Library
230 South Dearborn
Cbteaga. Winols 6Q604,
Prepared for:
U.S ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Pollutant Assessment Branch
Research Triangle Park, North Carolina 27711
August 1982
-------
DISCLAIMER
This report is issued by the U. S. Environmental Protection
Agency to provide information on sources of atmospheric inorganic
arsenic emissions. This report was furnished to EPA by Radian
Corporation, 3024 Pickett Road, Durham, North Carolina 27701 in
partial fulfillment of EPA Contract Number 68-02-3513, Assignment
18. The contents of this report are reproduced herein as received
from Radian Corporation. The opinions, findings and conclusions
expressed are those of the author and not necessarily those of the
EPA. A mention of company or product names is not to be considered
as an endorsement by the EPA.
U.S. Envfeonmemel Protection Agency
-------
PREFACE
On June 5, 1980 EPA listed inorganic arsenic as a hazardous air
pollutant under §112 of the Clean Air Act. This action was based on the
Administrator's findings that there is a high probability that exposure to
inorganic arsenic causes cancer in humatis, and that there is significant
public exposure to inorganic arsenic which is emitted into the air by
stationary sources. Now EPA must determine priorities for detailed studies
of source categories for possible development of standards. Major factors
considered by EPA in determining priorities include the following:
- the economic and technical feasibility of significant
improvements in emission control,
- the ease of expeditious development and implementation
of standards,
- the degree of public exposure associated with the
emissions and locations of the sources within each
category,
- the impacts of other Federal regulatory programs, and
the economic, energy, and environmental impacts of
requiring emission reductions beyond existing requirements.
This report provides preliminary information on seven source categories
of inorganic arsenic that will assist EPA in the development of priorities
for detailed studies for possible development of standards. Any such
standard would be in the form of a National Emission Standard for Hazardous
Air Pollutants (NESHAP) and would follow the procedures, including public
participation, for developing a NESHAP.
Information on source categories of inorganic arsenic other than the
seven source categories studied in this report can be found in earlier
reports such as the SRI report "Human Exposure to Inorganic Arsenic
(May, 1980)" and the PEDCo report "Extended Source Survey Report for Arsenic
(May, 1982)" (see reference list in Chapter 1 of this report). The exposure
analysis for the seven source categories studied in this report is found in
a companion report "Preliminary Study of Sources of Inorganic Arsenic -
Exposure Analysis (August, 1982)."
iii
-------
ABSTRACT
The seven source categories investigated as emitters of inorganic
arsenic are as follows: ,
- primary copper smelters (exclusive of ASARCO-Tacoma),
- primary lead smelters,
primary zinc smelters,
- secondary lead smelters,
- cotton gins handling arsenic acid-desiccated cotton,
- glass manufacturing, and
- arsenic chemical production.
These source categories were examined to determine the magnitude of
arsenic emissions corresponding to the current or existing level of control,
the magnitude of arsenic emissions resulting from compliance with all
applicable regulations, and the magnitude of arsenic emissions resulting
from the use of estimated best control (EEC) for arsenic. In the report,
the term baseline is used for all source categories to denote the lesser of
arsenic emissions that are currently occurring or the arsenic emissions that
would be expected if the source was complying with all applicable regula-
tions. These regulations include State implementation plans (SIPs),
Occupational Safety and Health Administration (OSHA) standards, arxd New
Source Performance Standards (NSPS). A potential EEC for each source
category was selected based on whether the emitted arsenic was primarily in
a particulate or gaseous form, or in a combination of both forms, and
whether the technology had been demonstrated as an effective method of
arsenic control within the subject or a similar source category. Once an
EEC option was selected, the objective was to determine the incremental
arsenic emissions reduction achievable over baseline by implementing EEC.
The resultant costs of the EEC increment were also determined.
The arsenic emissions and control data in the report were obtained from
published technical papers and reports, EPA data files on arsenic, and from
the subject companies within each source category. No source testing was
performed for this study. In instances where no arsenic emissions data were
available, estimates were prepared using the assumptions and calculations
given in Appendix A of this report. All data on the cost of arsenic control
iv
-------
equipment were derived from published cost curves in. economic and technical
reports. All cost bases and assumptions used in this report are detailed in
Appendix B.
The primary copper smelting category includes 14 plants located in
seven States that emit inorganic arsenic from both process and fugitive
sources. For many of the process sources, Radian's EEC for arsenic control
is already being used as a result of the existing particulate matter and
sulfur dioxide (SO ) regulations the smelters must comply with. Electro-
static precipitators (ESP's) and contact sulfuric acid plants are used for
particulate and SCL control, respectively. To reduce the occurrence of
arsenic vapor emissions, gas coolers prior to particulate control devices
are recommended as a requirement of EEC for arsenic. Fugitive arsenic
emissions from tapping and materials handling are currently controlled by
wet suppression and local hooding and ventilation to a particulate control
device or atmosphere. For arsenic fugitives from the converter sources air
curtains and secondary hooding systems appear to be viable EEC options.
Fugitive arsenic emission sources are not controlled to EEC levels in any of
the primary copper smelters. Agreements are currently being reached between
OSHA and individual smelters that will result in improvements in fugitive
emission control systems. Copper smelting source category arsenic emissions
under baseline control are estimated to be 233 Mg (256 tons)/yr. The
implementation of EEC is estimated to reduce total arsenic emissions to
119 Mg (131 tons)/yr. To implement arsenic EEC, the copper smelting
industry would experience an economic impact of varying proportions based on
plant size and current control status. To maintain the current copper
industry net present values (NPV), unit product price increases ranging from
0.6 to 4.6 percent would be required.
Currently there are five domestic primary lead smelters, three in
Missouri and one each in Montana and Texas. Primary lead smelters have
process and fugitive arsenic emission sources, with fi'gitive sources posing
the most significant air pollution problem. All process arsenic emissions
sources are controlled by either contact sulfuric acid plants or low
temperature (110°C) fabric filters, as dictated by existing State and
Federal regulations. For this reason arsenic EEC is assumed to be in place
for process arsenic emission sources in the primary lead smelting industry.
Fugitive arsenic emissions from primary lead smelters will be affected by
the National Ambient Air Quality Standard (NAAQS) for lead, impending lead
SIP's, and OSHA standards for lead and inorganic arsenic. It is assumed
that when compliance with standards is achieved, fugitive arsenic emissions
will be controlled to the level of EEC. Total arsenic emissions from
primary lead smelting are estimated to be approximately 43 Mg (47 tons)/yr
if EEC control is achieved industrywide.
The primary zinc smelting source category consists of five zinc metal
smelters and two zinc oxide smelters that emit arsenic. The four zinc metal
smelters that use electrolytic smelting techniques emit arsenic only during
the roasting step. All of these plants use contact sulfur acid plants to
-------
control roaster SCL emissions which concurrently control arsenic to the EBC
level. The one electrothermal zinc metal smelter also uses an acid plant
for its roaster emissions, therefore, EBC is in place at all roaster
sources. Electrolytic zinc plants have no arsenic air emissions other than
from the roaster. The sinter machine and smelting furnace arsenic emissions
from the electrothermal smelter are controlled by a low temperature (110°C)
fabric filter, which represents EBC for these emission points. In the zinc
oxide plants roaster emissions are also controlled by contact sulfuric acid
plants, therefore, EBC is in place for these sources. Other process offgas
streams are controlled by low temperature fabric filters such that arsenic
EBC is also in place at the majority of these sources. There is, however, a
proce.ss fabric filter system at one zinc oxide smelter that requires cooling
of the inlet stream before EBC can be in place. The current fabric filter
operating temperature of 140°C (280°F) needs to be reduced below 11.0°C
(230°F) to limit losses of vaporized arsenic trioxide. At the EBC level,
arsenic emissions from the zinc smelting source category are estimated to be
1.2 Mg (1.3 tons)/yr. Arsenic emissions from the zinc source category have
predominantly been controlled as a result of SIP and NSPS regulations.
The secondary lead industry consists of 66 plants that are located in
26 States and all ten EPA regions. Arsenic is emitted by secondary lead
smelters from process, process fugitive, and area fugitive sources. The
predominant sources of process emissions are the smelting furnaces and the
refining pots. Process fugitive emissions are generated during charge
preparation, furnace operations, and refining operations. Area fugitive
sources include raw material storage piles and the charge preparation area.
Arsenic emissions are currently being controlled to some extent because of
the smelters' obligation to comply with existing regulations for lead and
particulate emissions. Fabric filters and wet scrubbers are primarily us.ed
to achieve compliance with the NSPS and SIP's that are applicable to
secondary lead smelters. These devices have proven effective at controlling
process lead emissions, and are anticipated to be equally as effective at
controlling arsenic due to the similar behavior of the two compounds. A
fabric filter/scrubber configuration was selected as EBC for secondary lead
process emissions. The major arsenic emission problem from secondary lead
smelters appears to result from fugitive sources. The full implementation
and enforcement of the OSHA lead standard and the NAAQS for lead are
predicted to have positive effects on arsenic emissions reductions froir.
fugitive sources. Once all applicable standards are met by the secondary
lead industry, EBC for fugitive sources should be in place. The use of EBC
control in the secondary lead smelting industry, for process and fugitive
sources, is expected to reduce arsenic emissions by 22.4 Mg (24.6 tons)/yr.
The implementation of EBC in the secondary lead smelting industry would
result in a unit product price increase of 1.1 to 1.6 percent in order to
maintain current industry NPV.
The cotton ginning source category includes only gins that handle
cotton which has been desiccated with arsenic acid. These types of gins are
located primarily in Texas and Oklahoma and number about 320. The applied
vi
-------
arsenic acid spray forms a residue on cotton plant leaves, stems, and burs
which are emitted as gin trash during the seed cotton ginning process.
Cyclone collectors are used to control the trash, and consequently arsenic
emissions from a gin. Because gin operations are agricultural and seasonal
in nature they are not rigidly regulated as air pollution sources. For
these gins, EEC control was determined to be a secondary arrangement of
cyclones following those presently in use. The particular type of secondary
cyclones recommended are known as "long-cone" cyclones. Baseline arsenic
emissions from cotton gins are estimated to be about 3.2 Mg (3.5 tons)/yr
including both process and fugitive sources. Arsenic emissions under the
EEC level are estimated to be 2.4 Mg (2.6 tons)/yr for process and fugitive
sources. To implement EBC and maintain their NPV, cotton ginners would be
required to increase their unit product price by 0.5 to 2.1 percent,
depending on gin size.
The glass manufacturing source category consists of glass plants that
use arsenic trioxide or arsenic acid as a batch raw material for the glass
melting process. This study identified 15 domestic glass plants that were
using and emitting arsenic compounds. The high temperatures of the melt
furnace vaporize the amount of arsenic raw material that is not complexed in
the molten glass mixture. The vaporized arsenic is entrained in the furnace
flue gas and is emitted along with furnace particulate and S0? emissions.
Particulate emissions from glass plants are controlled primarily by ESP's to
comply with State regulations. Fugitive arsenic dust emissions from raw
materials handling are highly controlled due to OSHA regulations for
airborne inorganic arsenic in the workplace. EPA tests have demonstrated
that arsenic control efficiencies of 93 to 99 percent are achievable with
existing particulate control devices in use at glass plants. For glass
plant arsenic emissions EBC is defined as an ESP or fabric filter preceeded
by a gas cooling chamber (to reduce arsenic vapor losses). At current
regulatory levels, arsenic emissions from the 15 glass plants are estimated
to be 51 Mg (56 tons)/yr. If EBC for arsenic were implemented in all plants
arsenic emissions are estimated to be about 2.9 Mg (3.2 tons)/yr. The
implementation of EBC would require a unit product price increase, to
maintain glass industry NPV, ranging from 1,1 to 1.7 percent.
Arsenic chemical manufacture consumes approximately 90 percent of the
total arsenic used in the United States. The production of orthoarsenic
acid, monosodium methylarsonate (MSMA), disodium methylarsenate (DSMA),
dimethylarsenic acid (cacodylic acid), and chrome copper arsenate (CCA)
consumes about 90 to 95 percent of the arsenic used for arsenic chemicals.
Seven companies with a total of eight plants are producing these significant
arsenic chemicals and emitting inorganic arsenic. Arsenic emissions occur
when the dry arsenic trioxide raw material is dumped into storage hoppers
and process reactors. Arsenic emissions from the source category are well
controlled due to the value of the recoverable material and the influence of
the OSHA workplace standard for airborne inorganic arsenic. Manufacturers
have installed highly efficient hooding or enclosure systems, ventilation
systems, and particulate control devices to comply with the OSHA standard
vii
-------
and to recover arsenic trioxide dusts that can be used in the process. The
existing controls used at arsenic chemical plants to comply with OSHA
standards represent EEC for arsenic chemical plant air emissions. Total
arsenic emissions from the eight prominent plants are estimated, from data
supplied by the manufacturers, to be about 41 kg (90 lb)/yr.
viii
-------
TABLE OF CONTENTS
Chapter Page
1.0 INTRODUCTION AND SUMMARY 1
1.1 PURPOSE AND SCOPE 1
1.2 SUMMARY OF RESULTS ' 2
1.3 REFERENCES. 5
2.0 ARSENIC PROPERTIES AND CURRENT REGULATIONS AFFECTING
EMISSIONS 6
2.1 PROPERTIES OF INORGANIC ARSENIC COMPOUNDS 6
2.1.1 Arsenic and Air Pollution 7
2.1.2 Air Pollution Control 10
2.2 REGULATIONS AFFECTING THE EMISSIONS OF INORGANIC
ARSENIC 12
2.2.1 Arsenic-Specific Regulations 12
2.2.2 Particulate Matter and Sulfur Dioxide
Regulations 13
2.2.3 National Ambient Air Quality Standards 14
2.2.4 Worker Exposure Standards 14
2.3 REFERENCES 18
3.0 PRIMARY COPPER SMELTING 20
3.1 COPPER SMELTING DESCRIPTION 20
3.1.1 General Information 20
3.1.2 Process Description 20
3.1.3 Arsenic Emissions From Primary Copper Smelters . . 26
3.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS. . 28
3.2.1 Regulatory Impacts 28
3.2.2 Arsenic Removal Capabilities of Existing Control
Equipment 31
3.2.3 Inorganic Arsenic Emissions Under the Regulatory
Baseline 35
3.3 ARSENIC EEC OPTIONS FOR THE PRIMARY COPPER SMELTING
INDUSTRY 35
3.3.1 EEC Option Selection 35
3.3.2 Prevalence of Arsenic EBC Under Existing and
Baseline Levels of Control 38
3.4 INCREMENTAL IMPACT OF EBC OPTIONS 40
3.4.1 Environmental Impacts 41
3.4.2 Energy Impacts 41
3.4.3 Control Cost Impacts 41
3.4.4 Economic Impact Resulting from EBC Options .... 44
IX
-------
Chapter Page
3.5 POPULATION EXPOSURE DATA 44
3.6 EASE OF STANDARDS DEVELOPMENT 44
3.7 REFERENCES 49
4.0 PRIMARY LEAD SMELTING 51
4.1 LEAD SMELTING DESCRIPTION 51
4.1.1 General Information 51
4.1.2 Process Description 51
4.1.3 The Behavior of Arsenic in Primary Lead Smelting . 57
4.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS. . 57
4.2.1 Regulatory Impacts '....... 57
4.2.2 Arsenic Removal Capabilities of Existing Control
Equipment 59
4.2.3 Inorganic Arsenic Emissions Under the Regulatory
Baseline 60
4.3 ARSENIC EBC OPTIONS FOR THE PRIMARY LEAD SMELTING
INDUSTRY 60
4.3.1 Definition of Arsenic EBC 60
4.3.2 Prevalence of Arsenic EBC Under The Baseline
Level of Control 64
4.4 POPULATION EXPOSURE DATA 64
4.5 EASE OF STANDARDS DEVELOPMENT 64
4.6 REFERENCES 67
5.0 PRIMARY ZINC SMELTING 69
5.1 PRIMARY ZINC SMELTING DESCRIPTION 69
5.1.1 Electrothermal Zinc Smelting 71
5.1.2 Electrolytic Zinc Production 75
5.1.3 Zinc Oxide Production 78
5.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS. . 78
5.2.1 Regulatory Impacts 78
5.2.2 Effectiveness of Existing Controls for Arsenic
Removal 80
5.2.3 Magnitude of Arsenic Emissions from the Primary
Zinc Smelting Industry 80
5.3 ARSENIC EBC OPTIONS FOR THE PRIMARY ZINC INDUSTRY .... 82
5.3.1 Definition of Arsenic EBC 82
5.3.2 Prevalence of Arsenic EBC 82
5.4 INCREMENTAL IMPACTS ASSOCIATED WITH THE ARSENIC EBC
OPTIONS 83
5.4.1 Environmental Impacts 83
5.4.2 Energy Impact 83
5.4.3 Control Cost Impacts 83
5.4.4 Economic Impact Resulting from EBC 83
x
-------
Chapter Page
5.5 POPULATION EXPOSURE DATA 85
5.6 EASE OF STANDARDS DEVELOPMENT 85
5.7 REFERENCES 88
6.0 SECONDARY LEAD SMELTING 90
6.1 SOURCE DESCRIPTION 90
6.1.1 Industry Description 90
6.1.2 Process Description 90
6.1.3 Arsenic Emissions from Secondary Lead Smelters . . 100
6.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS. . 106
6.2.1 Regulatory Impacts 106
6.2.2 Existing Controls and Inorganic Arsenic
Emissions 107
6.3 DEFINITION OF EEC OPTIONS 117
6,3.1 Process Sources 117
6.3.2 Process Fugitives 118
6.3.3 Nonprocess Fugitives 118
6.4 DETERMINATION OF THE INCREMENTAL IMPACTS OF EEC 118
6.4.1 Model Plants 118
6.4.2 Incremental Impact of EEC Options for Process
Emissions 121
6.4.3 Economic Impact Resulting from 121
6.5 POPULATION EXPOSURE DATA 124
6.6 EASE OF STANDARDS DEVELOPMENT 124
6.7 REFERENCES 134
7.0 COTTON GINS 137
7.1 SOURCE DESCRIPTION 137
7.1.1 Arsenic Acid Desiccation , 137
7.1.2 Methods of Desiccant Application 139
7.1.3 Factors Affecting Potential Inorganic Arsenic
Emissions Prior to Ginning 140
7.1.4 Inorganic Arsenic Content of Gin Trash 142
7.1.5 Alternatives to Arsenic Acid Desiccation 144
7.1.6 Characterization of Cotton Gins Handling Arsenic
Acid-Desiccated Cotton 144
7.2 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING
REGULATIONS 149
7.2.1 Regulatory Impacts 149
7.2.2 Baseline Inorganic Arsenic Emissions and
Controls 150
7.3 DEFINITION OF EBC OPTIONS 157
7.3.1 EBC Options for Process Emissions 157
7.3.2 EBC Options for Fugitive Emissions 161
xi
-------
Chapter Page
7.4 DETERMINATION OF THE INCREMENTAL IMPACTS OF THE EBC
OPTIONS 163
7.4.1 Environmental Impacts 163
7.4.2 Energy Impacts 163
7.4.3 Control Cost Impacts 163
7.4.4 Economic Impact Resulting from EBC 163
7.5 POPULATION EXPOSURE DATA 167
7.6 EASE OF STANDARDS DEVELOPMENT 167
7.7 REFERENCES 174
8.0 GLASS MANUFACTURING 180
8.1 SOURCE DESCRIPTION 180
8.2 PROCESS DESCRIPTION 182
8.3 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING
REGULATIONS ..... 184
8.3.1 Regulatory Impacts 184
8.3.2 Baseline Inorganic Arsenic Emissions and
Controls 185
8.4 DEFINITION OF EBC OPTIONS 187
8.5 DETERMINATION OF MODEL PLANTS, EBC EMISSIONS, AND THE
INCREMENTAL IMPACT OF EBC 189
8.5.1 Model Plants 189
8.5.2 Incremental Impacts of EBC Option. 191
8.6 POPULATION EXPOSURE DATA 196
8.7 EASE OF STANDARDS DEVELOPMENT 196
8.8 REFERENCES. . 202
9.0 ARSENICAL CHEMICAL PRODUCTION 205
9.1 SOURCE DESCRIPTION 205
9.1.1 Orthoarsenic Acid 207
9.1.2 MSMA, DSMA, and Cacodylic Acid 209
9.1.3 Calcium Arsenate, Lead Arsenate, and Sodium
Arsenite 211
9.1.4 Chrome Copper Arsenate and Ammoniacal Copper
Arsenite 213
9.1.5 Summary of Producers 214
9.2 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING
REGULATIONS 214
9.2.1 Regulatory Impacts 214
9.2.2 Baseline Inorganic Arsenic Controls and
Emissions 214
9.3 DEFINITION OF EBC OPTIONS 219
9.4 POPULATION EXPOSURE DATA 219
9.5 REFERENCES 223
xii
-------
Chapter Page
APPENDIX A - BASES FOR THE CALCULATION OF ARSENIC EMISSIONS
ESTIMATES A-l
APPENDIX B - BASES FOR THE CALCULATION OF ARSENIC EMISSIONS CAPITAL
AND ANNUALIZED CONTROL COSTS B-l
APPENDIX C - METHODOLOGY FOR THE ECONOMIC IMPACT CALCULATIONS ... C-l
xiii
-------
LIST OF TABLES
Table
1-1 SUMMARY OF INFORMATION FOR THE ARSENIC SOURCE
CATEGORIES 3
2-1 ARSENIC TRIOXIDE (ARSENOLITE) VAPOR PRESSURE DATA. ... 8
2-2 ARSENIC CONTENT OF ORE CONCENTRATES SMELTED IN THE
DOMESTIC PRIMARY COPPER, LEAD, AND ZINC INDUSTRIES . . 9
2-3 EPA-PROMULGATED NSPS EMISSION LEVELS FOR PARTICIPATE
AND SULFUR DIOXIDE 15
2-4 NATIONAL AMBIENT AIR QUALITY STANDARDS FOR SULFUR
DIOXIDE 16
3-1 THE DOMESTIC PRIMARY COPPER SMELTERS 21
3-2 ARSENIC CONTENT OF THE FEED TO DOMESTIC COPPER
SMELTERS 22
3-3 DISTRIBUTION OF UNCONTROLLED FUGITIVE ARSENIC EMISSIONS
AT A PRIMARY COPPER SMELTER 29
3-4 SUMMARY OF THE EXISTING EMISSION CONTROL SYSTEMS
CURRENTLY IN-PLACE IN THE PRIMARY COPPER SMELTING
INDUSTRY 30
3-5 EXPECTED BASELINE COPPER SMELTER CONFIGURATIONS AND
EMISSION CONTROL SYSTEMS 32
3-6 ESTIMATED TOTAL BASELINE INORGANIC ARSENIC EMISSIONS
FROM THE PRIMARY COPPER SMELTERS 36
3-7 EBC OPTIONS FOR PRIMARY COPPER SMELTING 37
3-8 PREVALENCE OF ARSENIC EBC FOR PROCESS EMISSIONS UNDER
EXISTING AND BASELINE CONTROL 39
3-9 CAPITAL AND ANNUALIZED COSTS OF EBC FOR PRIMARY COPPER
SMELTERS 42
3-10 REVENUE AND PRICE IMPACTS FOR PRIMARY COPPER SMELTERS
TO ACHIEVE EBC 45
xiv
-------
Table Page
3-11 SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE IMPACTS
FOR PRIMARY COPPER SMELTERS 46
3-12 POPULATION EXPOSURE MODELLING INPUT DATA FOR THE
PRIMARY COPPER SMELTING INDUSTRY 47
4-1 THE DOMESTIC PRIMARY LEAD SMELTERS 52
4-2 ARSENIC ELIMINATION ROUTES AT THE AMAX PRIMARY LEAD
SMELTER IN BOSS, MISSOURI 58
4-3 MEASURED PROCESS ARSENIC EMISSION RATES AT THE ASARCO/
EAST HELENA LEAD SMELTER 61
4-4 MEASURED TOTAL PARTICIPATE, LEAD, AND ARSENIC FUGITIVE
EMISSION RATES AT THE ASARCO/EAST HELENA LEAD
SMELTER 62
4-5 MEASURED TOTAL PARTICULATE, LEAD, AND ARSENIC FUGITIVE
RATES AT THE ASARCO/GLOVER LEAD SMELTER 63
4-6 POPULATION EXPOSURE MODELLING INPUT DATA FOR THE PRIMARY
LEAD SMELTING INDUSTRY 65
5-1 THE PRIMARY ZINC SMELTING INDUSTRY 70
5-2 THE ARSENIC CONTENT OF THE ZINC ORE CONCENTRATES AT
SEVERAL DOMESTIC PRIMARY ZINC SMELTERS 72
5-3 EXISTING AIR POLLUTION CONTROL EQUIPMENT AT THE ST.JOE/
MONACA ELECTROTHERMAL ZINC SMELTER 76
5-4 UPPER-BOUND ESTIMATES OF INORGANIC ARSENIC EMISSIONS
FROM CONTACT SULFURIC ACID PLANTS IN THE PRIMARY
ZINC SMELTING INDUSTRY 81
5-5 TOTAL CAPITAL AND OPERATING COSTS ASSOCIATED WITH THE
INSTALLATION OF ARSENIC EEC AT THE NEW JERSEY ZINC/
PALMERTON ZINC OXIDE PLANT 84
5-6 REVENUE AND PRICE IMPACTS FOR PRIMARY ZINC CATEGORY. . . 86
5-7 POPULATION EXPOSURE MODELLING INPUT DATA FOR THE
PRIMARY ZINC SMELTING INDUSTRY 87
xv
-------
Table Page
6-1 MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES ... 91
6-2 CONSUMPTION OF NEW AND OLD LEAD SCRAP BY SMELTERS AND
REFINERS IN THE UNITED STATED IN 1980 96
6-3 COMPARISON OF FURNACE CHARACTERISTICS 98
6-4 FUGITIVE EMISSIONS SOURCES AND CONTROL METHODS 103
6-5 SECONDARY LEAD MATERIAL BALANCE 105
6-6 EMISSIONS TEST RESULTS FOR SECONDARY LEAD SMELTERS ... 109
6-7 ENRICHMENT RATIOS AND CONTROL DEVICE PERFORMANCE FOR
ARSENIC AND LEAD EMISSIONS FROM COAL-FIRED POWER
PLANTS 110
6-8 ESTIMATION OF ARSENIC EMISSIONS FROM PROCESS SOURCES OF
SECONDARY LEAD SMELTERS 113
6-9 ARSENIC EMISSIONS FACTORS FOR A SMALL SECONDARY LEAD
SMELTER 115
6-10 NONPROCESS FUGITIVE LEAD EMISSIONS AT A TYPICAL LARGE
LEAD SMELTER 116
6-11 MODEL PLANT PARAMETERS 119
6-12 MODEL PLANT EMISSION PARAMETERS 120
6-13 CAPITAL AND OPERATING COSTS FOR CONTROL OF ARSENIC
EMISSIONS FROM SECONDARY LEAD PROCESS SOURCES 122
6-14 NATIONWIDE IMPACTS OF APPLYING EBC TO ALL PROCESS
SOURCES 123
6-15 REVENUE AND PRICE IMPACTS FOR SECONDARY LEAD SMELTERS TO
ACHIEVE EBC 125
6-16 SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE INCREASES
FOR SECONDARY LEAD SMELTERS 126
6-17 EXISTING PLANT LOCATIONS AND CONTROL STATUS 127
7-1 PRODUCTION-RELATED DATA FOR THE COTTON GIN MODEL
PLANTS 146
xvi
-------
Table Page
7-2 BASELINE ARSENIC EMISSIONS FROM COTTON GIN MODEL
PLANTS 153
7-3 ARSENIC EMISSION REDUCTIONS ATTRIBUTABLE TO EEC
OPTIONS 162
7-4 SOLID WASTE IMPACTS FROM COTTON GIN MODEL PLANT
CONTROL 164
7-5 ENERGY IMPACTS FROM COTTON GIN MODEL PLANT CONTROL ... 165
7-6 CAPITAL AND ANNUALIZED COSTS OF COTTON GIN MODEL PLANT
EEC 166
7-7 REVENUE AND PRICE IMPACTS FOR COTTON GINS TO ACHIEVE
EBC 168
7-8 SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE IMPACTS
FOR COTTON GINS 169
7-9 COTTON GIN MODEL PLANTS' ARSENIC EMISSIONS DATA 170
7-10 COTTON GIN MODEL PLANTS' STACK PARAMETER 171
8-1 GLASS MANUFACTURING PLANTS EMITTING INORGANIC ARSENIC. . 181
8-2 ESTIMATED INORGANIC ARSENIC EMISSIONS AND PARTICIPATE
CONTROLS FOR EXISTING ARSENIC GLASS PLANTS 186
8-3 GLASS MANUFACTURING MODEL PLANT PARAMETERS AND
INORGANIC ARSENIC EMISSIONS 190
8-4 ARSENIC GLASS PLANT EMISSIONS UNDER EBC OPTIONS 192
8-5 CAPITAL AND ANNUALIZED COSTS OF EBC FOR GLASS
MANUFACTURING MODEL PLANTS 194
8-6 REVENUE AND PRICE IMPACTS FOR GLASS MANUFACTURERS TO
ACHIEVE EBC 195
8-7 SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE IMPACTS
FOR GLASS MANUFACTURERS 197
8-8 INPUT DATA FOR THE POPULATION EXPOSURE ASSESSMENT FROM
GLASS MANUFACTURING 198
xvii
-------
Table
9-1
9-2
9-3
9-4
9-5
9-6
ARSENIC COMPOUNDS AND THEIR USE
USES OF ARSENICAL PESTICIDES, INSECTICIDES, AND
HERBICIDES
ARSENIC CHEMICAL PLANTS WITH SIGNIFICANT INORGANIC
ARSENIC EMISSIONS
COMPARISON OF EXISTING AND ALLOWABLE INORGANIC ARSENIC
EMISSIONS
SUMMARY OF EXISTING CONTROLS AND INORGANIC ARSENIC
EMISSIONS IN THE ARSENIC CHEMICAL SOURCE CATEGORY. . .
INPUT PARAMETERS FOR THE POPULATION EXPOSURE MODELLING
Page
206
212
215
216
218
ANALYSIS 220
xviii
-------
LIST OF FIGURES
Figure
2-1 Effect of Temperature on the Performance of Arsenic
Emissions Control Devices 10
3-1 Schematic Diagram of a Primary Copper Smelter 23
4-1 Primary Lead Smelting 54
5-1 Fluid Bed Roasting System at the Jersey Miniere Electrolytic
Zinc Plant in Clarksville, Tennessee 73
5-2 Electrothermal Zinc Smelting at the St.Joe/Monaca,
Pennsylvania Plant 74
5-3 A Typical Electrolytic Zinc Smelter 77
5-4 The New Jersey Zinc/Palmerton Zinc Oxide Plant 79
6-1 Typical Blast Furnace System for Secondary Lead
Production 99
6-2 Typical Reverberatory Furnace System for Secondary Lead
Production 101
6-3 Secondary Lead Material Balance 104
6-4 Vapor Pressure of Pb, PbO, Sb, PbS, PbCl2, As,
and As203 7 Ill
7-1 Primary Areas of Arsenic Acid Desiccation 138
7-2 The Transformation Cycle of Arsenic Compounds in the
Environment 143
7-3 Cotton Gin Model Plant 147
7-4 Relative Dimensions for a 2D-2D, Small Diameter Design
Cyclone 152
7-5 Condenser Drum with Screen Wire and Fine Perforated
Metal Coverings 154
7-6 Schematic Diagrams of the Most Prevalent In-line
Filters 155
xix
-------
Figure Page
7-7 Intr-A-VAC^Particulate Matter Control System 158
7-8 Relative Dimensions of the Long-Cone Cyclone 160
9-1 Arsenic Acid Production Process 208
9-2 Production Schematic for DSMA, MSMA,
and Cacodylic Acid 210
xx
-------
1.0 INTRODUCTION AND SUMMARY
1.1 PURPOSE AND SCOPE
On June 5, 1980 EPA listed inorganic arsenic^ as a hazardous air
pollutant under Section 112 of the Clean Air Act. This action was based on
the EPA Administrator's findings that there is a high probability that
exposure to inorganic arsenic causes cancer in humans, and that there is
significant public exposure to inorganic arsenic which is emitted into the
air by stationary sources. After listing inorganic arsenic under
Section 112, EPA conducted a "Phase I" screening study to gather data on
suspected source categories of inorganic arsenic. This Phase I screening
study proved to be inconclusive for all the source categories involved,
therefore, EPA initiated this second study to finalize the data gathering
for arsenic source prioritization.
The purpose of this study is to examine seven stationary source
categories of inorganic arsenic emissions and provide information on each to
assist EPA in determining priorities for detailed studies for possible
regulatory development. The seven source categories are as follows:
- primary copper smelters (exclusive of ASARCO-Tacoma),
- primary lead smelters,
- primary zinc smelters,
- secondary lead smelters,
- glass manufacturing,
- cotton gins (that handle arsenic acid-desiccated cotton), and
- arsenic chemical plants.
These source categories were examined to determine the magnitude of
arsenic emissions corresponding to the current or existing leve^ of control,
the magnitude of arsenic emissions resulting from compliance with all
applicable regulations, and the magnitude of arsenic emissions resulting
from the use of estimated best control (EEC) for arsenic. In the report,
the term baseline is used for all source categories to denote the lesser of
arsenic emissions that are currently occurring or the arsenic emissions that
would be expected if the source was complying with all applicable regula-
tions. These regulations include State implementation plans (SIPs),
-------
Occupational Safety and Health Administration (OSHA") standards, and New
Source Performance Standards (NSPS). A potential EEC for each source
category was selected based on whether the emitted arsenic was primarily in
a particulate or gaseous form, or in a combination of both forms, and
whether the technology had been demonstrated as an effective method of
arsenic control within the subject or a similar source category. Once an
EBC option was selected, the objective was to determine the incremental
arsenic emissions reduction achievable over baseline by implementing EBC.
The resultant costs of the EBC increment were also determined.
The arsenic emissions and control data in the report were obtained from
published technical papers and reports, EPA data files on arsenic, and from
the subject companies within each source category. No source testing was
performed for this study. In instances where no arsenic emissions data were
available, estimates were prepared using the assumptions and calculations
given in Appendix A of this report. All data on the cost of arsenic control
equipment were derived from published cost curves in economic and technical
reports. All cost bases and assumptions used in this report are detailed in
Appendix B.
The source Chapters 3-9 present arsenic emissions estimates, arsenic
control device technical feasibilities, capital and annualized control cost
data, and resulting impacts of additional arsenic control for each source
category. The regulatory frameworks that influence the control of arsenic
emissions from each of the source categories are discussed in Chapter 2.
Recommendations on the ease of possible standards development are made based
on the technical and economic feasibility of requiring additional arsenic
control and the status of other regulatory efforts affecting arsenic
control.
1.2 SUMMARY OF RESULTS
Table 1-1 presents a summary of information for each arsenic emissions
source category. The data presented in Table 1-1 are specifically derived
and/or explained in each of the appropriate source category chapters or
their corresponding appendices.
-------
r.
— u
Zl
--. u i -
=6
i
1
i
00
LU
l— t
C
CD
LU
O
LU
O
C£.
O
(/}
O
LU
o:
LU
01
H-
0£
O
LL.
2
C
}w.
* 1^
SS ° °£ 2" S
£
irt w -o
« O- X 5
X "O |C flT TJ S"
8 — ft < 8.
« v ^ tn v x ••
i -1 2 j ifl
*= 5 -J - * «
< ^ -g J W 3
• < - W fc -V> < 1^ -w
« O (AX4CW2 C £ ffl
S
» "" * 0
8
si :: si
e — c a E -
u J "5 **'*' 5 ^ "
a, • (2 ^ ^ — 4J 1U *"
D W U UU — M ^ — h« ft* C
U1""*^^ 2*1 3"«ce *& § "<* e
££ J!J Jls^ &n ;
£ , , i £ , , , , &
-------
I
t-H
u-
fti
C
U
A
w
ft)
JJ
c.
5
o
t
en
u
2
E
14
B
f-i
41
|
*o
E
O
U
4)
O
u
B
•c
C
a
ti
w
kl
O
u
kl
4V
U
0
o
41
41
tl
tr.
B
J>
II
v=
ki
O
1*4
e
J
a
41
a
c
*
|
a
a
_o
a
a
"S
t)
u
4V
AC
3
hv
O
E
•5
^.
o
1-
kl
c.
t)
bl
tl
11
JE
U
&
«
_e
ti*"
41
L>
1C
E
U
U
41
C
0
4V
•5
SI
E
"a
u
4J
01
41
ti
ti
*-4
CO
41
3
CO
U
41
vv
•O
tl
0
E
£
B
1C
U
£
41
|
B.
Z
w>
a
a
ti
Cl
0
kl
0,
41
f.
C
u
e
41
V
14*
O
tl
c
f
E
O
e
e
•a
V
"a
PHI
3
'o
v<
Cl
e
B
O
4J
Cl
3
•O
a
u
to"
41
k>
tr.
e smeltel
•o
K
0
E
„
O
W
fj
i
•^
41
S
O
a
vt
~B
ec
K
E
O
CO
B.
O
a
Cl
a
u
3
eo
1
c
w
^
X
v=
Uv
O
tl
C)
E
ce
B
"e
_o
u
kl
a
u
*j
I4«
O
41
D
e
o
41
|
4>
k.
6
CC
4V
E
a
_
1
u
E
N
k>
K
E
s-
|
IM
a
E
C
.H
U
•J
41
tl
eo
u*
ji
ig process emissions only.
o
E
Q
Cl
£
ovenent 1
ki
B.
E
C,
•a
u
tc
3
E
•
le
S
R
4i
k.
C.
01
at
^
41
U
0
01
S
O
m
C-
M
•O
CC
tl
•13
E
3
•O
tl
kl
3
cr
41
b.
41
1»
10
rause of the uncertainty in what types of contri
cy
Ji
m
O
a
4)
>
ec
a
k*
tc
41
ki
CC
ki
O
•D
01
6
Cl
115
U
.a
"0
j»
C
u
u
CC
0
z
o
ki
c
in place, therefore, the Improvement would be ;
0}
J=
o
Oj
k.
tl
41
•0
10
a
»
u
CC
a
u
u
>
^
3
Ift
a
01
u
o
kl
pv
c
u.
LO
c-
5
CC
tl
kl
X
X
0
•u
E
C
kl
X
61
Z
o
CO
eo
E
ce
c.
c
u
ki
cc
o
E •
— 'c
•v o
tl -*
C ki
u
C9 3
•a
kl C
C k.
u- O.
a -c
T* CC
tl
kl •"*
cr.
C k.
U X
o ec
i1
• o
O ON
r» S
if* •
«V ON
tl
.C ki
eg
• B.
X
B •—
O CC
C
ki ce
C k.
C o
U iw
B —i
tl
k. O
k. 0
£ o
tl 0
ki r-
ce
kl U
2.8.
n
• 41
C "-
C E
tl C
Ci kl
tn 3
n o
5 B.
X
E
CO
tl
JO
•fl-
ee
0
c
s
o
"e
v*
tl
E
C
E
eo
S
tl
a
o
te
o
a
m
o
B
0
"S
CD
en
41
a
a
c
.c
s
to
0.
kl
a.
a
E
o
41
ki
tc
a
tr
c
—
tc
B
CO
**
O.
Q
e
1
>,
A
a
u
o
E
O
*4
4V
U
3
•0
41
^i
Ik,
O
E
U
tl
a.
o
u
i
s
r 8, Section 8.2.
due to OSIIA regulation for arsenic, see Chaptel
01
contrc
—
O
X
•o
to
u
kl
a
V
k>
a
e
o
tD
a
tl
41 ki
£ 01
4V -0
E
ffi C
ce i-
4J Z
E 4V
c c
a.
u "
01 O
C*. 4V
n 01
§'
•o
ft) OP*
U M
a «
HI IT
W O
C >
u c
c-
o
c
c
o
I
41
£1
I
i
CO
•D
E
a
a
rc
eo
41
j=
X
^
u
to
kl
B.
O
u
a.
c
E
CC
n
ti
u
c
B
ki
3
«M
kl
g.
B
«H
•o
£
••-I
u
B
^M
Uv
01
kl
X
«^
B
O
to
4V
B
ce
~4
B.
i
in
iei' nollulantH. J lmpro»««K»i. • ~
etniS5lonsba$ellne
t upon the definition of secondary lead smelter;
r thnn for smelters.
k> *•! •"
i c.f
•O -0 3
E C
CO X
^v 41
X J9 ki
« ec C
rsenic 01
considei
refiners
a r.
ti S
u- — O
C k. ki
o *•
-1 >
o a
k. a B
4V 4V O
E Ex
o R a
u ^ u
fi* ~*
i «- 5
4V 0
B k. •
0 tl HC
r c
-o i ^
ti a
V) 4V
,., k. 0
C E
C C.
CC. Oi 0
1«~ K T5
T k.
-------
1.3 REFERENCES
1. U.S. Environmental Protection Agency. National Emission Standards for
Hazardous Air Pollutants; Addition of Inorganic Arsenic to List of
Hazardous Air Pollutants. 45 FR 37886. Washington, D. C. Office of
the Federal Register. June 5, 1980.
2. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for Emission Standards and Engineering Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency.) EPA Contract No. 68-02-3173. May 1982.
-------
2.0 ARSENIC PROPERTIES AND CURRENT REGULATIONS AFFECTING EMISSIONS
This chapter is divided into two sections which discuss the physical
and chemical properties of arsenic pollutants and their relationship to air
pollution, and the various regulatory authorities which have a role in
reducing arsenic air pollution.
2.1 PROPERTIES OF INORGANIC ARSENIC POLLUTANTS
This section briefly discusses the physical and chemical properties of
inorganic arsenic pollutants and the effects these properties have on air
pollution emissions. The major thrust of the section will be toward the
properties of arsenic trioxide since this is the most common inorganic
arsenic compound.
Arsenic is the 47th most frequently found element in the Earth's crust.
Its total occurrence in the crust is estimated at about 0.00055 percent.
Arsenic occurs in nature primarily as a sulfide or bound to sulfur, and is
often found in association with iron. The most common form of
naturally-occurring arsenic, arsenopyrite, (FeAsS) exemplifies this. Metal
ore deposits of lead, zinc, gold, silver, copper, and tin are also natural
sources of arsenic.
Elemental arsenic is a nonmetal or metalloid that is classified in
Group 5a of the periodic table which also includes phosphorus, nitrogen,
antimony, and bismuth. The most common oxidation states of arsenic are -3,
+3, and +5, with +3 being the most prevalent in inorganic arsenic compounds.
Other important properties of arsenic are listed below:
Properties of Arsenic
Atomic number 33
Atomic weight 77.9216
Melting point
at 1 atmos., sublimes at 613°C
at 28 atmos., melts at 817°C
Density at 20°C 5.72 g/cm
Latent heat of fusion 88.5 cal/g
Latent heat of sublimation 102 cal/g
Specific heat at 20°C 0.082 cal/(g)(°C)
Lattice constants at 26°C a = 2.760A
b - 10.548A
Hardness (Mohs scale) 3.5
-------
The most important commercial arsenic compound is arsenic trioxide
which is also known as arsenous oxide, "white arsenic" and arsenic. Arsenic
trioxide is used as the raw material to manufacture the majority of all
other inorganic and organic arsenic compounds of commercial importance. A
detailed list of arsenic compounds and specific uses is given in Chapter 9,
Table 9-1.
Arsenic trioxide occurs as a white solid in several crystallagraphic
modifications. The most common form is the octahedral (or cubic)
modification known as arsenolite. Arsenic trioxide is appreciably volatile
at temperatures above 100°C (212°F) which is important from an air pollution
standpoint.. Table 2-1 presents some vapor pressure data for arsenic
trioxide. '
2.1.1 Arsenic and Air Pollution
Of the seven source categories under study, five are referred to as
"hot" sources and two as "cold" sources. The source categories
characterized as hot emission sources involve high temperature processes in
which some portion of the arsenic component of the materials being processed
is volatilized. A large portion of the volatilized arsenic is eventually
emitted in the vapor phase if the process exhaust gas temperatures are not
reduced by using cooling devices. The primary copper, lead, and zinc
smelters, the secondary lead smelters, an'd the glass manufacturers
constitute the hot sources. The cold emission sources, which only emit
arsenic as a particulate, consist of the arsenic chemical plants and the
cotton gins. The cold sources involve ambient or relatively low temperature
[less than 38°C (100°F)] processes that do not volatilize any arsenic
components.
As stated before arsenic and its compounds are ubiquitous and are found
in close association with many nonferrous metal ores. Table 2-2 lists some
nonferrous metal ores and their observed range of ore concentrate arsenic
content. When these ores are processed arsenic compounds are released. The
predominant source of arsenic and arsenic trioxide in the U. S. is as a
byproduct' of copper smelting. The arsenic occurs predominantly as the
sulfide which in most pyrometallurgical processes, oxidizes to the trioxide.
Both the sulfide and oxide are extremely volatile at the temperatures
encountered in nonferrous metal production. The arsenic trioxide or sulfiae
enters the gas stream, and as the gas stream cools down, some condenses and
is collected in the flue dusts. However, because of the high vapor pressure
of the trioxide, appreciable quantities of the oxide may remain in the gas
stream unless it is cooled to below 100°C (212°F).
Not all of the arsenic in the feed materials volatilizes. Some remains
in the solid phase and is carried through to the product metal, some is
released to the atmosphere in subsequent operations, and some ends up
complexed in the slag.
-------
TABLE 2-1. ARSENIC TRIOXIDE (ARSENOLITE) VAPOR PRESSURE DATA
3,4
Vapor
Pressure
(mm of Hg)
2.4 x 10~c
_K
2.5 x 10 I
4.6 x 10 ~^
_x
1.9 x 10 ^
2.2 x 10';:
_ 9
2.6 x 10 ^
— 9
1.0 x 10 ^
_2
2.7 x 10
1.0
5.0
1.0 x 107
2.0 x 10:
4.0 x 10:
6.0 x 10^
1.0 x 10^
2.0 x 10,
4.0 x 10,
7
7.6 x lO''
Equilibrium
Vapor Phase
Concentration Temperature of Arsenic Trioxide
(mg As/ni ) (°C)
3.9 x 10"?
_ i
4.0 x 10
7.4 .
1
3.1 x 10,
|
3.5 x 10,
9
4.2 x 10,
7
1.6 x 10,
4.3 x 107
1.6 x 10,
/\
8.1 x 10^
1.6 x 10^
3.2 x 10^
6.4 x 10^
9.7 x 10^
1.6 x 10,
3.2 x 10^
6.4 x 10^
1.2 x 10
60-61
81-86
101-105
117-124
119-124
149-152
153.5
165
212.5
242.6
259.7
279.2
299.2
310.3
332.5
370.0
412.2
457.2
-------
TABLE 2-2. ARSENIC CONTENT OF ORE CONCENTRATES SMELTED IN THE
DOMESTIC PRIMARY COPPER, LEAD, AND ZINC INDUSTRIES
ORE CONCENTRATE
TYPE
Copper
Lead
Zinc
RANGE OF
ARSENIC CONTENT
(wt %)
0.004 - 3.8
0.2 - 0.4
0.002 - 0.1
APPROXIMATE PRIMARY
METAL CONTENT
(wt %)
25
60
60
ARSENIC /METAL
RATIO
0.0002 - 0.15
0.0003 - 0.007
0.00003 - 0.002
-------
Inorganic arsenic compounds can also be volatilized and released to the
atmosphere any time an arsenic containing material is burned (example, coal
combustion) or subject to high temperatures. Many nonferrous metals are
recovered as scrap and reprocessed by smelting and refining. When the scrap
material is smelted some of the arsenic present as an alloying constituent
can be oxidized and volatilized and released into the environment. In glass
manufacture inorganic arsenic compounds are added to the batch materials and
then subjected to high temperatures. Some arsenic trioxide vapor is
released to the environment during these operations. Emissions of inorganic
arsenic can also occur when these compounds are being handled and used. For
example, during the manufacture of many arsenic chemicals dry arsenic
trioxide is the starting point. Some fugitive emissions of inorganic
arsenic particulate may occur during the handling and dumping of this raw
material.
2.1.2 Air Pollution Control
The type of air pollution control most effective on a given source of
arsenic will be dependent upon several of the source characteristics, the
two most important being exhaust gas arsenic concentration and temperature.
Sources operating at temperatures above 100°C (212°F) and handling materials
containing arsenic have the potential to emit appreciable quantities of
arsenic in the vapor phase. In order to control these emissions, the
exhaust gases must be cooled down to allow the arsenic species vapor to
condense, and then must be cleaned with a particulate removal device. For
example, the mechanism of cooling down vaporized arsenic trioxide to recover
trioxide dust is readily demonstrated by analogy to the "kitchens" or brick
cooling chambers used by the smelting industry to produce commercial grade
arsenic trioxide. As the gas stream exits the last kitchen its temperature
is less than 100°C (212°F) and the condensed particulate is 90 to 95 percent
pure. A more detailed discussion of the arsenic trioxide kitchens can be
found in reference 2.
The efficiency of the cooling/collection device system is a function of
the inlet arsenic concentration and the degree of cooling achieved, and is
bounded by the equilibrium vapor pressure of the arsenic species present.
Figure 2-1 presents a plot of control device inlet and outlet arsenic
concentrations vs. temperature for various hot arsenic sources. The data
shown were obtained from tests conducted on copper smelter process offgas
streams. The figure shows that the lowest arsenic outlet concentrations
occurred when the outlet gas temperature was below 100°C (212°F). The curve
also shows that the outlet arsenic concentration tracks that predicted from
the theoretical vapor pressure curve. There are of course practical limits
to the amount of cooling that can be achieved in a given situation. For
example, one must avoid cooling sulfur containing stack gases to below the
acid dew point or severe corrosion problems can result. Alternatively,
these streams with high dew points can be controlled with wet scrubbers
where the acid can be easily neutralized.
10
-------
• 3
£
s
Q
0)
O
O
u
01
M
A5
• 7
10
10
10
10
-1
10
-2
Legend
Symbol Device
A Electrostatic Precipitator
Q Fabric Filter
O Venturi Scrubber
O Acid Plant
• Blackened symbols represent inlet
conditions to the control devices.
Open symbols represent outlet conditions
from the control devices. E
• Superscript numbers designate the
inlet/outlet pairs.
Solid line corresponds to saturated As,0
vapor.
• 8
• Asterisk (*) indicates that a spray chamber unit
cooled the inlet gas prior to treatment by the
control device.
El 9
1.65 1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
10 X Reciprocal of Offgas Temperature, °K
300 250 200 150 100 80
Off gas Temperature, °C
60
40
20
Figure 2-1. Effect of temperature on the performance
of arsenic emissions control devices.
11
-------
Cold sources of inorganic arsenic emissions such as manufacturing of
arsenic chemicals from dry arsenic trioxide can be controlled by any
particulate control device. In these cases the ultimate performance is not
dictated by exhaust gas temperatures.
2.2 REGULATIONS AFFECTING THE EMISSION OF INORGANIC ARSENIC
Arsenic emissions from industrial sources are affected-by arsenic-
specific regulations, particulate and SO. regulations, workplace safety and
health regulations for inorganic arsenic and lead, and national ambient air
pollutant concentration limits. The specific influence and application of
these regulatory measures on the seven source categories of arsenic
emissions are discussed in the following sections.
2.2.1 Arsenic-Specific Regulations
No regulations could be identified on the Federal, State, or local
level that specifically limit the atmospheric emission of inorganic arsenic
on a mass emission or percent reduction basis. Some States and localities
do, however, have regulations and guidelines that affect arsenic emission
sources and do aid in reducing public exposure to inorganic arsenic. The
city of Philadelphia has regulations which identify arsenic as a toxic air
pollutant and require all emitters of the chemical to register with the
Philadelphia Air Control Board. The level of arsenic emissions from a
particular facility is determined and an ambient fenceline concentration
from the facility calculated through dispersion modelling. The determined
fenceline concentration is compared to an established acceptable ambient
concentration for that toxic pollutant to decide whether further controls
are required. If the fenceline concentration is greater than the guideline,
the city can require best available control technology to be installed. '
New York State has a similar system for toxic air pollutants which also
covers arsenic, however, the New York system is only a policy guide for the
environmental control department and is not part of the State's regulatory
framework. Threshold Limit Values are used to establish acceptable ambient
limits for a toxic air pollutant such as arsenic. Dispersion modelling is
used to determine the ambient concentration attributable to the source. The
determined concentration is compared to the established acceptable limit and
a decision is made by the environmental control agency as to whether further
controls are required.
In New Hampshire the general particulate matter process weight
regulation is used to control sources where arsenic is emitted as a
particulate. The general equation is modified for toxic pollutants such as
arsenicRby applying a stringency factor. The stringency factor for arsenic
is 0.9. Therefore, if a plant were allowed to emit for example, 4.5 kg
(10 lb)/hr under the general particulate regulation, they could only emit
4.1 kg (9 lb)/hr if the emissions contained arsenic.
12
-------
2.2.2 Particulate Matter and Sulfur Dioxide Regulations
State particulate matter regulations in the form of State
implementation plans (SIP's), and Federal new source performance standards
(NSPS) directly affect to varying degrees the extent to which arsenic
emissions are reduced from the seven source categories in this study. All
seven source categories are affected by SIP regulations, while only primary
copper, lead, and zinc smelting, secondary lead smelting, and glass
manufacturing are affected by NSPS. The glass manufacturing NSPS, however,
is currently under reconsideration by EPA and may be dropped. In the
process of controlling particulate matter by add-on control devices (fabric
filters, scrubbers, etc.) and process modifications, some level of arsenic
control is also achieved. Cold sources such as the cotton gins and arsenic
chemical plants emit arsenic as a particulate that can be effectively
controlled by the add-on devices used in these industries to meet SIP
requirements.
The hot sources, including glass manufacturing and all four smelting
industries, pose a somewhat different arsenic control situation because
arsenic can be emitted as a vapor in these instances. Particulate controls
such as fabric filters or electrostatic precipitators installed to meet SIP
or NSPS emission levels are not effective on high temperature gas streams
containing large amounts of vaporized arsenic. Therefore, even though a
glass or smelting facility may be doing an efficient job of controlling
gross particulate matter, it may be doing a poor job of reducing arsenic
emissions. However, facilities that cool down the arsenic-containing gas
stream prior to entering the control device have demonstrated efficient
arsenic removal levels. In these instances particulate matter regulations
would aid in helping to reduce atmospheric arsenic emissions.
Sulfur dioxide (S0~) regulations on the State and Federal levels affect
arsenic emissions from the primary copper, lead, and zinc smelting and
secondary lead smelting source categories. Arsenic -missions are directly
and effectively controlled through the techniques used by these four
smelting industries to reduce atmospheric SCL emissions. The other source
categories in this study are not regulated for SO emissions because they
either do not emit S0_ or emissions of SCL are negligible in comparison to
other pollutants such as particulates. The smelting source categories
control SO by the use of wet scrubbers and contact sulfuric acid plants.
These types of control systems are effective on the hot exhaust gases of the
smelters because they cool down the exhaust gas temperature such that
vaporized arsenic losses are minimized. The arsenic components are
condensed and are efficiently controlled by the devices' particulate removal
mechanisms.
The arsenic emissions from a smelter process offgas stream are often
controlled if the source is complying with its applicable S0? and
iculate matter regulations. However, efficient arsenic removal throu
and particulate matter controls may not be implemented at all Arizona
13
-------
primary copper smelters if the currently proposed Arizona SIP for S0? is
approved by EPA. The proposed Arizona SIP is not as stringent for S07
emissions as the EPA-designated, smelter-specific SCL emission limits that
these smelters will otherwise have to comply with. If the proposed Arizona
SIP is approved, some smelters may be able to comply with their SO.
standards without applying controls equivalent to the estimated best control
(EEC) technology for arsenic emissions.
The EPA NSPS limits for SCL and particulate matter from primary copper,
lead, and zinc smelting and secondary lead smelting are summarized in
Table 2-3.
2.2.3 National Ambient Air Quality Standards
The requirement for the arsenic source categories to attain the
national ambient air quality standards (NAAQS) for SCL, particulate matter,
and lead have or will have an impact on the level of arsenic emissions from
the subject plants. As plants are required through SIP's to reduce SO ,
particulate, and lead emissions to achieve the NAAQS, some level of arsenic
control will be achieved in the process. The direct effect of the NAAQS on
overall arsenic emissions reduction in the seven source categories is
difficult to quantify, particularly because some sources such as cotton
gins, primary lead smelters, and secondary lead smelters have not yet come
into full compliance with the applicable particulate (for cotton gins) and
lead NAAQS's (for primary and secondary lead smelters).
The NAAQS for SO., particulate matter, and lead are summarized in
Table 2-4.
2.2.4 Worker Exposure Standards
The Occupational Safety and Health Administration (OSHA) has published
concentration standards for workplace exposure to inorganic arsenic and lead
that play significant roles in reducing atmospheric arsenic emissions. The
OSHA standard for airborne inorganic arsenic., limits in-plant worker exposure
to 10 ug/m averaged over any 8-hour period. For lead, the exposure
concentration limit for workers is 50 yg/m averaged over an 8-hour
period. The lead standard helps reduce arsenic because the arsenic is
emitted with the lead and is controlled concurrently as lead is controlled.
The primary and secondary lead smelting source categories are affected by
both standards. The OSHA arsenic standard applies to all other source
categories except cotton ginning. Cotton ginning is excluded because it is
classified as an agricultural source and agricultural sources are exempt
under the regulation. '
In efforts to comply with the OSHA lead and arsenic standards,
facilities have had to install control measures on both process and fugitive
emission points of arsenic and lead. For the arsenic chemical manufacturing
source the OSHA arsenic standard is the prime regulatory force behind the
14
-------
prf
o
CO
W
2 O
0 M
M X
CO O
CO I-H
M Q
B cfi
g
CO PK
P-4 | 1
CO E>
2 co
w 5
H <
Is
O P
Prf C_J
P-l M
1 H
< Pi
W P-i
•
CO
1
CM
W
eg
H
Q
w
o s~*
iJ CJ
•< CO
h-5 "^.
W l-i
> 00
w ^— '
sTg
o -^
M 00
co s
CO
M
§
H
2
H
i-3
O
P-i
W
CJ
prf
ft
o
CO
o
M
CO
CO
M
W
S
O
o
H
CJ
W
CJ
P
O
CO
'—»
CM
O
•
o
o
l/^
CO
4-1
CO
rH
3
O
•H
>_,
Cfl
P-i
V^
CO
£r
Q
t-i
CO
cx
o
u
t-l
cfl
B
•H
t-l
P-i
8-3
in
vO
o
0
CM
O
CO
00 01
C 4J
•H p
4J ^J
•H >
0) C
B 0
co u
. «
v-> co
co cj
4J CO
co C
O 3
prf <4-l
00
c
'w
!_^
co
B
CO
o
*
o
m
o
•
o
CM
O
o
*~s
O
m
o
»
o
01
jj
CO
rH
O
CO
P-i
O
CO
CO
4-1
cfl
i-H
y
o
O
CO
,0
t-l
CO
^
co
^i
S-i
o
4J
en
co
w
^
CO
4J
C
•H
CO
«
co
o
CO
c
t-J
3
CO
00
u
cfl
o
Cfl
•H
Tj
co
c
•H
(-1
O
cfl
B
*
co
C
T-t
r"!
a
cO
B
^4
01
4-1
C
•H
C/3
00
f3
•H
4J
rH
CO
B
CO
a
•H
t^
4-1
U
CO
1— 1
CO
t-l
CO
4-1
t-l
CO
^
c
o
a
M
CO
o
CO
c
}_l
3
14-4
CO
c
•H
.C
a
co
B
t-i
O)
co
CO
CO
o
Pi
•a
co
CD
00
c
o
c
•H
CO
e
B
CO
00
)-J 4-1
CO i-l
e j
•H B
l-l CO
P-I
CM
CN1
O
O
m
co
4-1
CO
rH
3
a
•H
cfl
P-I
t-i
a)
C co
cfl CO
u
4-» CO
Cfl P
co t-
I—I 5
pa MH
•o
CO
0)
>h 00
>J C
cd >H
T3 4J
(= rH
o co
o B
a) co
co
15
-------
Pi
o
fe
W
q o
33
g^j
^ f— 3
H Z
H td
M H
rJ o
gg
cy u
M
pi H
H Pi
P*
H
Z »
w w
M O
a
z --
O 60
CJ 3-
Pi
W
H
2
<
1
Q
Z
H
pan
o
w
Pw
H
H
Z
<
H
P
h-3
o
p-
•^3" ^O
O -*
O O
O O
0 0
o o
^^ 'w'
o m
OO vO
CO
0 cfl
co a
O) 3
a a
iH
cj X
•H cfl
jj a
£ rH
x; co
4-1 3
•H C
r4 C
• •
M M
>> ^s
^ ^3-
CNl
r^
J^
cfl
a
•H
M
PM
co
01
*o
, J ^^^
X CM
O O
C/5
^4
3 co
<4H ca
rH *~*
^
C/)
1 —
m
o
o
o
o
o
o
cfl
a
3
•H
cfl
a
rH
cO
3
a
e
•
r4
,fl
cO
!>*,
H
CO
fl
O
CJ
o>
C/l
CO
o
o
o
o
o
in
r^»
C
CO
0
O
•H
^4
4-1
0)
a
O
Ol
60
rH
cfl
3
C
C
•
Jj
^c
^J.
cs
^1
^1
cfl
a
•H
}^
p^
Ol
4J
CO
rH
3
o
•H
4-1
}_l
CO
P-4
i-H
i— i
0
0
o
o
o
vO
01
CO
a
3
a
•H
X
CO
a
rH
CO
3
C
C
•
^
-G
vj.
CM
CO
O
o
o
o
*
o
o
C
cfl
01
e
a
•H
l-l
4-1
Ol
a
o
01
60
r_|
cfl
3
C
C
<:
•
M
.£
^.
CM
^
}^
CO
•a
c
o
o
Ol
CO
i — *
o
o
o
o
o
o
m
"
cO
a
£3
•iH
X
co
a
rH
CO
3
C
C
•
^4
JS
-------
controls applied during arsenic trioxide handling and dumping. In the glass
source category the OSHA arsenic standard was the reason that a major
manufacturer switched from using a dry arsenic raw material, with a large
opportunity for worker exposure, to a liquid arsenic raw material with no
threats to the workers from arsenic exposure. In the smelting source
categories much of the problem in not being able to achieve workplace
concentration limits comes from fugitive emissions. Because of the lead and
arsenic standards, improvements have been made in smelter fugitive emissions
control that have direct benefits towards reducing atmospheric arsenic
emissions. OSHA is currently in the process of working individually with
the smelting sources to control fugitive arsenic emissions that cause the
10 yg/m standard.to be violated. A similar program for lead is planned in
the near future.
17
-------
2.3 REFERENCES
1. Peters, A. The Use of Arsenic with Particular Reference to the Glass
Industry. Glastechn. Ber., 50:12, 1977. pp. 328-335.
2. Kirk-Othmer. Encyclopedia of Chemical Technology. Volume 3. Third
edition. USA. John Wiley & Sons, 1977. pp. 243-256.
3. Arsenic Emissions from Primary Copper Smelters - Background Information
for Proposed Standards - Draft Report. U. S. Environmental Protection
Agency. Research Triangle Park, North Carolina. February 1981.
p. 4-2.
4. Reimers, J. W. & Associates, Ltd. Study of the Emission Control
Technology for Arsenic in the Nonferrous Metallurgical Industry.
Performed for the Air Pollution Control Directorate - Environment
Canada. January 1977. p. 24.
5. Philadelphia Department of Public Health - Air Pollution Control Board,
Air Management Regulation VI, Control of Emissions of Toxic Air
Contaminants. July 1, 1981.
6. Memo from Schell, R. M., U. S. EPA to Patrick, D. R., U. S. EPA.
November 16, 1981. Regional Office and State Agency Visits.
7. Letter and attachments from Cashman, T. J., New York State Department
of Environmental Conservation to Blanchard, K., U. S. EPA. April 9,
1982. 3 p. Air guide policies.
8. State of New Hampshire. New Hampshire Code of Administrative Rules,
Chapters 100, 200, and 400-1200. Part 1203.02(b). February 19, 1981.
9. Air Pollution Control Association. 1981-1982 APCA Directory and
Resource Book. 1981. p. 153.
10. Reference 9, p. 151.
11. Reference 9, p. 143.
12. Occupational Safety and Health Administration. Code of Federal
Regulations. Title 29 - Labor. Chapter 17, Section 1018,
subsections a-c, Part 1910. Washington, D. C. Office of the Federal
Register. July 1, 1980.
13. Occupational Safety and Health Administration. Code of Federal
Regulations. Title 29 - Labor. Chapter 17, Section 1025,
subsections a-c, Part 1910. Washington, D. C. Office of the Federal
Register. July 1, 1980.
18
-------
14. Telecon. Brooks, G. W., Radian Corporation, with Gordon, C., OSHA.
April 12, 1982. Conversation concerning the OSHA arsenic standard.
15. Telecon. Keller, L. W., Radian Corporation, with Cassidy, M., OSHA.
March 18, 1982. Conversation concerning OSHA study of copper smelters.
19
-------
3.0 PRIMARY COPPER SMELTING
This chapter discusses the primary copper smelting industry and its
associated arsenic emissions. Section 3.1 presents a brief source category
description. Section 3.2 discusses the regulations currently covering the
industry and the resulting level of control and associated arsenic
emissions. Section 3.3 presents a discussion of the EEC options and
Section 3.4 discusses the impacts of applying these options. Section 3.5
presents the input data to be used in assessing the population exposure to
arsenic emissions from primary copper smelting. The ease of standards
development is discussed in Section 3.6.
3.1 COPPER SMELTING DESCRIPTION
3.1.1 General Information
Metallic copper is produced from copper ore concentrates at primary
copper smelters. The United States' primary copper smelting capacity
totalled 1.56 million Mg (1.67 million tons) in 1979, which accounted for
about 70 percent of the domestic copper consumption. There are currently
seven companies and 15 individual plants that produce primary copper in the
U.S. Table 3-1 lists the names, locations, and production capacities of the
plants. The 14 primary copper smelters exclusive of the ASARCO plant in
Tacoma, Washington will be considered in this report. A separate report
concerning arsenic emissions and regulatory alternatives for ASARCO-Tacoma
has been published by the EPA.
3.1.2 Process Description
The materials processed at the U.S. primary copper smelters are
predominantly sulfide ore concentrates. The most common copper-bearing
constituent of these ores is chalcopyrite (CuFeS_). The copper content of
the raw ore as mined is typically less than 1 weight percent, and flotation
methods are used to produce a copper concentrate that serves as the feed to
the primary smelting plant. A typical copper ore concentrate certains 15 to
30 weight percent copper and large amounts of iron, sulfur, and silicon.
Other important constituents include lead, antimony, zinc, cadmium, bismuth,
selenium and arsenic. Arsenic generally appears as the mineral arsenopyrite
(FeAsS) in copper-bearing ores and is concentrated by the flotation process.
The arsenic contents of the ore concentrates smelted at the 15 domestic
copper smelters are listed in Table 3-2. The amount of arsenic in the
concentrates ranges from 0.004 to 3.8 weight percent.
The copper-bearing minerals in the ore concentrate are separated from
the other materials and converted to 99 percent pure "blister copper" by the
20
-------
TABLE 3-1. THE DOMESTIC PRIMARY COPPER SMELTERS
PLANT
LOCATION
COPPER PRODUCTION
CAPACITY3
Mg/yr (tons/yr)
ASARCO/E1 Paso
ASARCO/Hayden
ASARCO/Tacoma
Kennecott/Hayden
Kennecott/Hurley
Kennecott/McGill
Kennecott/Garfield
Phelps Dodge/Morenci
Phelps Dodge/Douglas
Phelps Dodge/Ajo
Phelps Dodgs/Hidalgo
Copper Range/White Pine
Magma/San Manuel
Inspiration/Miami
Cities Services/
Copperhill
El Paso, Texas
Hayden, Arizona
Tacoma, Washington
Hayden, Arizona
Hurley, New Mexico
McGill, Nevada
Garfield, Utah
Morenci, Arizona
Douglas, Arizona
Ajo, Arizona
Hidalgo, New Mexico
White Pine, Michigan
San Manuel, Arizona
Miami, Arizona
Copperhill, Tennessee
104,000
163,000
91,000
73,000
73,000
45,000
254,000
161,000
115,000
64,000
127,000
82,000
181,000
136,000
20,000
(115,000)
(180,000)
(100,000)
(80,000)
(80,000)
(50,000)
(280,000)
(177,000)
(127,000)
(70,000)
(140,000)
(90,000)
(200,000)
(150,000)
(22,000)
1979 Production capacity of "blister" copper (99 percent pure copper).
21
-------
TABLE 3-2. ARSENIC CONTENT OF THE FEED TO
DOMESTIC COPPER SMELTERS
Plant
ASARCo-Tacoma
Kennecott-Garfield
Phelps Dodge-Ajo
ASARCo-El Paso
ASARCo-Hayden
Phelps Dodge-Hidalgo
Phelps Dodge-Douglas
Phelps Dodge-Morenci
Kennecott-Hayden
Kennecott-McGill
In sp irat ion-Miami
Magma-San Manual
Copper Range-White Pine
Kennecott-Hurley
Cities Service-
Copperhill
%
3.8
0.14
0.3
0.22
0.2
0.026
0.03
0.02
0.015
0.03
0.022
0.005
0.0035
0.0005
0.0004
Arsenic content of
kg/hr
204.1
137
68
60C
50
20
11
9.7d
8.5
7.6
7.6
4.4
0.7
0.2
0.05
feeda'b
(Ib/hr)
(4500)
(301)
(151)
(133)
(110)
(45)
(24)
(21.3)
(18.8)
(16.7)
(16.7)
(9.8)
(1.5)
(0.4)
(0.1)
The feed is a mixture of concentrates, precipitates, lead smelter
by-products, and smelter reverts.
DDoes not include recycled flue dusts and other intermediates except
where noted.
"23 kg/hr (51 Ib/hr) of this amount is fed directly to the converters.
3.4 kg/hr (7.5 Ib/hr) of this amount is fed to the roaster and
remaining amount is fed directly to the reverberatory furnaces.
22
-------
operations at a primary copper smelter. As shown in Figure 3-1, there are
three major processes involved at most smelters: roasting, smelting, and
converting.
Copper ore roasting is carried out at six of the 14 smelters considered
in this study. Ore concentrates generally require roasting if they are low
in copper content relative to sulfur, iron, and other impurities. Roasting
consists of heating the ore concentrates to about 650°C (1200°F) in an
oxidizing atmosphere. The roast is carried out below the melting
temperature of the charge and has three primary functions: (i) elimination
of a portion of the sulfur as SCL, (ii) removal of volatile impurities such
as arsenic, antimony, bismuth, and cadmium, and (iii) conversion of a
portion of the iron sulfides present to iron oxides. The degree of sulfur
elimination accomplished in a copper roaster depends on the desired quality
of the charge to the smelting furnace. The final roasted product is known
as the calcine.
There are two types of roasters used in the domestic copper smelting
industry, multiple hearth roasters and fluid bed roasters. Of these two
types, fluid bed roasters represent a more recent design. For a given
degree of sulfur removal, they require a shorter concentrate residence time
and produce a higher strength S0« offgas stream than do multiple hearth
roasters. Three of the six smelters that roast ore concentrates use
multiple hearth roasters and three use fluid bed roasters.
The emission control technologies that can be applied to the two types
of roasters depend on the SO content of the offgas streams. The SO,.,
content of multiple hearth roaster offgas is too low for economical
treatment with a contact sulfuric acid plant unless the stream can be
blended with a more concentrated SO- stream. Two of the three plants
currently using multiple hearth roasters have no SO- control on the roaster
offgas. Electrostatic precipitators (ESP's) are enjoyed for particulate
control. The third plant using a multiple hearth roaster (ASARCO/E1 Paso)
employs offgas blending with a lead sinter machine offgas to produce a
sufficient concentration of SO., to treat with a contact sulfuric acid plant.
All three existing fluid bed roasters in operation currently use contact
sulfuric acid plants to control roaster SO- emissions.
The second step in the production of metallic copper from
copper-bearing ore is the smelting process. The feed to the smelting
furnace consists of either hot calcine from the roaster or unroasted ore
concentrates ("green charge") along with siliceous or limestone flux. These
materials are melted in the furnace at a temperature of about 1500°C
(2730°F). In the molten state, several complex metallurgical reactions
occur in which copper and higher iron oxides are chemically reduced, and
sulfur is liberated as SO-. The smelting furnace operations results in the
production of two distinct phases: a copper-rich phase known as the matte
that consists primarily of iron sulfide and copper sulfide, and a
copper-lean phase known as the slag that can be ideally represented by the
23
-------
ORE CONCENTRATES
FUEL
AIR
ROASTER
FLUX
AIR
FUEL
RECYCLED
CONVERTER
SLAG
FLUX
AIR
ROASTER OFFGAS
CALCINE
SMELTING FURNACE
MATTE
CONVERTER
SMELTING FURNACE
OFFGAS
SMELTING FURNACE
SLAG TO DUMP
CONVERTER OFFGAS
BLISTER COPPER
TO REFINERY
Figure 3-1. Schematic diagram of a primary copper smelter,
24
-------
formula 2FeO + SiCL. The slag and matte phases separate within the furnace
by virtue of their mutual insolubility, and the slag layer floats above the
matte. The two phases are tapped from the furnace separately. The furnace
slag is discarded due to its low copper content and the copper-bearing matte
is sent to the converter for further processing. The net result of the
smelting step is the separation of the copper-containing portion of the
furnace charge from the non-copper containing portion.
There are several types of smelting furnaces employed in the domestic
copper smelting industry: the conventional reverberatory furnace, the
electric furnaces, the flash furnace, and the continuous (Noranda) furnace.
The reverberatory furnace is unique among the list because it produces an
offgas stream that has a relatively low SCL content, typically 0.5 to
1.5 percent SCL. The other furnaces produce offgas streams whose SCL
content exceeds 3.5 percent SO-.
The existing control equipment employed on smelting furnace offgas
streams depends on the furnace type. The S09 content of reverberatory
furnace offgas is too low to treat the stream economically with a contact
sulfuric acid plant, and other forms of S0_ control such as flue gas
desulfurization are not practiced in the domestic primary copper smelting
industry. Electrostatic precipitators are used exclusively within the
industry to remove particulate matter from reverberatory furnace offgases.
Contact sulfuric acid plants are used to treat streams from electric, flash,
and continuous (Noranda) furnaces because of their relatively high S09
concentration.
The final step in the production of blister copper from copper ore
concentrate is the converter operation. Molten matte produced in the
smelting furnace is transferred to Fierce-Smith type converters in ladles
via overhead cranes. Siliceous flux materials are added to the charge in
order to provide a medium in which the iron contained in the matte can be
removed as a slag. There are two stages in the converter operations.
During the "slag blow", the iron sulfide in the matte is preferentially
oxidized to iron oxide. The iron oxide dissolves in the siliceous flux
phase and forms a slag layer that is removed from the converter and recycled
to the smelting furnace. When all of-the iron has been removed from the
converter charge in this manner, molten copper sulfide is converted to
metallic blister copper and sulfur dioxide during the "copper blow", which
is represented by the following reaction:
Cu2S + 02 -* 2Cu + S02
The unique feature of the converter operation is that it is a batch process
occurring in discrete modes. The characteristics of the converter offgas
vary during the different modes. The SO content of the offgas stream is
higher during the copper blow than during the slag blow.
25
-------
Offgases from the converter operation that are captured by the primary
converter hooding system are treated in contact sulfuric acid plants at
11 of the 14 copper smelters considered in this study. Of the remaining
three plants, two use hot electrostatic precipitator systems
(Kennecott/McGill and Phelps Dodge/Douglas) for particulate control and one
(Copper Range/White Pines) has no converter offgas pollution control
equipment.
The 99 percent pure blister copper produced by the converter is
subsequently refined to reduce the level of various impurities. Virtually
all copper produced by matte smelting is fire refined in rotary-type
refining furnaces or in small hearth furnaces. Copper refining consists of
a series of operations in which the blister copper is first oxidized to Cu?0
and then reduced back to metallic copper.
3.1.3 Arsenic Emissions from Primary Copper Smelters
Less than 5 percent of the arsenic input to a primary copper smelter
typically remains in the blister copper product. The two mechanisms of
arsenic removal from the copper-bearing material are volatilization and
slagging. Arsenic is volatilized primarily as arsenic trioxide (As^CL) in
the roaster, the smelting furnace, and the converter. Slagging of arsenic
in the form of metallic arsenates such as iron arsenate [Fe_ (AsO,)~] occurs
from the smelting furnace and the converter. The distribution of arsenic
among the offgas and slag streams varies considerably among smelters, but
some general conclusions can be drawn regarding the magnitude of arsenic
removal from each of the various process and fugitive sources as discussed
below.
3.1.3.1 Process Arsenic Emissions. The three major sources of process
arsenic emissions are the roaster, the smelting furnace, and the converter.
The magnitude of the arsenic emissions from each of these process emission
sources is considered below. A detailed discussion of this topic can be
found in a paper by Weisenberg, Bakshi, and Vervaert.
The fraction of arsenic in the ore concentrate that is volatilized in
the roaster may vary from as little as 4 percent to as much as 90 percent,
with 25 percent being a typical value. The fractional volatilization
depends on several variables including temperature, residence time, arsenic
content of the ore concentrate, and the oxidizing/reducing atmosphere within
the roaster. These operating variables are indirectly determined by the
desired grade (copper to sulfur ratio) of the matte that will ultimately be
sent to the converter. In general, volatilization of arsenic as arsenic
trioxide (As?0_) is enhanced by relatively low roasting temperatures and
mild oxidizing conditions. High temperature, strong oxidizing roaster
conditions can cause the formation of the less volatile arsenic pentoxide
(As_0-) that interacts with available metal oxides such as iron oxide (FeO)
to form stable, non-volatile arsenates that remain in the calcine.
26
-------
Multiple hearth roasters tend to eliminate more arsenic by
volatilization than do fluid bed roasters, although the difference between
the two roaster types is not substantial. It has been theorized that the
difference may be due to the local variability of conditions within a
multiple hearth roaster, which is assumed to give rise to areas within the
roaster that are favorable to arsenic volatilization. The increased
residence time of the ore concentrate in a multiple hearth roaster as
compared to a fluid bed roaster may also be, a contributing factor to the
enhanced arsenic volatilization.
The relative amounts of arsenic volatilized and slagged in the smelting
furnace depends mainly on the furnace type.and on the arsenic content of the
furnace feed material. Weisenberg, et al. made the following observations
concerning arsenic elimination from reverberatory furnaces based on the
arsenic content of the ore concentrate entering the smelter complex:
1. With high arsenic feed (> 0.2 percent), 55-75 percent of the
arsenic entering the reverberatory furnace leaves in the gas phase
and 10-25 percent is slagged out.
2. With low arsenic feed (< 0.2 percent), 5-37 percent of the arsenic
entering the reverberatory furnace leaves in the gas phase and
16-55 percent is slagged out.
The behavior of arsenic in an electric smelting furnace is reported to be
similar to that in the reverberatory furnace. The arsenic volatilization
from a flash furnace is approximately 76-85 percent of the arsenic in the
furnace feed while the amount slagged varies from about 7-17 percent of the
furnace feed.
Arsenic removal in the converter is dependent on the matte grade of the
furnace product. Low grade mattes (low copper to sulfur ratio) tend to
result in greater fractional arsenic volatilization than do high grade
mattes. This is believed to be due to the combination of the early
appearance of metallic copper in the converting of high grade matte and the
fact that arsenic is thermodynamically more stable in copper than in copper
sulfide. In normal practice approximately 70 percent of the arsenic
entering the converter is volatilized and about 16 percent is left in the
converter slag that is subsequently recycled to the smelting furnace.
The EPA has developed arsenic material balance data for a]1 14 copper
smelters contained in this study. The total fraction of arsenic entering
with the smelter feed that is volatilized in the smelters ranges from about
40 percent to nearly 100 percent.
3.1.3.2 Fugitive Arsenic Emissions. Fugitive arsenic emissions from
primary copper smelters are described in detail in reference 2. Based on
measurements made at ASARCO/Tacoma, the magnitude of the uncontrolled
fugitive arsenic emissions is estimated to be about 2.4 percent of the
27
-------
arsenic entering with the smelter feed. Table 3-3 lists the approximate
distribution of fugitive arsenic emissions among the seven largest fugitive
sources. Greater than 90 percent of the total fugitive arsenic emissions
are associated with the converter operations (about 76 percent) and the
smelting furnace matte and slag tapping operations (about 15 percent).
3.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS
3.2.1 Regulatory Impacts
The emission control systems currently in place in the primary copper
smelting industry are summarized in Table 3-4. Improved emission controls
will be required at several smelters in order to achieve compliance with the
regulatory baseline SO and particulate regulations. Varying degrees of
arsenic control are achieved by the requirements to comply with these
baseline limitations. The estimated impacts of existing regulations on the
copper smelting industry are discussed below.
Primary copper smelters are currently regulated by particulate and S0_
SIP emission standards as well as NSPS for particulate and SO... These
baseline regulations have motivated a considerable amount of retrofit and
redesign activity within the primary copper smelting industry. Most of the
proposed and completed plant changes have been aimed at the reduction of
stack SO- emissions by the introduction of smelting furnace process changes
and the installation of add-on pollution control equipment. In general, the
S0_ standards have caused the industry to shift towards smelting
technologies that produce offgas streams of sufficiently high SCL
concentration to be treated economically in contact sulfuric acid plants for
sulfur removal.
The impact of the regulatory baseline is expected to affect the level
of copper roaster arsenic control currently in place at two of the six
copper smelters using roasters: the ASARCO/Hayden and Phelps Dodge/Douglas
smelters. The need for the multiple hearth roaster at ASARCO/Hayden will be
eliminated when a new flash smelting furnace controlled by a contact
sulfuric acid plant is installed to replace the existing reverberatory
smelting furnace. The Phelps Dodge/Douglas smelter has not announced plans
for achieving regulatory baseline compliance and is expected to close down
in the future. The remaining four copper smelters using roasters (ASARCO/E1
Paso, Kennecott/Hayden, Phelps Dodge/Morenci, and Cities
Services/Copperhill) will be able to achieve regulatory baseline compliance
without changing their roaster emission control systems.
The largest impact of the regulatory baseline will be on the smelting
furnace operations. At least four copper smelters (ASARCO/Hayden,
Kennecott/Hurley, Phelps Dodge/Morenci, and Phelps Dodge/Ajo) have announced
plans to modify or replace their existing reverberatory furnaces so that
they can install acid plant units for furnace offgas SO control and achieve
compliance with the regulatory baseline. Two Arizona smelters
28
-------
TABLE 3-3. DISTRIBUTION OF UNCONTROLLED FUGITIVE ARSENIC
EMISSIONS AT A PRIMARY COPPER SMELTER*
FUGITIVE EMISSION
SOURCE
ESTIMATED UNCONTROLLED
FUGITIVE ARSENIC EMISSIONS
(as % of arsenic input
in smelter feed)
ESTIMATED UNCONTROLLED
FUGITIVE ARSENIC EMISSIONS
(as % of total uncontrolled
fugitive arsenic emissions)
1.
2.
3.
4.
5.
6.
7.
Calcine transfer
Matte tap
Slag tap
Converter operations
Anode furnace
Flue dust handling
Slag dumping
TOTAL
0.02
0.3A
0.01
1.83
0.11
0.09
Unknown
2.4%
1.0
14.0
0.5
76.3
4.7
3.7
Unknown
100%
Based on estimates at ASARCO/Tacoma, exclusive of fugitive emissions from the
arsenic building. (See Reference 2.)
29
-------
ll
1
N^
H c;
—4 [d
Z 6-
1 §
>_J y
CJ
rrt ""w
S P-
W H
H en
en D
>• Q
cn Z
C 0
OH Z
E— i t — 1
Z H
O fJ
CJ W
^£4
Z en
*^ ^j fj
1— t CG <;
CO W Z
CO PU °^
M PH £
so
W CJ «
U Jx H
^^ ^4 S
i-l S S
E-i 2* e/3
en M
^_| Q^
X PH
W
W X
a H
H
IHH M
o
w
{H O
fYj ^^
^^ . 1
W I-l 3;
W
CO
-3" O
! *
en
W
J
«3
H
<— c
-» CO O
F-H prt U IH
^ ass
CO
i.
01
IH
01
>
B
O
0
01 01
•O E co /-~
IH ^ iH O. IH -H ce 3
C •-' cr u co
cj 0) C 5
oo jr IH oi cu
E " *H !», O u
IH 3 3 U IH ^ (0
•o o UH eo 01 -H x
OS irt IH O.-H g)
.co 3 r* •**
4J
IH
B
CO
I
01
u
IH
01
CU
,C U CO IH O* Ou
oi x cr< '•^ cn i
u t, u co o eg
CJ
rt
E
01
CO
LJ
CO
J
E
01
o ^^
M —
0) -^
CL
r-t
n o
i» ^ IH
i-4 4J
O eg B
§CU •• CO 4J > I 4J > O
> W. 4J J= O 4JOU
IHC 3 B4JCCE O O 5
c. cj cu cj 5 s*^ u s
1 1 1
^ cj ea
O 1-4 >
C cu ^ e a
^-» eo .- oi -a
0]
cu
u
E
^ri
0
IH
CU
IH
01
01
C7\ W '^ w E
»— CO " CO
10 CU CJ IH
K u E cn — i o
E s cu u; c n.
41 01 4J -^. 41
"1 U O
•* IH B
1
^^
CO O
a. E E
•a 6 to
•rt CX IH
CJ iH CO
CO 3
O" U
CJ u E
•rt 01
Kb X CJ U JZ| 3IHIH
X CJ COJ2C3 CO UnCOCU
eo a. g « L» ^4 IH c.
PH eo i. -Hl-J- 3 r*
Cu in CA £ cS b |^ to IH ey*
K i-i CJ -
c ce : x IH u
s 4^ > -a a o i-i
U O n IH O. IH
o o e o c. 4-
JC en cu cj co \c ej
-1
ce. \ \ u
^^
en cj
E en
^i
01
CO
CO
O
*-
u
eo
—
01
a
i—4
2E
a -a ^,
<~- a ~a eu
CN IH IH -C3 CO
^ a OB IH
co eu 01
CO 4J I— ^* 4J 4J
SB O J3 -H E CO
CU O iH >^ 0) CO
4-1 U IH CtO CJ O
CO IH 3 B E IH at
>> 01 UH -H CO 11
co c. I-H u cu a..<-> -a
3 to IH ^ 01
cum eocuuiTxcj ea
c/3 r*- .^ IH a tjs s
U r-4 4J 4J CU *C
OS CJ CO O IH, 01
Z4J> C04JC04JC1 H
u 0 4J B UO -rt -H
r — • u eje-o^tu "3
I-H
1 1 U.
« US l-H
0) CJ CO O CO
U (0 4-1 >
CO 4-1 jZ o
c e w oo s
IH O -rt O\ CU
9 0 3 ^ IH
th
1
^^^
en
~~*
fj
T3 6
IH X CU
H E
eo to B.
o irt 3 E
•rt -rt cr 01
.H X 01 CJ
39 IH
UH CO 00 CU S-*
r-l S B*s-* 03
3 JS i-l x "—
co 4-i B a< cj
•rt CO W C IH
4-1 3 CU 01 0)
CJ i-4 O -rt 4J
CO 4-1 CJ 4J CJ CO
U B 1 -rt CO
c to eu co u o
CJ fi> Cb ' ' CU
o
1 Z
CO
0>
i-4 CO
CO
IH
01
U
0)
B
O
u
4-1
E
CO
i-i 4j 03
CO CO iH
CJ X C E
eo cu o co
iH IH I-H
4J &0 4J O.
oi E n
IH
jz eo o> 4J
0) *J CO
o- x E ec
>N u ce E •«
4J CO irt *rt CU
1 1 "O CL T3 -rt
CU
CJ
CU
IH
CU
•a E f-t i-i
cu o M IH a
X CJ 3 3 a
i-i eu o -a re
t4H CO UH
1 1
01
IH IH CO
•rt O *J 13 i '
3 « CO O CO
cr co o) 3 o 3
U frt E CO CO Si to
IH O O J= -0 J= CO
IH CO X O CU >! -O
to
01
u
CO
E
IH
3
0
0>
"
1-^
**
4J eu o 1-1 o O
E cu 00 A "O O
o & E 4-1 eo u jz
CJ CO IH IH M I— 1 W
f S O eu o IH
oi o a. -o eu a. o
> co 1-4 e -H: -a
1H4JI-I OB O. 3 O3 •• C. E
UEO Bco eejco 00 cs 3
•rtCO^ 1H4J I-44J B4J CC
OO^H a. o IH r-4
3 a. oi O.-H co a) co i-^H co
UH JE COCO*O4>|H CLCQ*OOO
&04J wuo4^^ eocjcco
u c OOeooi t»ooi-i
CO*IHUH oil— iJESfcH i— 4 .E eo
p- 4J O *J 00
— 1 CO 4J 1 | 1 tO 1 1
IH x »H z cn
CO CU CO
CO
o
to
u
01
4-)
to
to
o
ct
j:
4J
to
eu
X
01
irt
0.
i-l
i— 1
3
Z
co u
01 B
"O IH CO O
C CO W U
3 01
CO U 01
IH U IB p>
CO Ui TO —I
19 0 JJ
MJ= a IH
B CJ CU 00
•rt CO CC tJ 3
•O 1-4 3 CU U-4
o -a o BS >,
Of CO IH
JZ 01 O 01- &. -C UX
E u eo co eu 4J o
*-4 iH ^ U N X CO
CO CJ 4J i-4 01 01
cj i-l 6 1 1 -O O
O 10 CU 11 C 0)
rt CJ > 3 B E
I-H
I b, i
a
E
O
u
10
H
O
CJ
CO
cu
T>
3
CJ
E
i-l
4J
0
B
CO
01
c
a
30
-------
(Kennecott/Hayden and Magma/San Manual) are awaiting approval of the
proposed Arizona SIP for SCL emissions before announcing compliance plans.
If the Arizona SIP is approved, the Kennecott/Hayden smelter will most
likely be able to achieve regulatory baseline compliance without making
changes in either its smelting furnace or its smelting furnace emission
control system. The Magma/San Manual smelter will most likely require a
partial SCL scrubbing system on its reverberatory furnace offgases to ,
achieve compliance under the Arizona SIP approval regulatory scenario. If
the Arizona SIP is not approved, the Kennecott/Hayden and Magma/San Manual
smelters will be required to comply with a set of EPA-promulgated SCL
emission limitations. Under these EPA SCL limits, major furnace
modifications and acid plant installations at Kennecott/Hayden and Magma/San
Manual would probably be required to achieve compliance. The Phelps
Dodge/Douglas and Kennecott/McGill smelters have not announced plans for
achieving regulatory baseline compliance and are expected to shut down their
smelting furnace operations in the future. The remaining six smelters
(ASARCO/E1 Paso, Kennecott/Garfield, Phelps Dodge/Hidalgo, Copper
Range/White Pines, Inspiration/Miami, and Cities Services/Copperhill) will
achieve regulatory compliance without changes in their smelting furnace or
smelting furnace control systems.
The impact of the regulatory baseline will not affect the level of
copper converter arsenic control at the 11 copper smelters using contact
sulfuric acid plants to reduce converter SO emissions. Of the remaining
three copper smelters considered in this study, the Copper Range/White Pines
smelter will be able to achieve regulatory compliance without adding any
converter emission controls, and the Phelps Dodge/Douglas and
Kennecott/McGill smelters are expected to shut down operations.
The primary copper smelters are also required to comply with the
Occupational Safety and Health Administration (OSHA) workplace inorganic
arsenic standard. Agreements are currently being reached between OSHA and
five individual smelters that will result in improvements in fugitive
emission control systems at these smelters.
Table 3-5 summarizes the process and fugitive emission control systems
that are expected to be in place under compliance with all existing
regulations. The assumption has been made that the Arizona SIP will be
approved and that Kennecott/Hayden and Magma/San Manual will not require
furnace modifications to achieve regulatory baseline compliance.
3.2.2 Arsenic Removal Capabilities of Existing Control Equipment
The effectiveness of available process and fugitive emission control
systems in controlling arsenic emissions are discussed in the following
sections.
3.2.2.1 Process Emission Controls. The process emission control
systems used in the primary copper smelting industry include baghouses,
31
-------
CO
2
w
H
en
^H
en
,_j
O
w
H
S3
O
CJ
2
O
HI
en o
M 1-1
W CO
(3
a
P§ 3
CD 60
O CU
M Pi
to
53 r-l
O «
CJ >
O
W Cu
H P.
Cd
S to
en H
en
es
Cd C
to O
to
O *^3
cj cu
co
cd ca
53 CQ
>j
Cd
en
ca
Q
Cd
H
O
Cd
to
X
,
m
1
rn
Cd
Ǥ
H
C,
O
•H
CO
CO CO
•H r-l
€ o
Cd V-i
4-1
CU G
> 0
i-l O
4-t
•H
60
3
to
C
O
*rH
co
01 CO
1-1 1— 1
6 0
Cd i-i
4-1
CO (3
W O
cu cj
a
o
^
to
co
4-1
C
0)
o
Cu
o
u
4J
c
CO
iH
to
£*•*
4-1
-H
O
CO
ft
CO
CJ
l-i /— *
>
O 60
CJ £
*^s
Wi
cu
4-1
CO
•H
iH
pa
4J
C3
^ ^
f^
•o
c c o to
o o o to
•H i-l 33
4J w
CO CO o -f-
rH i-H CM
•H iH -^
4J 4J • O
C C CJ cfl
(U 0) • >
> > •< Cd
en
to Cd to
CJ
^
j2
o
Pi > CJ
33 Cd
2 Pi oo
o
o
o
o-
o
t-H
o
Cfl
CO
PU
r-4
W
•*^,K1
o
QJ
oi
en
•<
^
iH
>. C
-1 0
C
O t)
o
C 0
0 33
i-l
4J o
Cfl CM
i-l
i-l -O
4-1 CU
C X
a) 1-1
> to
cO
to to
^J ^
CtJ I>
33 53
38
,j
to m
o
0
o
ro
v£)
i — i
C
CU
T)
CO
"•^
O
CJ
Pi
en
•. 0
r-l 0
C
O V4
o
e
<; o cu
53 i-l U
4-> 3
CO 4-1
i— 1 Cu
i-l CO
4J CJ
c
cu o
> S3
to to to
< en <
Cd
J>
^
o
Pi > cj
PQ [--"]
to pi m
o
0
o
en
px.
c
CU
TJ
ct)
4J
4J
O
CJ
CU
CJ
c
> o
r-l CJ
C
O ri
O
(3
O CJ
•H l-i
4J 3
CO 4J
r-l Cu
1-1 CO
4J CJ
C
53
CO
to to
CO >
a 53
en O
< CJ
to -tf
O
O
0
en
^.
^
0)
I-l
I-l
3
33
4-1
4-1
O
O
CU
c
c
cu
>X4
r-l I-l
0 0
4-1 4J
C C
0 0
CJ CJ
V«4 W
O 0
CU CU
*J ^
3 3
4J 4J
Cu CU
CO CO
CJ CJ
o o
to to
en en
Cd Cd
r^*
52
O
o
jVjf)
gv! rvi
O
0
O
in
^J
CJ
iH
1-1
o
a
S
4J
4-1
O
0
CU
qj
C
CU
^
c c
0 0
•H i-l
4J 4J
CO CO
i-l iH
1-1 *rl
4-1 4J
C C
CU CU
t> k-*
to to
^
53
as cj
o
S3
4-1
4-1
O
O
CU
C
C3
(U
Ni
^
r-l
>, C
rH O
CD *O
o
£3 O
0 X
< -H
53 4J 0
CO CM
1-1
•rl T3
4-1 CU
C3 X
cu 1-1
> to
r*
to to to
CO
W
pi >
•z
T) O
pi cu cj
pa •!-!
to "4-1 r--
i-l
•O
O
0
o
o
1— 1
vO
1—4
•rl
CJ
C3
0)
t-l
o
J22
•^^
CU
60
O
n
CO
p.
I-l
cu
r^
to
iH
O
to h
to en 44
to Cd C
0
CJ
•C ^.
O
C C
o o cu
•rl -H V4
4J 4J 3
Cfl B) 4J
r-l rH Cu
i-l -H cfl
4J 4J CJ
C C
OJ CU O
> > 53
to to to
en en en
Cd Cd Cd
^
1z
Pi > CJ
§ pi m
o
o
o
in
,-H
i — i
o
CO
CO
rH
60
3
O
f-*)
•^•^
CU
60
O
P
CO
p.
i-H
cu
^«—
to
32
-------
•
TD
CU
3
G
•H
4-1
C
O
CJ
•
in
1
fl
w
tJ
^3
H
C
O
•H
CO CO
•H rH
E 0
Cd rl
4-1
CU C
> O
•H CJ
4-1
•H
00
3
C
O
•H
Cfl
CO CO
•H rH
e o
W rl
4-1
Cfl G
CO O
CU CJ
0
o
M
PH
CO
4-1
G
0)
C
o
p.
e
o
CJ
4-1
c
CO
rH
PH
^
4-1
•H
O
CO
p.
CO
CJ
rl ^-,
CU rl
P. >-,
d* "**— .
0 00
0 S
^•^
VH
0)
4J
Cfl
•H
t-H
P3
4-1
C
CO
rH
PH
^
rH
t^ C-
rH O
G
O T3
0
C O
O 33
•H
4J 0
Cfl CN
rH
•H TJ
C X
01 i-l
t* fri
CO
PH PH
CO
W
Pi
^
*"O tf*H
0) O
•H CJ
•H ro
•a
o
IS
o
o
0
. G
i— H O
c
O T3
O
C O
0 5C
•H
4-1 0
CO CN
•H -a
4J CU
C X
0) -H
^ PH
PH PH
< <
^
EC 2
CO O
< CJ
rJ
PH CN
O
O
O
•*
(^
CN
rH
O
00
rH
CO
-o
•H
JC
•^^
0)
00
*X3
o
o
CO
p.
rH
01
x:
PH
XI
rH
O
rl
4->
XI C
>> 0
rH CJ
O VJ
O
C
O CU
•H rl
*•> 3
CO 4J
rH P.
•H CO
4-> CJ
C
CU O
> 2
T3
CU
PH G
CO O
W 2
^
2
O
> CJ
§ CN
O
o
0
CN
00
CO
01
G
•H
PH
01
4-1
•H
f.
3
^^.
0)
00
c
cfl
Pi
^4
CU
p.
p.
o
o
rH
O
4-1
C
?*~t O
rH CJ
G
O rl
O
G
O 01
•H U
4J 3
CO 4-1
i-H P,
•H CO
4J O
G
CU O
> 2
CO
Pi
CJ
Cfl
rH
Cfl
•H
4J PH
rl <
CO
PH
•*
PH
Cfl
W
p>
$3
O
> 0
w
PH CN
O
o
o
ft
r-t
00
1 — 1
rH
CO
3
C
CO
32
C
cfl
CO
^^.
CO
s
So
CO
£3
rH
O
rl
4-J
C
?". 0
rH CJ
C
0 rl
O
C
O CU
•H t-l
4J 3
CO 4-1
rH P.
•H cfl
4J CJ
c
41 O
> a
PH PH
<£ <£
p>
^
O
o
Pd in
O
o
o
^JD
CO
^-,
•H
s
CO
•H
*jg£
**^.
C
o
•H
4_>
CO
rl
•H
p.
03
C
M
rH
O
rl
4-1
f> C
^ O
rH CJ
C
0 rl
O
C
O 01
< iH SH
a *j 3
CO 4J
1 — 1 Pi
•H cfl
4J CJ
G
01 O
> a
PH PH PH
<3 <3 <3
^
^
O
Pi cj
PH Cd CN
O
O
0
A
o
CXI
•-^
CO
C1J
o
•H i— 1
> rH
i-l -H
CU JZ
CO rl
0)
CO P,
01 P.
•H O
4-1 CJ
•H
O
OI
o
G
cfl
•H
i-H
P.
E
o
o
CU
rl
3
CO
G
•H
O
4-)
4-1
G
01
•H
O
•H
>4H
14H
3
CO
01
fl
t*^
rH
01
**?
•H
i— 1
4J
CO
O
E
-o
rH
3
O
4-1
3
0)
CJ
cfl
rH
P.
G
•H
^
rH
4-1
G
01
^4
^H
3
O
4-1
O
C2
Cfl
•H
T3
0)
4-1
Cfl
O
•H
""O
C
•H
4-J
G
cu
£2
P.
•H
3
cr
CO
•
01
o
•H
4-1
0
cfl
P.
rl
01
4J
rH
0)
E
CO
C
o
£2
Q
O
CJ
c
o
TJ
OI
CD
CO
•
CD rH
G 0
0 rl
•H 4-J
4-1 G
CO O
i-H O
3
OO U-l
CU O
i-l
CO
00 3
G 4-1
•H CO
4-1 4-1
CO CO
•H
X -a
0) OI
E
4= 3
4-1 Cfl
•H CO
S <
XI
p^l
00 rl
C 0
•H 4-1
4J CO
CO rH 4-1
•H 3 C rl
X 00 cfl O
01 OI rH 4J
rl P. CO
rr J_i 4J
4-1 01 01 T3 -H >,
•H > 4-1 -H P. E l-i
5 0) >>H O -H 0) CO
•H 01 CO O 4J T3
o) x: > cu co c
0 CJ CO ri >-> O
C cfl o -H p, co cj
CO CJ rl 0)
•H O SrlOrlOOCO
P* 4-1 rH X^ ^-* ^-1 "H C
E rl *H 3 B Cfl rH XI 'H
O 01 ECOC04J-HX1CO
O 4-i c/3 X* cfl UH 3 ^n
1 rH | 4J O O rl rl
C 01 OIO (-IOO3
O E OCfl>,4J-HCDO
C Cfl !H 4-J Cfl O rl
01 C i-l 01 X) CNi-4
O CO -HOP-rHcOO-H
4J 0) PnOCOCUUHCflCfl
c
CU T-l
3 PH II II II II II II II
01 •
01 4-1 *O ^ PH CJ PH PL, p^ o
CO -HOI JS^COCOpcnCJCN
O JS 4-1 O Pd tO ~»-
rH IS rH CJ
O 0) CJ
o x; co <;
4-1 4-1
0)
T3 4-1 >-l
CU cfl o 0)
4-1 O
O CO CU Cfl
01 c x: c
P- O 4-J rl
X fH 3
01 co 14-1 KH 0)
CO O rl U
O) i-l OI 00 0) CO
rl B 4J 4-J C O C
CO CU (2 CO -H CO rl
0) CO 4J G 3
Cfl rl 4-1 O rH i-l 4-1
CO CU C rl rl 01 3
i— 1 4-1 O 0) E MH 00
oo i-i o x: *-i cfl G
3 0) 4J CO 00 -H 0)
O ^ r! rl CO ^% C- 4-1 rH
O C3 COOrl'HrHXl
--. OMH 01rlO4-IO)Cfl
OI OrHX! 4-JrHEO
00 3 T) cfl 0) cn -H
13 i-ICOOICUrlB rH
O O rHXlOlCOCOP.
O U-H S P. Xl TJ P.
o -H T3 vi x: C co
CO "Oi— i 4-1-HOICOtO
P- OI rH3>CflM4-l
i-H 130) 3rHO)rHOO
01 OIX! El*Hrl'4-HGC
x: cu 4-1
PH G
id 4-1 o
C 0
co jE fQ fd CO O a
rH -H CO 1-3
•HO PH
O 4-1 01
O • C X) Cfl
23 CD co c
~~~ G rH 0) O
4-1 O P. O -H
4J -H G 4-1
o 4J *a cfl ca
O Cfl -H -H -H
01 rH O rH >
C 3 cfl p, cu
C 00 E rl
01 OI C O XI
K^ i-l ^ O XI
O T3 <3
33
-------
electrostatic precipitators, venturi scrubbers, contact sulfuric acid
plants, and auxiliary spray chamber/gas cooling systems. Although not
specifically designed for arsenic removal, the existing pollution control
devices at the domestic primary copper smelters achieve various degrees of
arsenic emissions control. The EPA has conducted a test program aimed at
quantifying the arsenic removal efficiencies of existing equipment at
several copper smelters. Arsenic removal efficiency tests have been
conducted on portions of the process emissions control systems at: the
ASARCO/E1 Paso, ASARCO/Tacoma, Kennecott/Hayden, Phelps Dodge/Ajo and
Anaconda/Anaconda (now closed) primary copper smelters. The arsenic removal
efficiencies of existing baghouses, electrostatic precipitators, venturi
scrubbers, and contact sulfuric acid plant/gas precleaning systems were
evaluated at these facilities. The inlet temperatures to the emission
control systems tested ranged from 80 to 350°C (175 to 660°F), and the inlet
loading of arsenic ranged from 0.07 to 885 mg/Nm . A systematic effort to
separate the effects of varying temperature and inlet arsenic loading on the
performance of the systems was not undertaken. The conclusion of the test
work was that baghouses, ESPs, and venturi scrubbers have essentially
equivalent performance capabilities for the collection of arsenic (96 to
99 percent collection efficiency) when each of the devices is operated at
temperatures below 115°C. The arsenic collection efficiency recorded for a
double contact sulfuric acid plant system at ASARCO/E1 Paso treating a
relatively concentrated arsenic stream (252 mg/Nm ) was found to be greater
than 99 percent. The extensive precleaning system associated with an acid
plant, which consists of a hot ESP in series with an impingement-type
scrubber, a cold ESP, and occassionally a venturi-type scrubber, is largely
responsible for the effectiveness of these systems in removing arsenic. In
other tests, a single contact acid plant at the Phelps Dodge/Ajo copper
smelter treating an extremely dilute arsenic stream (0.07 mg/Nm ) had an
average arsenic removal efficiency of 75 percent. Although the removal
efficiency of 75 percent for the single contact acid plant system is low
compared to the test results for the other devices, the stream being treated
had an inlet arsenic concentration more than two orders of magnitude less
than in any of the other tests. Since single and double contact acid plants
require the same level of gas precleaning it is assumed that they have
comparable arsenic removal capabilities when treating offgas streams of
comparable inlet arsenic loading.
The EPA has demonstrated the importance of gas temperature in
controlling arsenic emissions from copper smelters. An analysis of the
temperature effects on arsenic emissions control is presented in Chapter 2.
Test results at Phelps Dodge/Ajo show that an ESP operating at elevated
temperatures (above 315°C in the test case) can have good particulate
removal efficiency (96.7 percent based on in-stack measurements) but poor
arsenic removal efficiency (27.8 percent based on out of stack
measurements). Due to the volatile nature of arsenic trioxide, significant
amounts of arsenic can pass through a particulate control device in the
vapor phase if the device is operated at a sufficiently high temperature.
From the standpoint of arsenic emission control, offgas temperatures should
34
-------
be reduced as low as is feasible prior to entering a particulate control
device in order to minimize the amount of arsenic escaping in the vapor
phase. A lower bound on the feasible offgas temperature reduction is
imposed by the acid dew point of the gas stream. Although the SO content
of a smelter offgas stream is fairly low, dilute SCL/H.O vapor phase
mixtures condense to much stronger sulfuric acid mixtures in the liquid
phase. This situation can cause corrosion problems if the acid is allowed
to condense onto the particulate control device. Thus, the recommended
operating temperature of a dry control device such as a baghouse or ESP is
10° to 25°C above the acid dew point of the gas stream being treated.
Although the acid dew points of smelter offgas streams are expected to vary
considerably, practical experience at two smelters (Anaconda and ASARCO/E1
Paso) have led EPA to recommend an operational temperature of 110°C (230°F)
as within tolerable limits. Gas cooling is commonly provided by one of
three methods: spray chamber/evaporative cooling, dilution air cooling, and
convective/radiative cooling.
3.2.2.2 Fugitive Emission Controls. In general, fugitive emission
control systems consist of ventilation followed by particulate control. The
overall efficiency of the fugitive control systems is the product of the
ventilation capture efficiency and the particulate collection efficiency.
At present the overall efficiency of fugitive emission control systems in
the primary copper smelting industry is limited primarily by the capture
efficiency of the ventilation 'systems. In the case of converter fugitive
controls, the effectiveness of the fixed converter secondary hood systems in
place at domestic copper smelters was judged to be low. The best
ventilation capture efficiencies achievable for converter fugitives is about
90-95 percent using the air curtain/secondary hood technology developed in
Japan. Reference 2 contains a complete discussion of the test work that has
been done to quantify the effectiveness of copper smelter fugitive emission
controls in reducing fugitive arsenic emissions.
3.2.3 Inorganic Arsenic Emissions Under the Regulatory Baseline
The estimated baseline inorganic arsenic emissions of the copper
smelters considered in this study are shown in Table 3-6, The emissions
estimates range from 0.0045 to 5.2 kg/hr (0.01 to 11.5 Ib/hr). The methods
used to obtain the estimates are detailed in Appendix A-3.
3.3 ARSENIC EEC OPTIONS FOR THE PRIMARY COPPER SMELTING INDUSTRY
3.3.1 EEC Option Selection
Table 3-7 summarizes the EEC options for arsenic emissions from the
primary copper smelting industry. The EEC options for both process and
fugitive arsenic emission control systems were selected based on the test
work discussed in Section 3.2.2 and on the feasibility of the controls as
demonstrated at existing domestic and/or foreign facilities.
35
-------
TABLE 3-6. ESTIMATED TOTAL BASELINE INORGANIC ARSENIC EMISSIONS
FROM THE PRIMARY COPPER SMELTERS
PLANT
ASARCO/E1 Paso
ASARCO/Hayden
Kennecott/Hayden
Kennecott /Hurley
Kennecott/McGill
Kennecott/Garfield
Phelps Dodge/Morenci
Phelps Dodge/Douglas
Phelps Dodge/Ajo
Phelps Dodge/Hidalgo
Copper Range/White Pine
Magma/San Manuel
Inspiration/Miami
Cities Services/Copperhill
TOTAL BASELINE ARSENIC EMISSIONS
kg/hr (Ib/hr)
2.9
3.1
4.4
0.0045
2.0
2.0
0.89
0.96
5.2
0.53
0.45
3.1
1.0
0.011
(6.4)
(6.8)
(9.7)
(0.10)
(4.4)
(4.4)
(2.0)
(2.1)
(11.5)
(1.2)
(0.99)
(6.8)
(2.2)
(0.024)
36
-------
C/3
w
3
2
CO
fv'
w
P-I
o
CJ
s
l^H
M
OS
PH
o
En
CO
Z
0
H
O
U
w
•
1
ro
W
, t
CQ
5
OS
s
u
z
s
w
u
z
OS
e
c
z
H
CO
OS
E-
cn
<
O
cn
a
CC
a.
•o
h>
01
hi
01
E
0
J=
4J
•H
e
cn
i
u
hi
41
•H
a.
hi CA Y4
01 3 CJ £
> 1-2 CO CO 01
o a. to u
U 00 CO
C ~ h. X 0)
•H 3 3 hi
as r-t IH c
O 3 i-l -H
£ hi CO T-i E
0) X CO
X U 4-1 3 01
hi hi U tO I-I
to 41 tO U
-H C B 4J 41
h. O O 1* hi
O. u cj 3 G,
1 1
00 X-N
C J= X CO
H b -H 4J 13 rH TJ
CUb
CO
o
u
CO
£
3
b
f
o
CO
hi
01
hi
41
41
os
CO .O 4J 3 .E O CO 4J Z iH X
E "H co u to o x cj u g
C- CC O 1 CO 00 CO 4-1 E hi TJ tO CO 01
cn .C CMCU E 00 CJ *^ CU B -r4 4J
hi x .0 I oo w o E o> 41 - hi x cn
b O. tO • I-H O CM4I X hi 4J CO rH B
b K O CJ O 4J O M 4J CO i-l 3 — -rl
o at .a cn hi 4J cn xb co iJ c
••r*cuo Chi|ai4^B co *HCO
B i-H OOU — ^ > OC C C C5 E Q CJ 0) U -i
o o • ^- gai-HOoiiHo u IHU c i-i o
•H CJUWO OUJSUCO (X'O b hi CO 4-1 B 1
4JI-^O4J 0 4JCBCO)
a. o h o — h>
Ox-- x-v /~.O3OO.D.
rM CN en i-i b
00
C J2
S b i-l u
CU b i-H 00 *O CO
4J — 0 C J= B tO
CO
hi
01
4J
CO
CO
o
Ji
u
CO
01
s
41
l-t
a
•rl
rH
X
CO U O U4JCOt3 00
X 41 O hi ft i-l T3
cnja 4-I3COU -HX
E uj cn^cuto cjhig
OiCOO 1 CO 4-1 W C9CCCJ
co jr CNICU oo to 4-t -^ 4j
uutu c cn c oo cj U^HCO
^H cocoeu>».co hi i-ii-ix
hi XJ3 1 00 h It- 4J E 01 UXCO
Ocoto 3UUOECU 4-> 33
hi &.x-> OI4-ICO O4-I cn IMtflOO
bO.cO> r-HOIhiCJCO CO r— 1 B
b CO CJ U NO X O 3 J= -r-l
0 OIUO4JQCD 02 COUB
•• z - c- o scucnhiu -4co
CO T3 • cn i— I *H4-1|CU4-I "X? 4->3CU
gi-ieOb1— < J2 CO J= > T5 C CU Ur-l
O*^^ gcooOBBco ca couu
•H UCUhIO OOf-IOCUrH 4JB1
4J r -— o 4-1 cj tj j= cj co a. TJ ECOOI
a .H o i-< hi
O /^ x-^ 3 CJ O. O.
— 1 CN i— 1
a B
U 01
•« J= 4-1
B 4J CO
CO -rl X
X 6- 3 CO
01
hi
01
B
o
u
•£
«H
e
• o cn
00 41 U 00 00
C E BE
•H ht O -H
"O U-i O 1-1-0
O O co 0
Ob O
r-l 1H
CO hi 1-1 hi CO
•a hi s 3 -a
cn BOI u B
C': B O <4-l CU O
i o 0 o» a. u cj
i
CJ
hi
4)
iH
p>
•H CU hi X -rl CJ
4- cn a. 4-i co cn
c.
O ^* X—
| Vr* ^^
4-1 X
00 CO r-l
CO 3 *-4
*-i CO 3
CO J3 4-1
X
•CO) 01 hi
c j: O
CO 4J 4J
cn
01
a
4J
01
u
CO
£
3
!4~
l_(
r-4
^
hi cn
cu o a. o
4J O
1-1 rH C. J=
CO *H CO
6 tO •• 4-1 hi
'00 CU
J= • E rH CO T3
4J 00 i-l tO T5 E
C E S 0 0 3
.a -r* o o o co
O..-I rH f. l-<
hi Chi— 1
O to O 1 1
14-1 4-1 ^
1
CU
4-1
CO
01 3 O
01
cn
CO
c
u.
s.
hi
CC
CU
01
I— 1
a.
tH
3
X
E 0- 4J
11 01 Q-i
cj •o t3 cn
r-l CO CU W
CO U
CJ J2 C u
4J ai o hi
T3 IH •> CU
O3 hi 4J
co » a) cn
Q CO C 4J CO
r-l CU O r-l O
u hi 1-1 -H ai
E tO 4J U-l
CU CO T3
hi 1— I CJ CJ
x cu *H *H pa
l-H O. 4J h<
r-i o. e ,0 -o
3 O 41 CO T(
r-l
1 b
4>
3
4J
a.
CO Cu
o cn
B W
O CO
•H B hi
4-1 O O
CO 1-1
P-i 3 co b
cn cj co b
b* co «H
> E Ji
hi 01 01 4-1
O i-l
oo cu 3
b C _>
•O 4J 01
£ i-H -H 4-1
4J iH 00 CO
•H 3 3 X
3 ja u- cn
x—.
cn
cfi
•a
o
o
f.
CU
hi r-l
CU TJ
"O tO Cu
E rH cn
3 u;
ce cu
rH i-H hi
^ o
•O CO
0) *J b
co cj b
0 CO
rH hi O
U 4-1 U
C 41
0 hi 4J
B
1 01
1
X
rH
o
u
4J
c:
o
CJ
ra
B
o
4J -0
•H 01
•a hi
•a IH
CO 3
er
O 01
B hi
1
37
-------
3.3.1.1 Process Emissions EBC Options. Contact sulfuric acid plants
are recommended as arsenic EBC for any copper smelter offgas stream with a
SCL concentration of 3.5 percent or more. Acid plants have been
demonstrated as feasible for treating fluid bed roaster offgases; electric,
flash, and Noranda smelting furnace offgases; and converter process
offgases.
Baghouse or ESP systems operated at 110°C (230°F) or below are
recommended as arsenic EBC for primary offgas streams with low SO- content.
Available information supports the feasibility of operating the baghouse or
ESP unit at about 110°C (230°F) with no deleterious effects due to
corrosion.
3.3.1.2 Fugitive Emissions EBC. Fugitive arsenic emission EBC
consists of local ventilation appropriate to the source under consideration
followed by particulate collection in a fabric filter or ESP system. The
air curtain/secondary hood system listed in Table 3-7 as an EBC option for
converter fugitives has not been demonstrated in the domestic primary copper
smelting industry although it has been demonstrated in Japan.
3.3.2 Prevalance of Arsenic EBC Under Existing and Baseline Levels of
Control
3.3.2.1 Process Emission Controls. Table 3-8 shows the status of the
existing and baseline process emissions control systems in comparison to
arsenic EBC for each of the domestic primary copper smelters considered in
this study. At the level of control currently found in the industry only
five of the 14 smelters have arsenic EBC in place for each of the three main
sources of arsenic process emissions (roaster, smelting furnace, converter).
The emission control changes that will be made to achieve compliance
with S0? emission limitations will result in the application of arsenic EBC
for process emissions at several additional smelters. Two baseline
scenarios are shown that indicate which copper smelters are expected to be
controlled at the arsenic EBC level as a result of their efforts to comply
with SO^ regulations. Under the scenario that will result if the proposed
SIP S0? limits are approved, at least nine of the 14 copper smelters will be
controlled at the EBC level for arsenic process emissions once S02
compliance is achieved. Of the remaining five smelters, Phelps
Dodge/Douglas and Kennecott/McGill are considered likely to shut down
operations as a result of their inability to meet S0_ emission limitations.
Three smelters: Kennecott/Hayden, Magma/San Manual, and White
Pines/Michigan, are not likely to be at EBC for arsenic process emissions
when complying with the regulatory baseline.
Under the scenario that will result if the more stringent
EPA-promulgated SO limits are applied to the copper smelters, 11 of the
14 smelters will be controlled at the arsenic EBC level upon achieving S0_
compliance. Of the remaining three, Phelps Dodge/Douglas and
38
-------
C/BO
Z •>
h— t **
W CO
§ w
w >
w
CO
U 0
O P5
P5 E-
CU Z
O
Pi CJ
o
C&4 W
z
CJ M
PQ jJ
U <
H PQ
W Q
CO Z
*3l
O
P^ Z
O H
E-i
W to
U 1— 1
Z X
W W |
< 05
> W
W O
fy] £5
cu a
1
.
CO
1
en
w
3
c-l
J
c
C£
z
o
u
eg
z
lp_-
u:
en
CO
J
£
'g
O
0
z
c/!
s
en
5
i
• c/:
U E"1
Et. BO S
8C —
U
Q£ CM
>• H^ O
.J 3 m
^ U- <
i— * a c-
• en
O H
i-t >• M
b. ca s
O O 2
CJ g ^
>H h- C
24 U3 CM
(— « Ofi HH
j tn
z
C
H C
11
z <
c
u
z
3
ft.
CSI
o
tn •
•o u
CO l-l <*-« -*4
00 -H U
u-» er ex oo
O 41 US
U tJ u-i
4* X
U 4J 0) Li
« o w
C C "O 13 4J T3
W 0) 01 U O
u-t 3 ca a* > cc
X -H "HO *^
u CD u u u u
41 F- -H -r4 v4
01 -rt C CO C
WJ W O CC (C
ca u
cow we w
i-l CX 41 4) 4) • 4>
CC tO £ U OC
u) *H c c 41 -~i e
CO J- £. J= O t- X
C 4J U U U 4J U
c? c ex e
A o o o e o o
u u z z -H u z
en en en en
w: en en tn
tn to tn c/i
tn
u H
X > - C
H C ~l 0 ^
41 f-i *H N 4J
*^- o ^ ca o
O W (A U Wi •> U
U n 4) 41 "H -H 41
K &- -H o. a E c
^ 4U O. CO Q C
enrH -HO C*H 4> H
<; b3 UW t-^X ^6^
^
41 * <
C W -- B-
C W 4) CO ^
X 8> 41 O • C -*•.
(Q -0 O. U JD • C 4-. M U
iH 4) 3 Q.W*Hf4(Q
>M i-< 4> .a OJ O 1— i 3 E ^ C
r-4 U U 4J en<4-i*r4-i a cc u m «-• 3
n tn i-ios u c c(A O<-H
*O> O C O CtW*H O"O &•
cat o > o *-< >ow T3 c ^
• to J-i u QJ y f-i 4) < 4)003 ^O
•a p i [5 i^ 1-4 en c Q)
4) V4010-, '"'^hn^ O«H"O*i300
BJ u«^. eoc eoo» O!§M 3i-t
T-t « &. u *j j- *j>,4i c.u>3 a; ji •
U OW TO MOW tfiXC UCTC38!
•H WV4 C *H jp «H O W Uitn4l t~< V-i 4J
iJ ^ X»H X3 4JWjDCJ'H
c pO3 vuocv oo^ ts en ja K E
S 5*JiW 3*J C Cfl3C3-H
O U-ilM l4-t-H4> 300O t-t-ti-.
CO 13 Q- X 0 1 *-• C « *-• • IM-H U CO
oi < en >, c a nj -a 41 ^ w w ^ <
C 34) "— * *HOrH «H *T3 AJ »H P r-4 t^t—W
u c o ^ c is w c a 4-1 o a* 4-1 4-1*4)
>, p U 4-J W UiO.*?
o «3 c 41010 ai'-'oe o at a coi—c
z 3: "*- *-> ci 3 < QCO.U-H ZOC-H a.tn3
en en en en to tn en
C±3 bJ ti3 tt3 ttl til ti3
en en en cn en c O
tn O O O O O C
u z z z z z z
-»*. -^ -^. <
41 4) 4)
60 g 00 N CO
o <:*-izo o 4j^wo «< uc ^.c;
a, i— i o ai 4)4) o. c D. a a> K y~.
rHCOOS-O C t-t i-( 0* r-i» CT3 E
41T3<>vCI- 4IW 4VO C>% OCC
cu X tn o T3
> z i- «
o « «-< c a>
^-30 o z
0
^J= C t-
rH 4J O ^,41 *
(X,H -H to u -a
a. > 4J s •*-* "r4
jj CU 00 O *J
U u C W U C
o-*a o. cj o o-
f. W 3 U
^ w 41 O u O C
4) CJ 4J C >-i "H R)
t-H r- aj CD > T-4 a
o; QJ a. c >
E e x ecu o
tn en 41 z u t- Z
^- e- §
O
ceo
'
-^ -^ *-
41 4)
o *-j z ra c
W O - EL
CS. i— Ol 1—" 41 4)
41 3 C U C- -H
Ou C t^ 2 U S
39
-------
Kennecott/McGill would be expected to shut down their operations, while the
White Pines smelter would achieve SCL compliance without the application of
arsenic EEC.
On the basis of the above discussion, at most three of the smelters
that are expected to continue operating will not be controlled at the EEC
level for process arsenic emissions upon achieving compliance with the
regulatory baseline. These smelters are: Kennecott/Hayden, Magma/San
Manuel, and White Pines/Michigan. To reach arsenic EEC for process
emissions each of the three smelters requires a spray chamber/gas cooling
system to lower the operating temperature of existing electrostatic
precipitator units and thereby reduce arsenic trioxide vapor losses.
3.3.2.2' Fugitive Emission Controls. Fugitive arsenic emissions EEC
from both the furnace tapping and converter operations currently are not in
place at any of the domestic primary copper smelters. Although adequate
local ventilation is generally in place for furnace tapping operations, most
smelters (with the exception of ASARCO/E1 Paso) vent the captured furnace
fugitives uncontrolled to the atmosphere. Converter fugitive controls are
not well developed in the domestic primary copper smelting industry. The
existing secondary hood and building evacuation systems currently in place
do not meet the level of arsenic control demonstrated by the best foreign
converter fugitive control systems.
The fugitive emission control changes that will be made at the domestic
primary copper smelters due to OSHA arsenic regulations will result in the
installation of arsenic EEC at just one smelter (ASARCO/E1 Paso). The
OSHA/industry agreements call for the implementation of a push-pull air
curtain system for converter fugitive control at the ASARCO/E1 Paso plant
provided that the technology is successfully demonstrated at ASARCO/Tacoma,
where a test system of this type is now in place. This development will
leave ASARCO/E1 Paso at EEC for the two largest sources of inorganic arsenic
emissions, the furnace tapping and converter operations. All of the other
plants considered in this study will lack both an arsenic removal device for
the collection of ventilation-captured fugitive arsenic emissions from the
smelting furnace as well as an EEC system for converter fugitive emissions.
3.4 INCREMENTAL IMPACT OF EEC OPTIONS
The impact of the arsenic EEC options on the primary copper smelting
industry has been assessed for each of the 14 smelters considered in this
study. The impact analysis is based on the information presented in
Table 3-5, which summarizes the most likely baseline process configuration
and pollution control systems for each of the plants. The table is based on
the assumption that the copper smelters will achieve compliance with the
baseline SO- and particulate regulations as defined by their respective
SIP's. Incremental environmental, energy, and economic impacts associated
with the implementaion of EEC for the control of arsenic emissions from the
primary copper smelters will be discussed in the following sections.
40
-------
3.4.1 Environmental Impacts
By far the greatest environmental impact associated with the
implementation of arsenic EEC in the primary copper smelting industry would
be the reduction of atmospheric arsenic emissions. Table 3-9 includes the
achievable incremental arsenic emission reductions at each plant under EEC
control. It is estimated that the total arsenic emissions reduction
achievable for the entire primary copper smelting industry is about
114 Mg/yr (126 tons/yr) below the emissions level associated with the
regulatory baseline. The methods used to obtain the emissions reduction
estimates are presented in Appendix A-3.
The water and solid waste impacts associated with implementing arsenic
EEC are minimal. No additional wastewater would be generated by the new
spray chamber units operated at the three smelters requiring improved
process arsenic emission controls. All cooling water would be evaporated in
the spray chamber, leaving only a mud-like residue. This mud is then
periodically disposed of in a secured landfill along with other smelter
slags that cannot be recycled. From the standpoint of solid waste, there
are no other impacts of arsenic EEC implementation except for a possible
increase in the impurity content of the smelter slags. This could occur as
a result of recycling additional flue dust enriched in arsenic and other
volatile impurities to the smelting furnace, thus increasing the effective
feed of these materials to the slag-producing furnace and converter
operations.
3.4.2 Energy Impacts
Additional energy would be required to operate the new pollution
control equipment needed to attain EEC for arsenic. This would consist of
the energy used to operate the spray chambers for the process emission
control systems (water pumping and flue gas fan requirements) and the energy
required to operate the fabric filter systems and air curtain/secondary hood
systems for fugitive emission controls. The annual estimated incremental
energy requirement for a typical smelter requiring both process and fugitive
emission control systems is approximately 2 x 10 Gj (5 x 10 kWh).
3.4.3 Control Cost Impacts
The capital and annualized costs associated with the implementation of
the arsenic EEC option in the primary copper smelting industry are
summarized in Table 3-9. The total incremental annual cost in final quarter
1981 dollars associated with implementing arsenic EEC in the entire industry
is estimated to be $36.7 million. This estimate is based only on the cost
of the pollution control equipment required beyond that associated with the
regulatory baseline. The methods used to obtain the cost estimates are
summarized in Appendix B-3.
41
-------
C/2
{"v^
r-fi
c-i
W
Cfl
u
PH
PJ 1
O
O
*~.
Si
M
OS
PH
O
Fn
a
w
o
CO
0 1
CJ
•
P
W
N
I— f
^•3
l~^
p
•^ri
j
H
] -- (
^ri
ON
CO
U
ij
3
H
c u
ee t-i
c « a
t^ es w x~
b. § z
O 60 3C ^"*
r E v>
r- n: >-
O W
u a.
', a c
• u u
tN] DC
_! S3 H U
1 § £f §i
i*ir
88
H H a
s- p os z >-.
G l-< [=3 «•
0 3 E
U _: Q-O*-I
_j < E g o
< *•* o*>-'
H C. Bu U
to < O
£
j
u c
en > £
Z M H
-: to J o
BS w {£ ^^
< hi « CJ U
> S Z £
Id £ Z >- >-.
|_l O -3 06
gSpg£
«!§£
« a
S££
u
bl CJ
H a. c z
Z M U U
O 3 a: «
tf 3 ^
i1
BS
Z
cc o
z S
e^ e/3
W £ U
0 j*
gl
*
>.
H4
o
Si
&4
H
5=
1
1
1
t
a
c
o
z
CM
|
O
o
to
*->. (9
0 0-
tj
4 U
C/l
m m
o o
x x
CO CM
— • "T
g2
O *n
O O
O CM
» «
"^
•ff
— O
n
|
*n
"^- 0)
o -a
W3 ""
tn m tn
o o O
XXX
cs n r«^
•* m ro
<• o n
-^ sr «*
^r oo r^»
(^ 00 1**
PH O
^- CN PH
**
1
PO
r*.
^
w C
o
— **
\O fO
(M
O O
vO r*-
CO «ff
^4 O
o o
o o
o o
e*-* ro
in
O
O
O
1
f.
^
O 0)
U i— I
OJ V-
C 3
C X
in \o
c o
x x
-a- NT
ON m
o m
m co
O CM
NO NO
o <•
in
si
0 0
o
CSt
o
o
o
*
m
4-4
O -H
U i-t
Cj O
c u
X
in in
G O
X X
O ON
ft ^
2|
CM
O r~
NO CO
ON -*
NO
O —
— O
fM CO
O
CM
O
§
in
esi
*-* -o
O -H
U **-*
C 00
^ fi
*"•
— 0>
C C
ON
00
O
O
O
o
\
\ oc
o •*-»
g
U5 41
0. U
s *
m
o
x
CM
CO
o
ON
O
-a-
O
~*
m
O
CM
sO
ON
O
o
§
0)
0)
00
•o
o w
c «
D. 3
U O
JS
w
c: x
n u
?• o
CJ 4J
*3 f2
00
6) 41
M H
(C
4)
i*
to u
•H
C
C 41
•rH U
4J e
- Q
"O rH
(U O.
o
e u
o
^ OJ
to >
tn 41
"e .e
1) U
CQ
"3
C O
SO 4-*
01 03
U E
X O
* 0 AJ
o to
u n
S s
41 C
09 O
O U
^ to a
u c c c
O CO
O "H -H f-l
4J en K tn
co tn ic
»O «H «d -H
U 41 4) tU
O
41 cn u u
o. en «f* -H
X 0 C C
4i u ai ai
o to «
41 M t-i J-
M O. o «
cc
oO tn 41
to c cn >
U *H 41 i-* to
ai 4-1 u u c
U CO O f( O
i-J -r4 J-i GO i-(
a; x c. s tn
E 41 w-j w
tfi 41 -H
v- U i- E
tn -H co 41 C
O 41 C4>
O t3 WOW
•»•*. n; o a i-
41 t- 1M CQ
00 00 >-
T; a. >, o tu
O 3 *J M-i >
O ft «rt
O y [*< 4J
tO 4J CO [M -H
a- a. oo
—(0 eg JZ 3
0-
j: e « z — CM m
H O ^0 O
fl O
42
-------
r)
QJ
CJ
•H
4J
tf
o
0
a
1
1
pd
l _ J
H
in m
0 0
X X
o^ in
J -'
O CM
-T CM
—4 CO
*
CM
00 —
r—
so m
CM CM
-i o"
CM en
CM
m
O
o
o
•»
^o
0)
00
"C
0
CO
CL O
r
v°2°2
X X
f! f*l
0 P-
J -'
O "^
in J5 -• <
00 ft
^- o
c o
in
O
g
o
pj"
CM
OJ
00
•o
o o
Q 00
(0 (Q
C.T3
U X
Si
P-
m
x
SO CM ^
— * in oo
-J O 00
O m in
m ^ *•*>
•
O ^3 sC
-3- r- •£!
— ' O r~
-* \O
v£ in
1-1 O O
o o o
-- CM m
in
0
O
O
CM
00
a u
00 C
C -rt
n P.
«
at
>- u
1
OC >C
2 2
X
•9 \G •&
l--"1
O O C*4
in m »j
*
« ^^
^0
o •«•
*
o m
rn —
•M M4
™,
in c^
•tf O
O O
CN m
O '
«i
|
sO
f*1
C
o
•H
2 *e
f-i CO
O.-H
to ^
C
I-H
00 00
c o
X X
*n o^
CM — *
O in
in CM
\O CM
•>
^ r*t
r^ in
O oj
vO
1 I
O O
X X
r** ~*
CM m
O
O
§
O
o
CM
i-H
41 *U
co -H a.
01 > CL
•H U 0
•a
41
i— (
Q
CO
C
01
4) y
J5 «
C
>. V-
CO 3
E u^
JJ 00
CO C
.C i-l
r-4
0) E
U U
CO
SK CJ
t¥ S
•H 0
& •*-
3
U CO
u C
(A O
rH CO
CO CO
U £
M 0)
ra
a. u
-H
1) C
t-l
•U 0}
CC
JS C
4J C
CO 4J
e 01
CO M-i
CO &Q in w
T-l i-i CO CO
« 00 E E
> 01 01 0)
V- C
0) O CJ
CO Q i-> *H
C C C
O 01 CJ CJ
CJ > CO W
w a5 CO CO
c
CO «-* CO 01
»-H -- CO >
O. T- 0) 1-1 CO
3k cj u c
r-l O -H O
CO 0) »-i 60 *-t
3 C O> *H CO
C t- **- CC
CO r- 1 0) *H
co re cj a)
C co c u
to ^ u u u
**-. >• U-, > C
en ^ 3 OJ
BO * O CO
00 4J O O J-
w CQ u-* ro
s: ^ u
3 >, O V
x: ct 1-4 •*-!
*J t- CJ fi. 4-1
ra t-, MI
CO> C. CD
O J= CO X 3
•a -H
E u co
O3 «-< 0 C
H- O O Ui
Irf Ot — O .C 3
01 O >* CJ LM
CL C W -^O
co u; u CM u
CO IH 01 ~^. O
v-l *H H ^3 C "«
co a. z E -H
>v £ u: co cc E
CO U CU cj U> U
C H- 3 W
CO 01 Z3 >, CJ >^
> O" C W
*•* O) CC ^4 U<
CO fH O. -^ EK.
O .£ — 10 < X
cj u C
ca ei (I 11 u
at H
.£ O Z «— « CM ro
E-t *J C
U3 0
43
-------
The costs shown in Table 3-9 served as the inputs to an industry-wide
economic impact analysis. This analysis is discussed in the next section.
3.4.4 Economic Impact Resulting from EEC
An economic impact analysis was conducted for the primary copper
smelting plants to estimate the potential impact of requiring EEC. Using
the cost data in Table 3-9 and the methodology given in Appendix C, revenue
and product price increases, required by smelters to maintain the same net
present values before and after the installation of EEC, were estimated. To
test the sensitivity of the economic impact results to the various input
data, a sensitivity analysis was conducted for each smelter, assuming
simultaneous 15 percent decreases in baseline price and annual output and a
15 percent increase in the weighted cost of capital.
The annual revenue increases required by the primary copper smelting
plants to maintain their profitability range from $1.7 million up to
$4.1 million. The plant with the lowest required revenue increase indicates
the highest unit price increase of nearly He per kg (5C per Ib) and the
greatest percent unit price increase of 4.6 percent. The lowest unit price
increase is 1.5<: per kg (0.68C per Ib) or 0.6 percent. The revenue and
price impacts of EEC in primary copper smelters is summarized in Table 3-10.
The sensitivity analysis showed that plants would incur a slight rise
in the price increases needed to maintain profitability. The plant with the
EEC-required increase of 4.6 percent also indicated the highest increase in
the sensitivity analysis, 6.7 percent. Sensitivity analysis results for
primary copper smelters are summarized in Table 3-11.
3.5 POPULATION EXPOSURE DATA
The data in Table 3-12 were developed during this study for use in an
air pollution dispersion model to estimate the population exposure to
arsenic from primary copper smelters. The Strategies and Air Standards
Division (SASD) of EPA is using these data in an ongoing effort to estimate
exposure to arsenic from this source category.
3.6 EASE OF STANDARDS DEVELOPMENT
The major impediment to arsenic standards development in the primary
copper smelting industry is that the EEC options for the largest generally
uncontrolled arsenic emission source in the industry (converter fugitives)
have not been adequately demonstrated in the United States. In particular
the air curtain/secondary hood system recommended as converter fugitive EEC
has been used successfully in Japan but not in the United States. An
experimental air curtain/hood system recently installed at the ASARCO/Tacoma
facility has not yet been tested for arsenic removal efficiency. Until such
test data are available the feasibility of applying the domestically
available converter fugitive emission control designs is uncertain.
44
-------
O
ea
w
w
M
Si
O
o
H
C/3
PM
P-l
O
Pi
PM
Pi
O
C/)
H
CM
S
w
u
M
Pi
<
w
I
CO
r4
a
<
H
01
4-1 CO
a co
H rH
Pi CXrH
O
4«J *O
, i ^^/
c
3
0)
3
C .X"V
01 to
01 CO cO
0) rH
T3 )H O
O) O T3
t i j-4
H t*
•H -H O
3 0
cr o
0) vx
ttJ
rH 4-1
01 C
T3 CO
O rH
S Ca
— < r» o rH vo r- O -H
CO CM CM -H CO > -H 0) JH
01 01 rH 1-t O
T3 rH tH <4H S
C f^ M O >H -->
0) CO 3 0 CO 0)
•O 32 33 S O 60
^ **** "**»«. ^^. *^s^ *T3
CO 4J 4J 4-1 4J O
33 4J 4J 4J 4-1 p
-•» O O O O
O O U O O CO
CJ OI 01 01 (U PM
Pi C C C C rH
c/) cfl
co co co >H ~^ )-i
£X CX CX OI CO -H
rH rH rH (X S CX
Ol 0) 0) CX 00 CO
r3 A ^ 0 CO C
P-l &< PH C_) S rH
«3
sj-
00
o
r-H
ro
o
t — i
r^
rH
rH
rH
•H
^
CU
CX
Cu
O
O
•^^.
0)
o
•H
^
}H
01
cn
CO
0)
•H
4-1
•H
CJ
45
-------
CJ
U
I— I
Pi
U
as
u in
33 OS
H W
O W
C/3 C/J
H
CO OS
>-> W
I-J CXi
>> 5H
H OS
M <
> 53
M M
H OS
M O-i
C/3
w o
CO fi
I
CO
pa
0)
CO
4J CO
C -i i-l
OS fXi-H
O
4J T3
•H *^
C
3
01
3 /-.
C co
01 ^
> CU CO
flj MJ *vJ
l-i CO i-l
0) O
T3 )-i T3
cu o
1-4 C 0
•H -i-l O
3 0
cr- -^
O)
OS
iH 4J
4) C
•O CO
O iH
S3 P.
NO i/*t ON i— t ON
i-< CN CM CO O
— t i— i oo i— i oo
co in to NO —• *
o o o o O
O"\ ^5 co o r*1*
CO 00 O^ O O
r**- o u"i r** r^-
<• in co oo ^
CO CN CNJ ^-l CO
•d
r—(
>> rH 0)
Ol iH -H
C r-l -H >4-l
C «1) rl O H
01 TJ 3 0 CO
•a >, a S3 o
X CO -^ -- --
CO 33 4J 4-J 4J
W 4J 4J 4J
— A O O O
O 4J CJ O O
CJ a) 01 01 01
OS C C C (2
< (3 C3 C C
C/^ 01 CO 01 Ol
-i-l
S 0 < 3=
01 01 01 0)
60 60 60 60
*T3 *rj 13 T3
O O O O
O O O O
CO CO CO CO
p. Qi O« &i
i-l r-l iH r-l
01 Ol 01 01
f. r*. r*. r*.
PJ O-i PJ O-i
CM
CM
st
sr
o
CM
CM
CO
-*
CM
O)
C
•l-l
p j
01
•H
t*.
^
^%^
0)
60
C
cfl
^4
01
a
a.
0
CJ
ON
O
00
r— 4
O
o
CM
vD
CM
CM
i— 1
01
3
C
cfl
C
CO
c/>
•*s^
cO
g
60
CO
SB
ON
_,
p^
CO
0
0
O
CO
•
-------
^
6-t
H
^
A
ffS
bJ
PC
H
PS
O
En
3
O
w
f.'
»-*
S
as
<
CO
OS
1
K
O4
<
t-
to
o
M
E
H
*y
—
^
U
[£
S
0)
u
OQ
X
O
a
91
g
U*
a
to
_2
tZ4
.'
H
4J
5
^
V
1
CO
a
00
«
£
I— <
O
u:
H
S
as
§
0
c
o
«
l_,
JI
^
»-f
V-
^.
60
^
^
«H
fc*
•«^,
00
u
CJ
.^
4J
^
u
0)
"e
J
0
CO
o
CJ
to
m
S
fa
o
•""X
O
*£
^
e
a-
s
'i"
»i
C&N CO
0 x^ —
^-f ~I O*
u m •*--
a. co
e m o
CJ CM C
X
.a m
01 x-v
> W* CO
M • O
•H in *— *
60 CO
oi m c
2 • •
CM 0
in m
"%sV«. x-s
m «* so
04* co co
CM — -C-
v-x N^ >w
-JT sO CO
r-«- m -ff-
m
in — o
O O O
§c o
-c
* in -
o CM —
ON CM r*
m so
CM c m
^r — • m
-H CM CM
r- O O
-* o in
SK' S_^ ^^f
m so m
<— < CM CM
CTv — sO
co so in
<** -- s^
CM -tf —
— CM CM
— CO O
m CM O
— CO — *
O CM -H
•c in cn
CM
Jrf
o ,*;
1 CO U
^ 4-1 (C
l-i Cu
25-e .
C3 0) > C
A iJ 0) O
£ C Qi CJ
3 W 3 3
CJ CJ CJ
m in
\o r*
m m
m
01 4->
C i- ij
O O CJ
tsl X >
X
„
c
O w
CJ CO
o£ c-
*c w
in os
x^. CM
in — '
o o
m m
CM C
O O*
in os
x-«. CM
in —
c o
m m
CM O
o a
m
""^
ON
CO
in
*^
ON
r*
m m
o o
c c
c o
m CM
-i 4J DO
CtO 3
0 fa
> CC
C UJ •
0 0 j=
0) ej Jj
Qti O 01
W >
.0 CU 0)
W OS
in
m
csi
C
N.*
~4
~*
O
in
in
CM
O
.— '
0
CM
""*•"••
o
1
O
CO
CM
m
O
in
— «
*»*x
Os
m
m
m
CO
CM
CO
CM
in
CM
to
>.
CO
• c
00 O
13 f*
l-i 4-J
CO CO
3
> CO
c ?•
o w
u
cn
— in
_T —X-v
x-* ^M r*
. o" —
^ -*^ %«x
m r*
00 O r*
O O O
m ON in
r* m m
— • O CM
CO — P-
O O m
x-s
in
^H
— '
m S m
CM
O
C-
CM
m
CM |D
o o
o o
o o
m in
0 0
m m
J£ 60 C
030
CO fa Ot
W CO
CO ^
C 0) >
•H > C
CO 01 O
Z JC u
CO 0
CM ON CM
— CM in
m so
m
01 *J
C *-t V-
O 0 CJ
tsi ar >
<
o c
CJ 01
OS -C
CO to'
< X
m m m
^%s"x-s"l~ .
MT ** in
C CO O •-•
o • . .
to C O C
I-
co co CM r-
am o o
e ...
o o o o
u
i— 1 x-s CM CO
-H - 0 —
oo r- •— ' N-X
.— CM — •
60 in — CO
iy ...
S m o O
____
CM x-s
• in
m ON
•**> •**•
• c in in
CM m os c\
m m m m
CM
CM — i
"%. S"
• r*
c, c
• 4) >
(0 < > > C
o at o o
PS PS os u
»n M
p )-i ^
o o a>
CM SC >
^ 2
o
u c
01 0>
o -o
£ >.
0) (0
^ s
m m
m •$• •&
s*. C C
O XX
o in o
. c • •
o o — <•
s_x cfl %^x^
PM PI t
o re o O
O O CO CO
sO —
m s-s
-H &
x-s. OS ON
in o) x-s x-s
O .Ol 1
. ^ o o
O 00-" --
— ' v4 X X
CM 00 • •
O 2S ^N~-
O i I
00
X X
x-s U~i CM
m
r*.
2S
CM
CM
O CO
o o o
o o c
* • *
CM E; s-r CM
•sT
O r*. CM
O O O O
o in in m
^^»*
PM m en m
CM PM
*-" *—
oTcM1
m O
CM «-•
0.-
r^
CM CM
CO O
O O
in CM
m —
m so
u
sx j_i hn .
U CO 3 tO
CO 2 .3,
C ?> 41 >
•H P > C
co o at o
X u DS cj
CM (M
CSJ CTN -H*
-- vD CM
— sO
Ot 4-J
C M M
O O 0)
Cs] IK >
ui
O "
C rH
C M
^ IS
2
m~m
I -*
O -^
en y - ' *«-^
"*— '1 ON
O —
co — o
• X •
*-* m o
en x-*
o** » in
. o •
m *— ' o
CO O CM
— o o
jf^k
o
sO
•
"»
CM
O
in
r»
o\
CM
CM
* .
^ 00 >
O 3 60
« fa 3
4J fa
C OJ >
•H > C
CO 0 0
X ft£ cj
CM r**
— -H m
\o m
•o-
41 4-»
C 1- l^
O O 0)
NJ a: >
-u >
o -
'J FH
01 1-1
C ft
c cj
^ z:
m m
x-v x— >
>-.C CM
• O O
^
m o
O 0 O
• o •
— i •— ' CM
CO
c o —
X—.
o
r--
m
m
_J
"*
— m
x^— m
O x^—
O O X-N
coo
-co
C * 0
O CO *
O CM r-
*— sO
-4 s^ CM
in
CM « sO
P— O CM
O O O
in in in
^Z^
en en m
^
«i
ON
m
CM
m
^
—
§
CM
^,
MP
m
60
O fa 60
rc 3
U CJ U.
t/1 CJ
C C >
« 3 O
s: fa cj
-T r*
CM ON m
m m
*T
(U *J
c >»> u
O O 0)
CM X >
.£.
0 — i
0 -H
OJ CC
m x~* X***
— —. o
x-^ mm
sO O —
-- ON ON
—t -- CM m
o o o o
o c o o
o o o o
Q CM — r-
CM r* CO so
m -T "^
co P- > >
o < o> 01
01
c
o
Csl
C
C
47
-------
*O
0)
2
c
•H
C
O
CJ
.
f>4
i—4
1
CO
W
5
H
m m en en
NO vO fM ON
4 -^
P- CM • P-
en — r-* in
_ M CM _*
^« « m cn
>— S — V ^N, ^-.
O m in in
*? m •« *•*
O O ON ON
co m cn cn
*» .*• m m
CM CM
— «
£"0"
CM CO
CM — *
r-- in
vO in
CM CM
1* >
on n > a) >
C3 4J c > C
o co o o» o
d cj cc o
in NO
CM «tf CO
m
o> —
C w t->
O O V
01
GO M
*0 <
0
a *
cn
(fi 1C
a. I-H
<— oo
O> 3
E £
m
<* /—.
en en ^* r-»
^* •* ON CM
cn en • o
>— "* cn
«T CM
O m O •—
CM — < O O
cn en ON
/"* **••» ST ^""i
cn cn \o cc
-T en o •
O m CM •&
CM — O —
m
CM' a
CM Z
NW
ON
NO
cn
-^ ^5-
0 CO
o a CM o
•iT Z fl CM
CM ^ w
CM in
-T • •
-* in <•
— — ON
-- en c^
CM CO
— in in
CM a — —
£ZJs
en cn cn
— ,
— i
^
in o
— z
NO
> u >
> c > c
0 O 0 O
ce u a£ u
vO r-
CM m CM
— CM CO
m m
en
4) 4J
C U U.
O O U
M S >
. *-^
— o
/-» e "^ o
r^ O O •
• W >^ C
O "^ ^**
•^^ w in
a o CM
en ao O
0000
u
>^
JJJs ON
*-* r-l O — t m o
cn oo O CM
O Z O O
^"in
•' M
C U
O eo
U "
— u:
c ...
u jc e oo oo
3 CJ ^ 3 3
fa CO 3 ft* tbl
in oo c >
— I < •-" 3 O
U. w [Z u
0 0
CM U. 00
U 3
. V- «
> a >
c > c
O 01 O
u a- u
r^ co
g3 CM o
— 0 CO
m
1) •
oo tn
C 0)
1C C
ee -H
OM
p *4
o. ^
eS5£
cn
lff*.
~+ — o ^*>
^^ <-^ • CM
ON in O •
O O N—p
Xrf- %»X f^
o —
-or CM o O
O O O O
ON ON
*— v /•»• ^ CM
CM CM O *
. . . O
O O O *••'
CO — * •— ' -^
• * CM •
CM O O O
d
en ao
ON m
ON ^
in (*\
-^ m
"Vv O ^-^ O
o o o o
§000
-o *
o co -^ m
0 -* ON ~>
cn ^ eo
ON . . •
oo co cn CM
— co -a- NO
m m o C
r-. CM in m
ON in ON ON
— CM en cn
in cn cn rn
CM
— CM
o *-»
CM ON
^»
*~
CM CM
00 ON
0 -tf
m CM
in so
in r*.
jfi y
0 03 •
(0 4J 00
CU 3
. < . u.
P < U >
0) > 0) >
> c > c
0) O 0) O
06. 0 K U
r-t
01
3
C
cfl
«C X
g *
00 C tM"
s 5 "*
en m
x— >. s~*
0 • .
• o o
N^ ^"'
— ' •*
-TOO
O O O
ON
x-ON
ON CM y-s
*-% CM -H
JSC
O ON
-sO
GOO
O O O
-« CN en
P-- -C- 00
cn CM CM
m en NO
ft m cn
ON O O
CM u"\ in
— — « --
P*. ON ON
CM en en
en en cn
^4
—
CM
O
•-1
cn
-
o
o
CM
O
NO
>
C
o
-»- ^ 00 00
.033
C « b- t*.
3 cn • •
U. c ;»
o- u c
. < 3 o
Ul f*i O
-if r«
CM en \o
en
0) w
C J- w
o o a>
N X >
c
u <;
0}
P -
•H "-*
o. S
cn to
C -H
m m
*•*> x^-
m «^-
t i
O O
— ' •— X X
O on 'C-
O Oil
. o O
0 0 - —
"** ^ X X
en esi
2 22
x^ *~*
^
o oa- v»
« n'oU
O O O O
K X
CM OO
cn m cn
O O -N o'
o o o o
o o c o
-o -
— — 0 cn
V— ' -^^ V^, *^.
P*. P*» 00 00
cn cn *v CM
m m \c
cn rn en cn
O O O O
in tn m in
— — ' — ' — •
en cn en cn
rn en en cn
0 0
o" o
O m
fs) CM
O CM
— * sO
vO P«"
00 be
in
*• . •
o. e- 3 o
O 00
vo o
U £.
(A fH l-i
•H u a,
u 4t O, 2;
S" cS*-
-------
3.7 REFERENCES
1. U. S. Bureau of Mines. Copper. In: Mineral Facts and Problems.
Preprint from Bulletin 671. 1980.
2. U. S. Environmental Protection Agency. Arsenic Emissions from Primary
Copper Smelters - Background Information for Proposed Standards.
Preliminary Draft. Research Triangle Park, N.C. February, 1981.
3. Weisenberg, I. J., P. S. Bakshi, and A. E. Vervaert. Arsenic
Distribution and Control in Copper Smelters. Journal of Metals.
21:38-44. October 1979.
4. Telecon. Cassidy, M., Occupational Safety and Health Administration,
with Keller, L., Radian Corporation. January 29, 1982. Occupational
Safety and Health Administration arsenic Project.
5. Telecon. Cassidy, M., Occupational Safety and Health Administration,
with Keller, L., Radian Corporation. March 26, 1982. Occupational
Safety and Health Administration arsenic project.
6. Memo from Rathbun, R., EPA, to Pratapas, J., EPA. December 21, 1981.
16 p. Copper, lead, and zinc smelter information.
7. Telecon. Titus, J., EPA, with Keller, L., Radian Corporation.
February 18, 1982. Compliance status of primary copper smelters.
8. Personal Conversation. Vervaert, A., EPA:ISB with Keller, L., Radian
Corporation. March 5, 1982. Primary copper smelters.
9. U. S. Environmental Protection Agency. Preliminary Draft Report -
National Emission Standard for Arsenic Emissions from Primary Copper
Smelters. Research Triangle Park, N.C. EPA Contract No. 68-02-3060.
1979.
10. U. S. Environmental Protection Agency. Design, Operation, and Emission
Data for Existing Primary Copper Smelters. Research Triangle Park,
N.C. EPA Contract No. 28-02-2606. March, 1978.
11. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for Emission Standards and Engineering Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency.) EPA Contract No. 68-02-3173. May 1982.
12. Memo from Schewe, G. J., EPA, and A, E. Vervaert, EPA, to Tikvart,
J. A., EPA. March 24, 1978. Air quality estimates of arsenic due to
primary copper smelters.
13. Memo from Keller, L., Radian Corporation, to arsenic file.
May 27, 1982. Arsenic emissions from primary copper smelters.
49
-------
14. U. S. Environmental Protection Agency. National Emissions Data System.
Copper Smelting. January 7, 1982.
15. Letter frpji Richardson, J. B., ASARCO, to Fensterheim, R. J., Chemical
Manufacturers Association. May 17, 1982. ASARCO response to Radian
emission estimates.
16. Vatavuk, William M., and Robert B. Neveril. Estimating the Size and
Cost of Gas Conditioners. Chemical Engineering. January 26, 1981.
17. Vatavuk, William M., and Robert B. Neveril. Factors for Estimating
Capital and Operating Costs. Chemical Engineering. November 3, 1980.
18. Marsh, A. 0., Jr. ASARCO Incorporated: Converter Secondary
Hooding/Tacoma Plant. Salt Lake City, Utah. January 22, 1981.
50
-------
4.0 PRIMARY LEAD SMELTING
This chapter discusses the primary lead smelting industry and its
associated arsenic emissions. Section 4.1 presents a brief source catagory
description. Section 4.2 discusses the regulations currently covering the
industry and the resulting level of control and associated arsenic
emissions. Section 4.3 presents a discussion of the EEC options.
Section 4.4 presents the input data to be used in assessing the population
exposure to arsenic emissions from primary lead smelting. The ease of
standards development is discussed in Section 4.5.
4.1 LEAD SMELTING DESCTIPTION
4.1.1 General Information
Metallic lead is produced from lead ore concentrates at primary lead
smelters. The United States' primary lead smelters produced 578,000 Mg
(636,000 tons) of lead in 1979, which accounted for 41 percent of the total
domestic demand for lead in that year. There are currently four companies
and five individual plants that produce primary lead bullion in the United
States. Table 4-1 lists the names, locations, production capacities, and
estimated total inorganic arsenic emission rates of the plants. The
existing total annual inorganic arsenic emission rate from the entire
primary lead smelting industry is estimated to be about 43 Mg/yr
(47 tons/yr).
4.1.2 Process Description
The materials processed at the U. S. primary le^j smelters are
predominantly sulfide ore concentrates. The most important lead-containing
mineral in the ore concentrates is Galena (PbS). Commercially significant
raw lead ore as mined is typically 3 to 6 percent lead by weight, and
gravity/flotation methods are used to concentrate the ore. Two major types
of lead ore concentrates are smelted in the United States: Missouri lead
ore concentrates, which account for 83 percent of the domestic primary lead
production, and western lead ore concentrates, which account for 10 percent
of the domestic primary lead production. Concentrates originating from
foreign lead ore account for the remainder. Lead ore concentrates from the
Missouri lead belt typically contain about 75 percent lead and 15 percent
sulfur by weight, with silver, copper, zinc, iron, calcium, magnesium,
arsenic, and acid-insoluble material making up most of the remaining
10 weight percent. Western lead ore concentrates are of poorer quality than
Missouri lead ore concentrates and contain about 45 percent lead, 10 to
30 percent sulfur, up to 15 percent zinc, 8 percent iron, and 3 percent
51
-------
Cfl
pi
W
H
,J
W
a
(**>
^i
u
^4
cn
Jg
M
Pi
P-i
M
H
Cfl
W
a
O
p
W
H
•
,_^
1
U
03
^
U
i-H i-l
A CO
4-1 CU Ol
O CO 4J
H K CO
< Pi
CU O C
4J -H O
Cfl C -H
£ CO CO
•H 60 CO
4J VJ 1-1
COO6
W C W
M
CO
^
4J
1-4
O
CO
CU
cfl
U
C
o
•H
4J
O
2
T3
O
V"<
Pi
s-^
l-i
!>.
C
o
4-J
/""s
1-1
>*
00
SS
*"""'
x-*
^4
>,
^v^
CO
c
0
4J
SvX
^-^
1-1
>!
"oo
g
C3
O
i-l
4J
Cfl
y
o
j "^
4_1
£3
cfl
P-
u-i
cn
•-1
cn
•
CM
.— 1
O
O
o
o
0
o
o
CM
co
cn
cfl
^
CU
H
««
O
cn
CO
P-i
.— 1
W
0
co
Cfl
P-J
iH
W
8
pi
^
W
^
01
PM
TD
0)
o
3
T3
0
i-i
ca.
T3
CO
0)
rH
M-l
O
CO
e
cu
0
•H
•^
cu
4-1
CO
v4
^t
^^
CO
cu
3
60
•H
U-l
^^
4_)
1-4
O
CO
0.
cfl
O
p^
o^
r— 4
«
•
^
^
i-l
C
CU
Cu
a
C3
•H
c;
5
O
4=
CO
CO
*TD
o
4-1
cu
e
cu
r4
4_)
(3
0
T3
CU
CO
cfl
_o
cn
cu
4J
CO
^1
c
o
i-4
CO
CO
•H
G
cu
o
,-J
*^
c
0)
CO
1-4
cfl
I-l
cfl
3
c
c
cfl
-a
cu
4-1
CO
e
•H
4-1
CO
Cz3
^3
52
-------
copper. Other constituents of western lead ore concentrates include gold,
silver, calcium, magnesium, antimony, bismuth, and arsenic.
Arsenic generally appears in the form of arsenopyrite (FeAsS) or
arsenic sulfide (As?S_) in lead-bearing ores. The gravity/flotation methods
used to concentrate tne lead bearing fraction of the raw ores also
concentrate arsenic. The arsenic content of Missouri lead ore concentrate
is typically about 0.02 percent by weight, while western lead ore
concentrates contain about 0.1 to 0.4 percent arsenic by weight.
Fugitive emissions associated with lead ore handling and storage are
controlled by ventilated enclosure systems and wet suppression methods. Ore
transfer points are generally hooded and vented to fabric filter or venturi
scrubber systems.
As shown in Figure 4-1, primary lead smelting is a complex process with
several distinct operations. The three key operations common to all U. S.
primary lead smelters are sintering, blast furnace reduction, and dressing.
Other significant process steps are those associated with the dross
reverberatory furnace, the slag fuming furnace, the deleading kiln, and the
lead refinery. Impure lead bullion produced by the smelting process can
either be refined on-site (as is the case for the Missouri lead smelters) or
shipped elsewhere for refining (as is the case for ASARCO/East Helena and
ASARCO/E1 Paso).
The first step in the processing of lead ore concentrates is sintering,
which is carried out in an oxidizing atmosphere at a temperature of about
600°C (1100°F). Lead sintering is essentially a roasting operation that
also serves to agglomerate the feed materials into a form suitable for blast
furnace reduction. Several functions are performed by lead sintering:
(i) a large portion of the sulfur in the ore concentrate is removed as
SO., or S0_, thereby allowing for the opera^on of a low
temperature, high efficiency blast furnace operation.
(ii) readily melting complex silicates are formed that agglomerate the
charge into a hard, porous blast furnace feed.
(iii) species such as cadmium, arsenic trioxide, and antimony trioxide
are removed by volatilization.
Updraft-type sintering machines are used exclusively in the domestic
lead smelting industry. The ore concentrate charge is spread onto the grate
bottoms of iron pallets that move through the sinter machine o^ a circular
belt. Air is introduced through wind-boxes located below the sinter charge
and passes through the charge to participate in the sintering oxidation
reactions. The sintering reactions are sufficiently exothermic to sustain
the desired temperature of 600°C (1100°F) once the charge is ignited.
Approximately 80 percent of the sulfur in the ore concentrate is eliminated
as SO^ from the sintering machine. The product sinter is screened to remove
fines and sent to the blast furnace.
53
-------
0 8
z tn
b£ U.
d °
I!
i s
* §
id u:
u >
< OS
< c
^^
Ul
IS
<
< %
ty, b.
^
as u:
Hfc
^
z
a -J
U5
QC
U
CC
<
tu
)
x.
a
<
9*
UJ
CC
z
<
u
p
J
.j K:
3: r-
g _j
Ss
1
§
5-
g
a
!I«
en H a.
N4 O tj
is
Ul U.
au.
0
ui
OS <
CO U?
en H u
St- b.
U b<
HOW
SZ U
ce u.
JO
M
S3
2§
11
JSSED BULL
TO REFINER
J ,
,..
S £
O J
§ I
Z i
w
a
o
5
!-
<
S
f
*
SINTER
u
Cc
c
S
*:
^J
z
as
u.
^
NTHDVW
31VHLN3D
C- _• I-I Q
< ^ ^ 5
AU'.
IS i
S
H u
W < CJ
c u z
0£ CQ 2
S
o u
c z <
< i-, Z
— z 2
"££
11
j_ *
U5 Z >•
=1 J OS
LJ " 3
r- Z ^J
J — Q£
^lg
5 o
3oi
u. f- <
\
>.
•j*.
o
co
cc:
I —
i
UJ
1 1 1
e£
LU
1 —
oo
UJ
00
c
0)
CO
CO
0)
cfl
•H
I
-------
Four of the five operating primary lead smelters use contact sulfuric
acid plants and the associated gas pre-cleaning equipment to treat some or
all of the sinter machine offgases. The ASARCO/Glover lead smelter has no
sinter machine SCL control, sinter machine offgases at ASARCO/Glover are
sent to a spray chamber/fabric filter system operated at 125°C (257°F) for
particulate control only. Three of the four smelters with acid plant units
(ASARCO/East Helena, AMAX/Boss, and St. Joe/Herculaneum) treat two separate
offgas streams from the sinter machine. A concentrated SCL offgas stream is
removed from the inlet end of the sinter machine and treated in an acid
plant circuit, and a dilute SCL offgas stream is removed from the outlet end
of the sinter machine and treated in a fabric filter system operated at
100 to 125°C (212 to 257°F). ' It is not economically feasible to treat
the entire offgas stream from a lead sintering machine with a contact
sulfuric acid plant unless a technique is used to increase the SO content
of the stream to an acceptable level of 3.5 percent S0? or greater. The
ASARCO/E1 Paso lead smelter makes use of a sinter machine offgas
recirculation technology that allows for the production of a single sinter
offgas stream containing 4 to 5 percent S0?. The entire sinter machine
offgas stream at ASARCO/E1 Paso is treated in a contact sulfuric acid plant
circuit.
Hooding and enclosure of emission points followed by particulate
removal using fabric filter or wet scrubbing systems are the commonly used
fugitive control techniques associated with sinter preparation and recovery.
Sinter machine fugitive discharges are generally controlled by total or
partial enclosure of the operation.
Screened sinter product is mixed with coke, recycled slag, silica,
lime, and recycled flue dust and is fed to the blast furnace. At the
operating temperature of the blast furnace (980 to 1035°C) the charge exists
in a molten state, and several oxidation/reduction type metallurgical
reactions occur. Lead oxide is reduced to metallic lead by the action of
coke. Iron oxide is reduced to metallic iron which in turn reduces lead
sulfide and lead oxide to metallic lead. The net result of the blast
furnace reactions is the production of a metallic lead-rich phase called the
lead bullion and a slag phase containing complex silicates of iron, calcium,
zinc, magnesium, and aluminum oxides. Depending on the non-lead impurity
levels of the ore concentrate and on the operation of the furnace, a
copper-rich phase called matte and a copper/iron/antimony/arsenic-rich phase
called speiss can also be formed. In most cases lead bullion and slag
represent the major products from the blast furnace. They separate on the
basis of their mutual insolubility, and the slag layer floats c.bove the lead
bullion. The various phases are removed individually from the furnace.
The SO content of blast furnace offgas is typically on the order of
0.05 volume percent and too low to facilitiate treatment in a contact
sulfuric acid plant. Particulate removal at all five domestic primary lead
smelters is accomplished by treatment in fabric filter units operated over
the temperature range of 100 to 125°C. Fugitive emissions from the blast
55
-------
furnaces are controlled by building enclosure of the source and by the use
of fixed or movable hoods for furnace tapping operations.
Blast furnace slag contains variable amounts of lead and zinc, and its
fate is determined by its composition. If the slag contains appreciable
amounts of zinc, which is typical of western lead ore, it is sent to a slag
fuming furnace where zinc oxide and lead oxide are separated from the gangue
material and recovered. Low-zinc slag produced at the Missouri lead
smelters is partially discarded and partially recycled to the sinter machine
for recovery of the lead value. The slag fuming furnace offgas streams
containing particulate at ASARCO/E1 Paso and ASARCO/East Helena are
controlled with fabric filters. The operational temperature of the fabric
filter systems is not known explicitly, but is assumed to be low (100 to
125°C, or 212 to 257°F) by analogy to the operation of baghouse systems on
the other lead smelter process offgas streams.
Lead bullion from the blast furnace is sent to a dressing kettle where
the molten bullion is cooled, thereby reducing the solubilities of the
various contaminants entrained and dissolved in the lead. A dross phase
forms on the surface of the lead that may contain portions of slag, matte,
speiss, and lead oxide. The dross is skimmed from the surface and sent to a
reverberatory dross furnace or an electric dross furnace, where it is
treated with additional reagents to produce as many as four phases: slag,
matte, speiss, and lead. Dross furnace slag is recycled to the blast
furnace, matte and speiss (when formed) are shipped to a copper smelter, and
lead is recycled to the dross kettle.
Dross kettle and dross furnace process offgases contain very small
amounts of SO . They are combined with blast furnace offgases at all five
domestic primary lead smelters and treated for particulate and fume removal
in fabric filter units operated over the temperature range of 100 to 125°C
(212 to 257°F). Fugitive emissions from the dressing kettles are generally
controlled by building enclosure or kettle hooding systems.
Lead bullion from the dressing operation contains a variety of
impurities that must be removed by lead refining. The three Missouri lead
smelters refine lead on-site, while the ASARCO/East Helena and ASARCO/E1
Paso smelters ship their drossed buillion to an ASARCO lead refinery in
Omaha, Nebraska. Lead refining can be a very complex operation and is
highly dependent on the level and types of impurities present. Domestic
lead refineries employ pyrometallurgical processes (furnace-kettle refining)
that are generally based on two impurity removal strategies:
(i) Drossed bullion impurities such as antimony, arsenic, and tin are
more readily oxidized than is lead. They can be removed by
oxidation/volatilization or oxidation/slagging techniques.
(ii) Drossed bullion impurities such as gold, silver, copper, and
bismuth can be removed in the form of insoluble intermetallic
compounds with the added reagents zinc, calcium, and magnesium.
56
-------
The final product of a lead refinery contains greater than 99 percent pure
lead.' Offgases from lead refining and other miscellaneous potential sources
of arsenic emissions at primary lead smelters are controlled with fabric
filter units.
4.1.3 The Behavior of Arsenic in Primary Lead Smelting
The arsenic content of refined lead represents less than 1 percent of
the arsenic fed to the smelter in the ore concentrate. The routes of
arsenic removal from the lead-bearing portion of the charge material include
volatilization and slagging as well as association with the matte and speiss
phases that are subsequently shipped to copper smelters. Arsenic is
volatilized as arsenic trioxide from sinter plants, blast furnaces, dross
reverberatory furnaces, zinc fuming furnaces, and reverberatory softening
furnaces (lead refining). Slagging of arsenic occurs in the form of
metallic arsenates such as iron arsenate [Fe., (AsO, )„] from blast furnaces,
dross furnaces, and slag fuming furnaces. Large quantities of arsenic are
removed in the matte and speiss phases that are tapped from dross furnaces.
Table 4-2 shows the approximate distribution of arsenic among the
various removal routes at a Missouri lead ore smelter. A total of
87.1 percent of the arsenic entering with the plant feed is accounted for in
the solid products leaving the facility. The dross stock pile is the
largest single source of arsenic elimination, followed by the matte stock
pile and the blast furnace slag. A total of 12.9 percent of the arsenic
entering the smelter cannot be accounted for in the solid waste and product
streams. This figure represents an upper bound on the fraction of the input
arsenic that is emitted to the atmosphere.
Comparable information on the distribution of arsenic elimination at
western ore smelters (ASARCO/East Helena and ASARCO/E1 Paso) is not
presently available. However, it is known that large amounts of arsenic are
removed from the lead-bearing material in the speiss phase tapped from the
dross furnaces at these facilities. A U. S. Bureau of Mines publication
reported that the average arsenic content of the speiss produced at three
unspecified lead smelters was 15.8 percent by weight. The data presented in
the same publication suggest that as much as 74 percent of the arsenic
entering a speiss-producing lead smelter is removed in the speiss phase.
Speiss produced at ASARCO/East Helena and ASARCO/E1 Paso lead smelters are
combined with copper ore concentrates at the ASARCO/Tacoma and ASARCO/E1
Paso copper smelters, respectively. The fate of arsenic at primary copper
smelters is discussed in Chapter 3.
4.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS
4.2.1 Regulatory Impacts
The existing regulations having a direct effect on arsenic emissions
from the lead smelting industry are the NSPS for SO and particulate, the
57
-------
TABLE 4-2. ARSENIC ELIMINATION ROUTES AT THE AMAX
PRIMARY LEAD SMELTER IN BOSS, MISSOURI
ARSENIC ELIMINATION ROUTE % ARSENIC ELIMINATION3
Dross Stock Pileb 42.4%
Matte Stock Pile 28.5%
Blast Furnace Slag 16.0%
Unaccounted Arsenic 12.9%
Refined Lead 0.3%
Speiss Stock Piled 0.0%
The percent arsenic elimination is expressed as a percentage of the total
amount of arsenic entering the plant with the ore concentrate feed.
Approximately 20 percent of the dross material produced at the AMAX smelter
is stored in the dross stockpile. The majority of the remainder is sent to
an electric dross furnace for recovery of copper matte and lead bullion.
The unaccounted arsenic was obtained by difference from material balance
figures.
Speiss does not form in the electric dross furnace at the AMAX smelter due
to the low arsenic and antimony content of the Missouri lead ore
concentrate.
58
-------
SIP emission limitations for S0? and particulate, the impending SIP emission
limitations for lead, and the OSHA workplace exposure standards for lead and
inorganic arsenic. These regulations are discussed in greater detail in
Chapter 2.
All five primary lead smelters are in compliance with their respective
SIP's for SCL and particulate emissions. The lead NAAOS is currently
exceeded in regions surrounding four of the five primary lead smelters. The
ASARCO/Glover smelter has maintained compliance with the lead NAAQS through
the useRof dispersion techniques facilitated by a tall [186 m (610 ft)]
stack.1 State implementation plans for achieving the lead NAAQS have not
yet been developed, and at the present time it is unclear what the effect of
the lead NAAQS on the primary lead smelting industry will be.
3
The OSHA lead standard of 50 yg/m is being exceeded at all five
primary lead smelters, and the OSHA arsenic standard of 10 ug/m.-is being
exceeded at two plants (ASARCO/El Paso and ASARCO/East Helena). The OSHA
workplace standard exceedances are primarily due to emissions from fugitive
sources. The ASARCO/El Paso and ASARCO/East Helena smelters have already
signed agreements with OSHA that will result in improvements in their
fugitive emission control systems. The modifications are specifically
designed to achieve compliance with the OSHA inorganic arsenic standard.
4.2.2 Arsenic Removal Capabilities of Existing Control Equipment
4.2.2.1 Process Emission Controls. Arsenic removal efficiency tests
have not been conducted on any of the process emission control systems at
existing primary lead smelters. However, the test work that has been done
in the copper smelting industry is also applicable to the lead smelting
industry. Both types of smelters are hot sources of arsenic emissions in
the form of arsenic trioxide. The acid plant/jas pre-cleaning systems in
use on sinter machine offgases are effective means of arsenic control due to
the extensive pollutant removal capability of the pre-cleaning system. The
relatively low operational temperatures (100 to 125°C, or 212 to 257°C) of
the lead smelter fabric filter systems in conjunction with the copper
smelter fabric filter test data, support the conclusion that high arsenic
removal efficiencies are being achieved on lead smelter offgas streams that
are low in SO content. Without the benefit of actual arsenic removal test
data, it is estimated by analogy to the copper smelter data that arsenic
removal efficiencies greater than 90 percent are currently being achieved by
existing acid plant and fabric filter systems.
4.2.2.2 Fugitive Emission Controls. Test data are not available to
determine the arsenic removal efficiencies of existing lead smelter fugitive
emission control systems. By analogy to the copper smelting industry, it is
expected that the best available ventilation capture systems should be
capable of approximately 90 percent fugitive arsenic emission capture. The
performance of existing ventilation systems in the primary lead smelting
industry is expected to be below this level.
59
-------
4.2.3 Inorganic Arsenic Emissions Under the Regulatory Baseline
Arsenic emissions testing has been performed at two primary lead
smelters. Process and fugitive emissions test results are available for the
ASARCO/East Helena facility, and fugitive emissions test results are
available for ASARCO/Glover. ' The testing was designed only to
determine arsenic emissions from the two plants. Arsenic removal
efficiencies of the existing control devices were not measured.
4.2.3.1 Process Emissions Testing. Table 4-3 summarizes the results
of the process arsenic emissions testing done at ASARCO/East Helena. The
sum of the measured process arsenic emissions rates from the plant was
2.2 kg/hr (4.8 Ib/hr). Tests more recently conducted at East Helena by the
Montana State Department of Health and Environmental Sciences/Air Quality
Bureau indicate that the total arsenic emission rate (process and fugitive
emissions combined) is only 0.63 kg/hr (1.4 Ib/hr) . A test report
describing this more recent work will be available by mid-1982.
4.2.3.2 Fugitive Emissions Testing. The results of the fugitive
arsenic emissions testing performed at ASARCO/East Helena and ASARCO/Glover
are summarized in Tables 4-4 and 4-5. The arsenic contents of the ore
concentrates processed at ASARCO/East Helena and ASARCO/Glover are about 0.4
and 0.02 weight percent, respectively. The relative magnitudes of the
arsenic emission estimates demonstrate that fugitive arsenic emissions from
a plant smelting high-arsenic lead ore concentrate (0.1 to 0.4 percent
arsenic by weight) are much greater than those from a plant smelting
low-arsenic lead ore concentrate (0.02 percent arsenic by weight). The
largest source of fugitive arsenic emissions at the ASARCO/Glover smelter
was found to be the blast furnace operation, which accounted for 84 percent
of the total fugitive arsenic emissions. The dross reverberatory building
was found to be the largest source of fugitive arsenic emissions at
ASARCO/East Helena, accounting for 96 percent of the measured fugitive
arsenic emissions. The disparity in the relative magnitudes of the arsenic
emissions from the various sources at the two plants is due largely to the
fact that ASARCO/Glover does not have a dross reverberatory furnace
operation. Fugitive arsenic emissions at ASARCO/East Helena represent about
22 percent of the total estimated arsenic emissions from the plant
(2.75 kg/hr). The arsenic emissions study more recently undertaken by the
State of Montana indicates that fugitive arsenic.emissions represent a much
larger fraction of the total arsenic emissions.
4.3 ARSENIC EEC OPTIONS FOR THE PRIMARY LEAD SMELTING INDUSTRY
4.3.1 Definition of Arsenic EEC
Arsenic EEC for process emissions from primary lead smelters is defined
by analogy to arsenic EEC in the primary copper smelting industry. Contact
sulfuric acid plants and the associated gas pre-cleaning systems are
recommended for the treatment of concentrated S0? offgas streams produced in
60
-------
H
W
«o> pao>
cu
CO
3
o
60
Cfl
m
61
-------
CJ
c
a
m
•*
•o
u
01
IS
3
u
CO
a.
a
o
H
CJ
C
** CJ
a CO
O i-< CO
w •- B
oo w
3
•rt CJ H
c u j:
CJ 10 — .
CO o£ -O
< C s-'
O
01 1H
> «!
•H m
•HEX
oo w ~^
3 00
b .*
T3
C8
i— ( tl CO
ra _: e
u O
C CJ -H
H > w
it W
U-t JJ *"<
O -H S
oo (3
N L?
^
•0 01 .C
CO U —
01 CO J3
ti e
U CO
•H CO Ul
00 -rt .C
bLfSo
4J
CO
3
1-1 o en
8 -H C
O V -H
H a a
0* 03
O CJ E
> W
M n
00
3
V
18 ^
3 'tl JS
CJ W *«~
•H a .a
tl CO
> 01 1-
-H *H JZ
u 3 ~-
•H U 00
01
C
0
1-4
ffi
o
o.
o
^O ^ *** •*** •**
— CM
•-4 r^. -^ O O
O in O O O
O O O O C3
r^ O ^O CO c^
l^ CM *» — < en
oo
>-* y—» ^-^. so m
en CO cs O —
O CM O O O
CM so I*- m \o
— CM O O O
o -• o o o
l/^ *— < 00 i— • U"l
««• i^- r« en r*-
*4 ^O
r^. m u"i ^fl «^
• • • • •
CM CM -4 O —
~
CM \O SO CM SO
-* tri O O O
00
c
•H 01
•O J3 o tl
-HI. C 00 U
•H cj co c co
3 > W -H C
a oj oo 3 e ^ >•<
U C b 3 u s
i-. — t-i [» 1-1 b,
4J CO f-i CQ O -H CJ
e o 1-1 co co c
•H U 3 *H i-t ra •**
en a pa pa Mb M
O
SO
O
O
o
o
vC
CO
ki i-l
O 6.
01
•M kri
B -rt
O C
H 6J
62
-------
w
M
EH
M
O
e
u
M
25
W
C/IS
<3 ccs
W
Q H
3 rJ
CO
Oi-l-4
00
3
CJ ,-v
•H Q) V4
C L' r*
41 CO
01 02 ^
l-l rH
< c •*-*
0
41 -H
•H (fl
00 b3 ^**
3 00
[u ^
•a
cd
•o
n
rH 4) 01
co _; c
u o
O 41 iH
•H 0)
l4-» 4J -r-t
0 -H 6
00 W
X 3
E*«
^^
•o ci .e
CO U —
4) IS -O
rJ « rH
4) C
? 0
•H -H
U 01
•H 01 W
oo i-i .e
3 3 ~-
fc. U 00
IV
4J
CO
rH
3
U
4-1
10
CM
rH
CO
JJ
O
H
4J
c9
rH 3
a u 01
0 u O
t- U -H
CO 01
IM CU 01
0 -H
41 B
S-8 > U
•H
00
3
41
CO /-.
3 41 J=
•H CO .£
4J C(£ rH
B C
0- O
4) X
> 01
•H -H U
*J S JH
•H W —
oo oo
3 ^
01
c
0
•H
CO
b4
41
a.
O
4)
ja
-H 0
oo
>O -T -H O
rH CO -H O
00 -H
41
Z
^
CN1 x^
O — «
O O 41
• • rH
O O JD
00
•H
oo
CO rH 41
O ~» Z "1
0 C C
O O 0
O C5 O
co . r^. in o
co r- en O
en in o
rH
^
^*v s~* in ^~*
CO vC rH in
^4 CM O
-------
the lead sintering operation. The S0? concentrations of the remaining
offgas streams at primary lead smelters are too low to make contact sulfuric
acid plants a viable EEC option for these streams. Fabric filter systems
operated 10° to 25°C above the acid dewpoint of the gas stream are
recommended as arsenic EBC for offgases generated by blast furnace, dross
furnace, zinc fuming furnace, and lead refining processes. Information
concerning the acid dewpoints of these streams is unavailable, but current
practice in the primary lead smelting industry supports the feasibility of
operating the fabric filter unit over the temperature range 100 to 125°C
(212 to 260°F) with no deleterious effects due to corrosion.
Arsenic EBC for fugitive emissions in the primary lead smelting
industry consists of enclosing of ore storage areas, ventilating and/or
enclosing material transfer points, ventilating and/or enclosing furnace
operations, and treating all of the ventilation gas streams with fabric
filter systems. The specific techniques to be applied will vary on a
plant-by-plant basis due to the diversity within the lead smelting industry.
4.3.2 Prevalence of Arsenic EBC Under the Baseline Level of Control
Arsenic EBC is in place for process emission sources at each of the
five domestic primary lead smelters. The application of EBC for process
sources has been motivated by existing S0» and particulate SIP emission
limitations. Arsenic EBC for fugitive emissions are projected to be in
place when compliance with the OSHA arsenic and lead standards and the
impending SIP emission limitations for lead are achieved. Thus, there are
no incremental environmental, energy, or economic impacts associated with
the requirement of arsenic EBC for the primary lead smelting industry.
4.4 POPULATION EXPOSURE DATA
The data in Table 4-6 were developed during this study for use in an
air pollution dispersion model to estimate the population exposure to
arsenic from primary lead smelters. The Strategies and Air Standards
Division (SASD) of EPA is using these data in an ongoing effort to estimate
exposure to arsenic from this source category.
4.5 EASE OF STANDARDS DEVELOPMENT
Process arsenic emission sources at the five domestic primary lead
smelters are currently controlled at the recommended EBC level. As
discussed in Section 4.3, existing regulations for the control of SO and
particulate have resulted in lead smelters installing control systems that
have the secondary benefit of effective arsenic removal.
The majority of fugitive arsenic emission sources at the primary lead
smelters are not currently controlled by EBC. However, industry compliance
with the OSHA lead and inorganic arsenic standards and the lead SIP's for
the lead NAAQS attainment is projected to result in EBC for fugitive sources
64
-------
<£
s3
M
p*i
CM
W
H
O
fa
E"^
O
1 >
CM
2
C ^
12
,J
1 1
U
Q
O ><
':> ' p^
t-i
W OO
OS 3
i— J £•*
O I—*
CM
X O
w z
H
2: H
O .J
M W
H IS
U
o v.
t- g
9
0) u-
u cj
CO CO
a o
o cy
•-I cfi
Eb -*.
CO
C£ ^
U
w a
<3 O Ei-
pS fr- o
* •— I P— • CM
^ lA
*-\ ^-s O CM
in as 2 H
N-- S«/ *— ' %»--
f*» ^x O "^
— CM CO CO
P"* "^ *ff *ff
"^^"!*v%r^.
o o m o
O ^ N-* ^^
O O >>O O
SO CM CM CM
-^ —i —
^ ^ ^ a
y y U ai
(C 03 cC JS >
W C/l W CD *J
o- u. > v: QO
P i* *-^ e
ot ai AJ 3 i-i
« W W &L. (t)
•H -H <-< C= O
to en CQ i*j H
a g
lA 00
CM OS OO
•— CM i/>
m
01 4-»
C V- P
O O U
t"J H >
«
C
0)
rH
0)
8 =
< » H
< wz
^^
c ^r -a- o
•H O O —
O.CM CM VO
o o' o' o*
U
^
gyp ^ M9
••H -* •* O
•H • « •
ooo o —
^ 4
00 CM CM M?
Ot > • '
zoo o
m m
*"%s x^_|
CM OO *-*.
• » 00
so in •
CM m o^
OOO
OC r«« CO
m
m m —
ooo
OOO
O O -
. . o
sO O CM
CO CM --
•+S in ""-^
CM \O
•— CM m
in tn in
S~* 1*+. y-N
r- CM in
-3- M3 CO
O ^- O
r*« in O
CO "^ <*1
(*i ^ co
in in
in — » —
>-V MM -^
OS . .
oo — —
r^« d os
CM -^ *ff
lA lA lA
— So*
co — O
—• ' \D CO
'*•' ^"^ ^^
O sO -H
Ji ^ . en
W 0 C 0>
flj TO 1_ >
OH > 00 OC
< 0) C 3
-^ OC -H (JU
p -^ a M
Ol *J 3 U r-4
4-* « U- <0 tO
•H r-« C U3 O
C/l M N3 H
CO lA
§ -*
C^l l/l
f^
O U
C f V-
O O 1U
N = >
X
t-
o
C tn
OS B.
VJ l-H
< U
2 •%.
•-T •
0
CM •— '
»A
CM r^
~ °
sO M3
f— V*
CM -^
r^. cM
o'
^ 3
CO
CO —4
o
0 o-
lA
^^
CO
^
^
IA
CO
lA
~*^
C
in
CO
00
Os
CO
m
fin.
^
CO
m
CO
m
y~v
o
o*
s*-
^_!
vO
m
1-4
CM
in
CO
N-y
CO
o
en
o>
en to u
CO i-t i-l
-< w oc
60 3
^^ ^) U,
01 0> *-*
*J > CO
•H Q£ O
JJ
o o
LA ^ CO
— < e^j f*t
r~ f-i
01 4J
C U ^
o _o a)
E
at
c
ai a
o .-•
"J 3
U
. P O
~b ~c
CM O
— O
in tn
*: §
0 =>•
>-s CM
-*»•
O O O-*
m m o o
^S ISiS^
r*. oO co co
CO •& 00 "^
M-I -- — t CM
'^r^-Ty-Tt-s
0000
§000
ooo
o o m in
CM •*-- *— • ^1
O co m — *
-H CO CO CO
f-
m PS.
-* in in ~*
— O O -*
CM p» r-- — *
in *^" "^ co
CO CM CM rO
r^.
/•sji* «• ••
"^ oSo
c • • •
— •« m ir\ ^%
^^^^
CM lA lA -M
fO — « — • CM
m r*. r-
"^-7^, -Tt-Cj
O O O --N
— • • • m
•^^ r*. p- co
Os
* co ao
m • • •y
oo co co ~*
^- CM CM —H
CO
^ V
O M >
*j a; . u
j<: w u c i-t
u w t- oo
n o* 3 3
jj en -H u- u-
O 3 W CJ r-4
re co oi i-t en o
X CO > « H
00 CM
1/S »O -^
""* r- O
01 M
C P )->
O 0 01
N EEi >
0
U cu
a: >
< o
< S
S ^
m tn co
. ^ «^
•^ B) O
CM g
r* O co
0 ° 0
^ -00
«- ey
. Opfl
3^S
00 .
CM Z '*-'
.
O
o
lA lA O
— ~4
00 M^"
CM CO
f> ^3-
%_y ^
O co
lA lA
O C
c o
o o
lA CM
fO CM
S-^ Sw'
\o o
iA
lA •—
O •
00 0
—" 00
in o
in o
CO CO
in
— « lA
,—* i
• CO
lA •
— * ro
\o o
<• —
in m
— i --
o o
0 0
CM — '
•*-'
u
U 00
CO ^ 3
4J O U-
4_t i— 1
C W 03
re Cu o
S < E-1
o
>< tn
< to
z o
< ca
65
-------
also being in place. The data are not available to accurately estimate when
all fugitive sources will be at EEC.
Thus, compliance with existing regulations will result in the
application of arsenic EEC for both process and fugitive sources in the
primary lead smelting industry. Arsenic emissions standards development
would have a negligible inpact on the industry.
66
-------
4.6 REFERENCES
1. U. S. Bureau of Mines. Lead. In: Mineral Facts and Problems.
Preprint from Bulletin 671. 1980.
2. U. S. Environmental Protection Agency. Industrial Process Profiles for
Environmental Use: the Primary Lead Industry. Research Triangle Park,
N.C. EPA Contract No. 68-03-2577. February, 1980.
3. Paul, Robert B. The Glover Lead Smelter and Refinery of the American
Smelting and Refining Company/Glover, Missouri. In: AIME World
Symposium on the Mining and Metallugy of Lead and Zinc. Proceedings.
St. Louis, Missouri. 1970. p. 777-789.
4. Gibson, F. W. The Buick Smelter of AMAX - Homestake Lead Tollers. In:
AIME World Symposium on the Mining and Metallurgy of Lead and Zinc.
Proceedings. St. Louis, Missouri. 1970. p. 738-776.
5. Beilstein, Donald H. The Herculaneum Lead Smelter of St. Joe Minerals
Corporation. In: AIME World Symposium on the Mining and Metallurgy of
Lead and Zinc. Proceedings. St. Louis, Missouri. 1970. p. 702-736.
6. Kelly, W. R. ASARCO in El Paso. Engineering and Mining Journal.
_182_: 79-98. September 1981.
7. Letter from Kearney, W. M., AMAX Lead Company of Missouri, to Lee Beck,
EPA:ISB. August 6, 1981. Arsenic material balance at the AMAX
Smelter.
8. U. S. Bureau of Mines. Occurrence and Recovery of Certain Minor Metals
in the Processing of Lead and Zinc. U. S. Government Printing Office,
Washington, D. C., Information Circular 8790. 1979.
9. Memo from Rathbun, R., EPA, to Pratapas, J., EPA. December 21, 1981.
16 p. Copper, lead, and zinc smelter information.
10. Telecon. Cassidy, M., Occupational Safety and Health Administration,
with Keller, L., Radian Corporation. March 18, 1982. Occupational
Safety and Health Administration arsenic project.
11. U. S. Environmental Protection Agency. Emissions Measurement at the
ASARCO Lead Smelter in East Helena, Montana. Research Triangle Park,
N.C. EPA Contract No. 68-01-4140. May 1980.
12. U. S. Environmental Protection Agency. Sample Fugitive Lead Emissions
from Two Primary Lead Smelters. Research Triangle Park, N.C.
Publication No. EPA-450/3-77-031. October 1977.
67
-------
13. Telecon. Maughn, D., Montana Air Quality Bureau, with Keller, L.,
Radian Corporation. January 12, 1982. Emissions measurement at
ASARCO/East Helena.
14. Telecon. Kelps, H., Montana Air Quality Bureau, with Keller, L.,
Radian Corporation. January 13, 1982. Stack parameters at ASARCO/East
Helena.
15. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for Emission Standards and Engineering Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency.) EPA Contract No. 68-02-3173. May 1982.
16. Memo. Keller, L., Radian Corporation, to arsenic file. April 8, 1982.
Emissions estimates for the primary lead smelters.
17. Letter from Richardson, J. B., ASARCO, to Fensterheim, R. J., Chemical
Manufacturers Association. May 17, 1982. ASARCO response to Radian
emission estimates.
18. Telecon. Reynolds, B., Missouri DNR, with Keller, L., Radian
Corporation. March 3, 1982.
68
-------
5.0 PRIMARY ZINC SMELTING
This chapter discusses the primary zinc smelting industry and its
associated arsenic emissions. Section 5.1 presents a brief source category
description. Section 5.2 discusses the regulations currently covering the
industry and the resulting level of control and associated arsenic
emissions. Section 5.3 presents a discussion of the EEC options and
Section 5.4 discusses the impacts of applying these options. Section 5.5
presents the input data to be used in assessing the population exposure to
arsenic emissions from primary zinc metal smelting and zinc oxide
production. The ease of standards development is discussed in Section 5.6.
5.1 PRIMARY ZINC SMELTING DESCRIPTION
Metallic zinc is produced from zinc ore concentrates at primary zinc
smelters. The estimated domestic production of primary slab zinc in 1978
was 407,000 Mg (448,000 tons), which accounted for approximately 49 percent
of the total U. S. demand for primary zinc metal. There are currently five
plants that produce primary zinc in the United States. In addition, two
zinc oxide plants have been considered in this study because they produce
zinc oxide directly from zinc ore concentrates and thus are sources of
inorganic arsenic emissions. Table 5-1 lists the names, locations,
production capacities, and estimated inorganic arsenic emission rates of the
plants. The total annual inorganic arsenic emission rate from the five
primary zinc metal smelters is estimated to be about 0.3 Mg (0.3 tons)
arsenic/year. Annual arsenic emissions from the two zinc oxide plants are
estimated to be 5.2 Mg (5.7 tons)/yr.
There are two major types of primary zinc metal production facilities,
pyrometallurgical smelters and electrolytic plants. The only domestic
pyrometallurgical primary zinc metal smelter in operation is the St. Joe
Minerals electrothermal smelter in Monaca, Pennsylvania. During the last
few decades, electrolytic zinc plants have achieved predominance over
pyrometallurgical zinc smelters in the domestic primary zinc industry. Four
of the five U. S. primary zinc metal plants are electrolytic facilities.
The materials processed by the domestic primary zinc industry are
predominantly sulfide ore concentrates. The most common zinc-bearing
constituent of these ores is sphalerite (ZnS). Raw zinc ores are treated at
the mine site with flotation techniques to produce a concentrate containing
up to 62 percent zinc and 32 percent sulfur. Other zinc ore constituents
include iron, calcium, silicon, lead, magnesium, cadmium, copper, aluminum,
cobalt, and arsenic. The arsenic content of the ore concentrates processed
in the domestic zinc smelting industry range from about 0.001 to
0.100 percent by weight. A U. S. Bureau of Mines Publication lists the
69
-------
TABLE 5-1. THE PRIMARY ZINC SMELTING INDUSTRY
PLANT /LOCATION
St. Joe/
Monaca, PA
ASARCO/
Corpus Christ! , TX
AMAX/
Sauget, IL
Jersey Miniere Zinc/
Clarksville, TN
National Zinc/
Bartlesville, OK
ASARCO/
Columbus, OH
New Jersey Zinc/
Palmerton, PA
PRODUCTION CAPACITY
PLANT TYPE Mg/yr (tons/yr)
Electrothermal 50,000 (55,000)
Zinc /Zinc Oxide
Smelter
Electrolytic 98,000 (110,000)
Zinc Plant
Electrolytic Zinc 76,000 (84,000)
Plant
Electrolytic Zinc 82,000 (90,000)
Plant
Electrolytic Zinc 51,000 (56,000)
Plant
American Process 20,000b (22,000)
Zinc Oxide Smelter
American Process 30,000 (33,000)
Zinc Oxide Smelter
ESTIMATED
BASELINE INORGANIC
ARSENIC EMISSIONS3
kg/hr (Ib/hr)
0.023 (0.051)
0.0014 (0.0031)
0.0028 (0.0062)
0.0030 (0.0067)
0.0014 (0.0031)
0.014 (0.031)
0.60 (1.3)
The arsenic emissions estimates shown represent a "best estimate" of baseline
emissions. See Appendix A for development of estimates.
Production capacity of zinc oxide (ZnO).
70
-------
average arsenic content of zinc ore concentrates processed in the
electrolytic portion of the industry as 0.07 percent by weight. Table 5-2
includes the arsenic content of the zinc ore concentrates processed at
several domestic facilities.
The first step in the production of zinc metal from ore concentrates at
both electrolytic and pyrometallurgical zinc smelters is the roasting
operation. Zinc roasting consists of heating the ore concentrates to
650-1000°C (1200-1800°F) in an oxidizing environment. The roast is carried
out below the melting temperature of the charge and has three primary
functions: (i) elimination of sulfur as SCL, (ii) conversion of zinc
sulfide to impure zinc oxide, and (iii) removal of volatile impurities from
the ore concentrate. The degree of sulfur elimination accomplished in a
zinc roaster varies from about 93 to 97 percent.
The fluid bed roaster is the newest roasting system for zinc sulfide
concentrates and is currently the only type in use. The offgas stream from
a zinc roaster typically has an SCL concentration of 10 to 13 percent, and
up to 85 percent of the roaster product (calcine) is carried out with the
offgas. Waste heat boilers, cyclones, and electrostatic precipitators are
used in series to capture the entrained portion of the calcine. The
collected materials are normally combined with the remaining portion of the
calcine and sent to either a sinter machine (pyrometallurgical smelting
facility) or a leaching plant (electrolytic facility). Contact sulfuric
acid plants are used for SO control at all of the fluid bed roasters within
the primary zinc industry. Figure 5-1 illustrates the fluid bed roasting
system at the Jersey Miniere Zinc electrolytic zinc plant in Clarksville,
Tennessee.
5.1.1 Electrothermal Zinc Smelting
The roasting operation and the remaining steps in the production of
zinc metal at the St. Joe/Monaca electrothermal zinc smelter are shown in
Figure 5-2. The roaster calcine is first mixed and pelletized with sand,
return sinter fines, coke breeze, and smelting furnace residue. The
pelletized mixture is distributed on grate-type pallets that move along a
continuous metal conveyor system within the sinter machine. The purpose of
the sintering operation is to remove residual sulfur and volatile impurities
such as lead, cadmium, and arsenic from the calcine, and to agglomerate the
feed for reduction in the electrothermal smelting furnace. Sinter offgases
typically contain about 0.1 percent SO . Low-temperature baghouse units
operated at about 93°C (200°F) are used for particulate control.
The sinter product is sent to the electrothermal furnace, where the
zinc content of the zinc oxide/coke/silica mixture is separated and reduced
to pure zinc metal. Heat is produced within the furnace by the resistance
of the charge material to the current flow between pairs of graphite
electrodes. The zinc metal produced is vaporized in the high temperature
atmosphere of the furnace (1400°C or 2600°F). Furnace vapors are drawn
71
-------
TABLE 5-2. THE ARSENIC CONTENT OF THE ZINC ORE CONCENTRATES
AT SEVERAL DOMESTIC PRIMARY ZINC SMELTERS
PLANT ARSENIC CONTENT OF ORE
CONCENTRATE (WEIGHT PERCENT)
National Zinc/Bartlesville, Oklahoma < 0.1
St. Joe/Monaca, Pennsylvania 0.005
Jersey Miniere Zinc/Clarksville, Tennessee0 0.0018
New Jersey Zinc/Palmerton, Pennsylvania ' 0.08-0.13
ASARCO/Corpus Christi, Texasf 0.001
Telecon. L. R. Meadows, National Zinc, with Larry Keller, Radian
Corporation. January 22, 1982.
Telecon. Thomas Janeck, St. Joe Minerals Corporation, with Larry Keller,
Radian Corporation. February 11, 1982.
CPainter, L. A., "Jersey Miniere Zinc: Plant Design and Start-Up."
Engineering and Mining Journal, July 1980.
Telecon. Maurice Silvestris, New Jersey Zinc, with Larry Keller, Radian
Corporation. January 22, 1982.
The ore smelted at the New Jersey Zinc/Palmerton facility is unique in
that it is a low-sulfur ore (Ogdensburg ore).
Telecon. Ken Nelson, ASARCO, with Larry Keller, Radian Corporation.
March 2, 1982.
72
-------
o
2 W
< 2
0)
0)
m co
i-< w
01 0)
•H C
c c
•H 0)
S H
to
OJ Ai
•u o
to
e -H
01
4-J 4-J
W C
f^ td
W rH
a
oo
c a
•H c
•u -H
05 N
03
o a
>-i -H
0) rH
XI O
M
73 4J
•H O
3 OJ
^ ^H
fa W
< J
O <
Cb U
b
O Q
W
OS Z
a 1-1
H <
w os
<: H
o z
M
O
o w
M Q H
3 Id CO
j ca <
u, o
00
FT
I oc I
o
H Z
OS O
O O
OS
M
<
73
-------
ELECTROTHERMAL SMELTING FURNACE OFFGAS
,TER SYSTEM
H ""
fa
0
M
at
1
g
> i
ZINC METAL
TO
DISTILLATION
ELECTROTHERMAL
SMELTING
FURNACE
NTERED
I
Cd
H
3
PROCESS
o
>•
cj
Cd
OS
t
iLAG TO RESIDUE TREATMENT
!
CO
CO td
< E-
O W
fa ^
fa CO
o
OS
Cd Cd
Z H
M -4 _
ss -*
So
gl
M
CO g
Cd
Ctf Z
Cd F-I
H a
5§
co £
Id
Z
s
C/3 Qtf
O U
gl
Q3
O Cd
M Q H
S Cd CO
-3 aa <
fa O
OS
O Q
S U
OS <
td
Cd S
BS H
O Z
Id
O O
Z Z
58
CU
C8
•H
C
I
-------
through a water cooled, U-shaped condenser and bubbled through a molten zinc
bath. The zinc metal vapor condenses and is subsequently cast into slabs.
Offgases from the electrothermal furnace are treated for particulate removal
by baghouse units operated at about 93°C (200°F). Fugitive emission
controls at the St. Joe/Monaca facility consist of local ventilation of all
materials handling systems followed by particulate collection in fabric
filter units. Table 5-3 contains a summary of the existing air pollution
control equipment at St. Joe/Monaca.
Arsenic elimination can be achieved by several mechanisms in the
electrothermal smelting of zinc. Arsenic can be volatilized as arsenic
trioxide (As?0») from the roaster, the sinter machine, or the electrothermal
furnace, and it can be slagged from the electrothermal furnace. Data
presented in a U. S. Bureau of Mines publication suggests that the largest
single source of arsenic elimination from pyrometallurgical smelters is the
furnace slag.
5.1.2 Electrolytic Zinc Production
The roasting operation and the remaining steps in the production of
zinc metal at a typical electrolytic zinc smelter are shown in Figure 5-3,
The roaster calcine is first leached in a dilute sulfuric acid solution to
dissolve the impure zinc oxide. Manganese dioxide (MnO ) is generally added
to the leach tank to cause the precipitation of an iron cake that contains
iron and significant amounts of arsenic, antimony, and silicic acid. The
leachate is then sent to a series of cold and hot purification tanks where
cadmium, copper, and cobalt are removed from solution. The precipitation
reactions,, that occur are induced by the addition of zinc dust, which reduces
Cd , Cu , and Co to their respective metallic forms. At least one
domestic primary zinc smelter utilizes arsenic trioxide as an activating
reagent in the removal of cobalt from the zinc-bearing solution.
The final step in the electrolytic production of zinc metal is
electrodeposition onto the surface of the cell house electrodes. An
electric potential is applied to the solution, causing the formation of zinc
metal and sulfuric acid. The sulfuric acid is recycled to the leach tank,
and the zinc metal is stripped from the electrodes and cast into slabs.
The fluid bed roaster SO emissions at each of the four domestic
electrolytic zinc smelters are controlled by a contact sulfuric acid plant.
The remaining steps in the electrolytic production of zinc do not generate
significant amounts of atmospheric emissions. Demisters are used to control
acid mist from the cell house operations.
Data on the distribution of arsenic at electrolytic zinc smelters are
presented in reference 2. Although the information given is incomplete, the
largest source of arsenic elimination is shown to be the leach residues.
Most of the arsenic entering an electrolytic zinc plant is removed in the
solid or liquid phase with the iron and copper cakes.
75
-------
TABLE 5-3. EXISTING AIR POLLUTION CONTROL EQUIPMENT AT THE
ST. JOE/MONACA ELECTROTHERMAL ZINC SMELTER
EMISSION SOURCE
EXISTING AIR POLLUTION
CONTROL EQUIPMENT
Ore Concentrate Dryer
Venturi Scrubber
Fluid Bed Roaster
Cyclone/ESP/Contact Sulfuric Acid Plant
Sinter Machine
Fabric Filter (93°C) (200°F)
Sinter Sizing & Crushing
Fabric Filter (93°C) (200°F)
Electrothermal Furnace
Fabric Filter (93°C) (200°F)
76
-------
ZINC
ORE *•
CONCENTRATE
AIR *•
PRE-CLEANED ROASTER OFFGAS
TO ACID PLANT CIRCUIl'
MANGANESE
DIOXIDE
LEACH
TANK
SULFURIC
ACID
ZINC
METAL
ZINC
DUST
IRON
CAKE
COLD
PURIFICATION
REACTOR
CADMIUM,
COPPER CAKE
ELECTROLYTIC
CELL HOUSE
HOT
PURIFICATION
REACTOR
•COBALT
CAKE
ZINC ARSENIC TRIOXIDE
DUST (ACTIVATING AGENT)
Figure 5-3. A typical electrolytic zinc smelter.
77
-------
The only potential source of airborne arsenic emissions from
electrolytic zinc smelting is the roasting operation. The available data on
arsenic elimination during the roasting of zinc concentrates are not
conclusive. By analogy to copper roasting in fluid bed roasters, it is
reasonable to expect that at least a small fraction of the arsenic entering
the roaster with the ore concentrate is volatilized as As_0 . However,
since contact sulfuric acid plants are used for roaster offgas SCL control
at all of the domestic electrolytic zinc plants, arsenic emissions from
these facilities are low.
5.1.3 Zinc Oxide Production
Zinc oxide is produced by two methods in the domestic zinc smelting
industry, the French and American Processes. In the French Process, zinc
oxide is made from the oxidation of purified zinc metal that is volatilized
and oxidized to zinc oxide. Arsenic emissions from this type of zinc oxide
facility are assumed to be small due to the purity of the feed material
(zinc metal).
American Process oxide is produced directly from zinc ore concentrates,
and thus has a larger potential for arsenic emissions. The two domestic
American Process zinc oxide smelters (ASARCO/Columbus and New Jersey
Zinc/Palmerton) utilize different types of ore concentrates.
ASARCO/Columbus processes a zinc sulfide ore concentrate, and the operation
consists of roasting in a fluid bed roaster followed by processing in a
densifying kiln and a wetherill zinc oxide furnace. A contact sulfuric acid
plant is used to treat the roaster offgas, and low temperature baghouse
units are used to treat the other offgas streams. New Jersey Zinc/Palmerton
processes a low-sulfur zinc ore concentrate (Ogdensburg ore), and the major
process steps include waelz kiln operations, sintering, and horizontal grate
furnace operations. Particulate removal at the New Jersey zinc plant is
achieved using baghouse units on each of the process offgas streams. The
waelz kiln product baghouse is operated at 140°C (280°F) and the sinter
machine, horizontal grate furnace, and the fume kiln baghouses are operated
below 100°C (212°F). A schematic diagram of the New Jersey Zinc/Palmerton
zinc oxide plant is shown in Figure 5-4.
No data are available on the behavior of arsenic in primary zinc oxide
production. The assumptions used to estimate arsenic emissions from these
facilities are listed in Appendix A-5.
5.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS
5.2.1 Regulatory Impacts
The existing regulations affecting the primary zinc smelting industry
are the SO and particulate matter NSPS, the SO and particulate matter SIP
emission limitations, and the OSHA inorganic arsenic workplace standard.
All of the primary zinc smelters are in compliance with their respective
78
-------
u
C
cu cu
2 E
3 w
&U CQ
cu
O M
E- H
^ k
CU
fa
en
cu c 6 ^
3 3 "& ~~
fa _4 td
CO
-H cu
S' >,m a
T3 cd
O 00
E
r*
cu
T3
•H
u X
O O
3
•o u
O C
W *H
/ i
H
•H
X
0)
6
fa
/
0) f~ u-j
E OOu-j
r3 eg o
.u
en
3
C
CO
01 eg
C 00
~ *M
y o
jd
en
u 3
o o
J-* -C
k
•u en
a cd
•3 on
13 u-i
CU O !— I CU
C MO cu u co
1-f
cu a cj 3
CU - M 3 0
0) en ^ (3 T) J3
CJ 3 M O 00
td O 3- M cd
c _; fa t- sa
)_, QO 1 i 1
3 eg A
fa 03
0)
cj en
ed cd
C 00
~ OJ
3 'i i
a, b °
c
o
eg co
£ cd
00
M U-4
CU 01
M C
: co
01 -H 3
u J= 0
c cj j:
•H Cd 00
co S eu
•H eg
C/3 CQ
pa •
•^ w
C
•H
w
CU
c
1-4
_c
U
5
^
01
J-i
J2
^
0)
jj
C
•rt
C/5
N
i-H
0)
cd
cu
.u
cd
cj cu
ej
tH Cd
cd C
C 3
O fa
N
•H
iJ
o
ex
(U
•H
O
O
c
o
01
iH
CD
PL4
O
c
OJ
en
-------
SIP's for S02 and particulate emissions. In addition, all of the primary
zinc plants are in compliance with the OSHA inorganic arsenic workplace
standard.
5.2.2 Effectiveness of Existing Controls for Arsenic Removal
Arsenic removal efficiency tests have not been conducted on any of the
process emission control systems at existing primary zinc smelters.
However, the test work that has been done in the copper smelting industry is
also applicable to the zinc smelting industry. Both types of smelters are
hot sources of arsenic emissions in the form of arsenic trioxide, and
similar control technologies are applied. The acid plant/gas pre-cleaning
systems in use on roaster offgases have been shown to be the most effective
method of arsenic emission control. Fabric filter systems have been shown
to be capable of greater than 90 percent arsenic removal provided the
operational temperature is maintained at or below about 110°C (230°F).
5.2.3 Magnitude of Arsenic Emissions from the Primary Zinc Smelting
Industry
Arsenic emissions testing has been performed at only one facility in
the zinc smelting industry. Test results from the St. Joe/Monaca
electrothermal smelter indicate that the largest source of controlled
arsenic emissions from that plant is the sinter machine offgas stream. The
estimated arsenic emission rate from the sinter machine based on the test
results is 0.023 kg/hr (0.05 Ib/hr). Arsenic emissions from other process
and fugitive sources at the St. Joe/Monaca facility are reported to be
negligible in comparison to sinter machine emissions.
There has been no test work done on arsenic emissions from electrolytic
zinc smelters. The fluid bed roaster/contact sulfuric acid plant tailgas is
the only source of appreciable arsenic emissions from these facilities. An
upper bound on the arsenic emissions from a contact sulfuric plant can be
calculated from the maximum concentration of arsenic that can be tolerated
as an impurity to the catalyst bed in the SO. ->• SCL conversion process,
which is 1.2 mg/Nm . Measured arsenic concentrations in the offgases from
contact sulfuric acid plants tested in the primary copper smelting industry
were consistently below this level regardless of the inlet arsenic content
of the gas stream entering the pre-cleaning system. The arsenic emissions
estimates in Table 5-4 were calculated using this maximum arsenic
concentration and the measured acid plant tailgas flowrates in the primary
zinc industry. The upper bound arsenic emissions estimates for the acid
plant units range from 0.022 to 0.047 kg/hr (0.010 to 0.048 Ib/hr).
Arsenic emissions from the New Jersey Zinc/Palmerton, Pennsylvania zinc
oxide plant were estimated using the procedure outlined in Appendix A-5.
Based on a projection of the arsenic distribution among the various slag and
vapor streams at the plant, arsenic emissions at New Jersey zinc were
estimated to be about 0.60 kg/hr (1.33 Ib/hr) under the baseline level of
80
-------
TABLE 5-4. UPPER-BOUND ESTIMATES OF INORGANIC ARSENIC EMISSIONS
FROM CONTACT SULFURIC ACID PLANTS IN THE PRIMARY
ZINC SMELTING INDUSTRY
PLANT/LOCATION
MAXIMUM INORGANIC ARSENIC
EMISSION RATE FROM CONTACT
SULFURIC ACID PLANT3
kg/hr
(Ib/hr)
ASARCO/Corpus Christ!, TX
0.023
(0.051)
AMAX/Sauget, IL
0.045
(0.099)
Jersey Miniere Zinc/Clarksville, TN
0.047
(0.10)
National Zinc/Bartlesville, OK
0.023
(0.051)
ASARCO/Columbus, OH
0.022
(0.048)
Estimates made based on maximum permissible arsenic impurity limits for
acid plant feed gas (1.2 mg/Nm ).
81
-------
control. However, no arsenic emission test data are available for
comparison purposes to verify the estimate.
5.3 ARSENIC EBC OPTIONS FOR THE PRIMARY ZINC INDUSTRY
5.3.1 Definition of Arsenic EBC
Arsenic EBC for primary zinc plants is defined by analogy to arsenic
EBC in the primary copper smelting industry. Contact sulfuric acid plants
and the associated gas pre-cleaning systems are recommended for the
treatment of offgas streams containing sufficient S09 such as those produced
by fluid bed roasters. The S0« levels of the remaining offgas streams at
primary zinc and primary zinc oxide smelters are too low to make contact
sulfuric acid plants a viable EBC option for these streams. Fabric filter
systems operated 10° to 25°C above the best estimate of the acid dew point
of the gas stream are recommended as arsenic EBC for offgas streams low in
SO,- content. Operations controlled in this manner include the sinter
machine and the electrothermal furnace offgases at St. Joe/Monaca, and the
Waelz kiln, sinter machine, horizontal moving grate furnace, and fume kiln
offgases at New Jersey Zinc/Palmerton. An operational temperature below
110°C (230°F) is suggested by analogy to the available data in the primary
copper smelting industry.
Arsenic fugitive EBC in the primary zinc smelting industry consists of
the best available ventilation technology for each fugitive arsenic emission
source followed by particulate removal in a fabric filter system operated
below 110°C (230°F). This includes enclosure of ore storage areas,
ventilating and/or enclosing material transfer points, properly ventilating
and/or enclosing all furnace operations, and treating all of the ventilation
gas streams with fabric filter systems. The specific techniques
constituting arsenic fugitive EBC will vary from plant to plant due to the
diversity of plant types within the primary zinc smelting industry.
5.3.2 Prevalence of Arsenic EBC
Arsenic EBC is in place for all arsenic emission sources at the
electrothermal and electrolytic primary zinc metal plants and at the
ASARCO/Columbus zinc oxide plant. In addition, all arsenic emission sources
at the New Jersey Zinc/Palmerton zinc oxide smelter are controlled at the
EBC level except for the waelz kiln offgas stream. The fabric filter system
serving this stream is operated at 140°C (280°F), a relatively high
temperature for effective arsenic capture. Arsenic EBC for this particular
emission source consists of lowering the operational temperature of the
fabric filter unit closer to the acid dew point of the gas stream. The acid
dew point of the stream is likely to be low (below 100°C or 212°F) because
the sulfur content of the ore concentrate smelted at New Jersey Zinc is very
low and hence there will be little SO- in the offgas. The cooling capacity
required to lower the temperature of the waelz kiln offgas stream could
easily be provided by a water spray chamber system.
82
-------
5.4 INCREMENTAL IMPACTS ASSOCIATED WITH THE ARSENIC EEC OPTIONS
The incremental impacts of the arsenic EEC options have been assessed
for the only primary zinc smelter for which arsenic EEC is judged to be not
fully in place (New Jersey Zinc/Palmerton zinc oxide plant). The
environmental, energy, and economic impacts resulting from the EEC option
are discussed in the following sections.
5.4.1 Environmental Impacts
The greatest environmental impact associated with the implementation of
arsenic EEC in the primary zinc smelting industry would be the reduction of
atmospheric arsenic emissions. It is estimated that arsenic emissions at
the New Jersey Zinc facility will be reduced by 0.50 kg/hr (1.1 Ib/hr) below
the baseline emissions level. The methods used to obtain the emissions
reduction estimate are presented in Appendix A-5.
The water and solid waste impacts associated with implementing arsenic
EEC are minimal. No additional wastewater would be generated by the spray
chamber operation. All cooling water would be evaporated in the spray
chamber, leaving only a mud-like residue. This mud is then periodically
disposed of in a secured landfill along with other smelter slags that cannot
be recycled. From the standpoint of solid waste, the only other potential
impact of EEC implementation would be an increase in the impurity content of
the process slags generated at New Jersey Zinc due to enhanced recycling of
volatilized materials. However, this effect has not been demonstrated and
is not expected to significantly alter the disposal characteristics of the
slags.
5.4.2 Energy Impact
The recommended EEC option would require that additional energy be
consumed beyond that needed for baseline control. Additional energy would
be used to operate the gas cooler/spray chamber system for the waelz kiln
sinter offgas stream. The estimated annual power requirement of the cooling
system is 75 Gj (2 x 10 kWh).
5.4.3 Control Cost Impacts
The estimated capital and operating costs associated with the
implementation of arsenic EEC at New Jersey Zinc are shown in Table 5-5.
The total incremental annual cost in final quarter 1981 dollars associated
with implementing arsenic EEC is estimated to be $112,000 per year. The
methods used to obtain the cost estimates are summarized in Appendix B-5.
5.4.4 Economic Impact Resulting from EEC
An economic impact analysis was conducted for one zinc oxide plant to
estimate the potential impact of requiring EEC. All other facilities in the
83
-------
o
M H
H Z
jj p?
H M
to O
H Z
H
H
3: O
H
Q pd
w w
§ 1
o ---
CO CJ
CO Z
<; M
N
H >H
CO W
O CO
iyi
a >-3
z
M 3
H W
•rf J2J
&-3 F*1
PLI P3
O H
Q H
z <
<^
^a
H
M CJ
CU M
< Z
CJ W
CO
^•J p~l
H
>*-V f+yj
H O
iX
1
W
,_]
^-«
H
o o
as >-»
H Z
Z W
O tn Q
o os fei <-»
< > eo
0 sci-5
z S «.
H oi -^
O W
U C-
§Q
W
C*4 &.
J 0 H "fa
< o- z x
3 U U — »
z a: S «•
«4 pb I>-P")
-.°&2
< E- U •**
H 03
fi H O
O to W H
H O OS Z <-.
U m U «>
Q = Z
U -J O1 Ot*1
_J < U I— O
•J E- Btf = -H
H e- fa. H
« < o
z u
M
1x2 O
U3 > OS
Z M H
O H Z
S3 M < O
e w fcj ^-
< M fg U fa
ua ua z M ~-
M o iJ eo
U M H C/3 -^
< z 5 <
u s oa
en c
o: w o
< Oi H
Hg
u
OS CO
J Z O
O W H O
a! Z HH
H e- o z
O CTM B!
w 3 «e
o-
OS
W
Ed O
Z w
1-1 «
»J W
en z fa
< til f
pa ~-
u eo
J M J^
< z >-•
fr^ td
O O3
H OS
g
u
a. fa
< X
CJ •«.
eo
£ N— '
«s
0-
H
c
CU
ST
o
X
in
CN
esj
-
(N
•3-
o
in
c
X
u
•H
~~ o
fa ™
0) Cu
|3
J= 00
CJ C
X ^~*
a o
fa 0
O.CJ
^
m
CJ
o
o
o
o
r-i
.g
^
X C
01 O
01 U
fa fa
01 U
O f<
SCO
0) ~* &4
Z N
1
1
1
l
01
B
o
z
u
fa
41
0
fa
a
>:
fa
01
B.
i
N
s-/
13
-H
X
o
u
e
•H
N
eo
Z
c
•0
V
a
t-4
X
«H
U
Q
C.
a
u
u
c
a
i^
0-
e
x
01
CO
fa
o>
<->
9
01
z
eg
34
-------
zinc smelting and zinc oxide industry were determined to have EEC in place.
Using the cost data in Table 5-5 and the methodology given in Appendix C,
revenue and product price increases, required by the plant to maintain the
same net present values before and after the installation of EEC, were
estimated. To test the sensitivity of the economic impact results to the
various input data, a sensitivity analysis was conducted for the plant,
assuming simultaneous 15 percent decreases in baseline price and annual
output and a 15 percent increase in the weighted cost of capital.
The annual revenue increase required by the zinc oxide plant to
maintain its profitability under EEC is slightly less than $100,000, which
is equivalent to a product price increase of 0.3C per kg (0.14C per Ib).
The percentage increase in the product price is 0.3 percent.
Under the sensitivity conditions the analysis showed that the plant
would incur a unit price increase of 0.4C per kg of product in order to
maintain profitability. This price increase is equivalent to a percentage
unit price increase of 0.4 percent. Table 5-6 summarizes the economic
analysis results for the EEC and sensitivity cases.
5.5 POPULATION EXPOSURE DATA
The data in Table 5-7 were developed during this study for use in an
air pollution dispersion model to estimate the population exposure to
arsenic from primary zinc smelters. The Strategies and Air Standards
Division (SASD) of EPA is using these data in an ongoing effort to estimate
exposure to arsenic from this source category.
5.6 EASE OF STANDARDS DEVELOPMENT
The primary zinc smelting industry is predominantly controlled at the
EEC level for arsenic emissions, and standards development would be likely
to impact only one plant. There are no known technic-1 impediments to the
implementation of EEC at that plant. As discussed in Section 5.4.4, the
economic impact of requiring EEC would cause a product price increase of
0.3 percent. Economic feasibility does not appear to be an impediment to
EEC use at the zinc plant in question.
35
-------
TABLE 5-6. REVENUE AND PRICE IMPACTS FOR PRIMARY ZINC CATEGORY
MODEL
PLANT
REQUIRED REVENUE
INCREASE
(000 dollars)
REQUIRED UNIT
PRICE INCREASE
(dollars per kg)
REQUIRED PERCENT
PRICE INCREASE
EEC Case
99.6
.003
0.3
Sensitivity Case
104.6
.004
0.4
86
-------
3=
H
03
O
fe
<
H
<
Q
H
P
P-t
Z
M
O OfJ
Z H
M 02
. T ' — 1
i-3 O
W 2
O M
O
S 0
Ed M
Pd H
3 ,_!
CO W
o a
^
w o
z
Z M
O N
H ><
<£ Pi
i-j * O
jj a;
-r« n
c — .
O u
ei ^s
>
u
3 a
O CO
u. n
's
O *4«4
4-1 U
ac •— '
» 0
o tu
i— 1 fO
1*1 "-.
« "e
a:
u
£ i
Z E «•>
< a u.
S H L,
•H M
U! X 0
U U
<
c/:
u
0) y*.
0) y-t
e — •
a
•H S
a
w ^^
.s *•*
£0 u_<
01
= e
EMISSION
POINT
PLANT UTM
LOCATION COORDINATES
(km)
co co co ac
—• *— ' *-* ••— ' to to
(13 R (6
CLPO — co — a. o.
s — o o o E e
0 O O O O O C
Cj . . . . U U
0 O 0 0
>. X ^
.a ^= .a
00 CO CO CO
01 /-s *•"»/-> ^ 0) 01
f— I CTi CM r*. P-i i-^ ^H
^ CM o — ' o j; *a
•H O O O O -H «^
00 . . . . ttf 6C
•H O O O O •« «rt
60 00 00
at ro — co — ' 01 a
2 — 0 0 0 2 Z
O O O O
O O O O
^ oo co CM r*» o aO
m CM i/"» in m o o
cc
r>. pn -3- r^ m — • —
co \O rn xO —
lA «C sD i/^ vC O O^
\o m un *Io m — \£
m pst CM fi CM vo o
C CM O% Oi vO 0s- O
oo — * u"> ^ ^o "*o m
— -• S^- N^ -^X CM N^-
m -a- co •-» o IN o^
m en — es CM co O
P* n*.
^-v ^-i x-s ^-s /-v Q\ O
O O CM —4 O- • •
O O O CM O ON O
»a- es — ^- — i co -->
>»». =ai as» •*. 3 3
U. t»-
CL, i_i !-i M l-i
•< CJ 01 01 0) -CO
— . *j <_) *- jj y a-
0) u
C I- U
O O 4)
N X >
<:
D-
0) -
o ce
"-) U
. s
4-1 O
ua 3;
m
X-s
O
iTi
O
o
en
CM
o
o
r»
CM
g
0^
O
o
0
r^.
~*<~*
00
r-
CM
r^
c
o
o
QO
m
00
r-.
O
-»
cn
CO
•a-
vC
CM
-a-
o
o
en
m
o
o-
<
P£
ASARCO Zone 14 FI
Corpus Hor 658.0
Christ!, TX Vert 3075.9
m
o^
o>
O
0*
in
•a-
O
o
m
CM
•o
o
o
o
CO
CM
o
o
o
Ch
m
a^
o>
o
o
o
vO
en
N«*
r-^
0^
CM
cn
r*.
ON
O
in
m
"
o
in
r-
in
•a-
e-
•<
Qi
tt-
_3
1-4
U
01
% $
Z (0
<: oo
o
r-j
•w*
vO
O
O
o
•a-
CO
^o
O
C-4
*3-
^p
r-l
O
a*
in
00
0
S
o
CM
O
«
a.
<
tn
Jersey Zone 16 FI
Miniere Hor 463.8
Zinc Comp. Vert 4041.0
Clarksville,
TN
m
/-^
o
in
0
o
e*"i
CN
O
O
m
r-i
O
O
-T
O
C
o
<~*
CO
m
S"
C
v£)
in
^
vC
CM
in
CM
m
-
X— >
m
o-
m
_-
o
o
LT
O
n
&-
<
C£
National Zinc Zone 15 Ff
Bartelsville, Hor 233.7
OK Vert 4070.0
m
oo
CTv
r*
O
vC
CM
<
ce:
ASARCO Zone 17 Mf
Columbus, OH Hor 331.0
Vert 4429 2
tO
MD «4D \O —
^-v x-x x^, rM
CT\ CM \O C
O O O •
. . . o
O O O •*•'
p^.
^T — mo
O O O C
O O CO
\O >d
— vO ^
^ ^-V -H X-s
— CT> x-* vO
x-v O CM O
en o o o
-*" o* o c
m *i
r* r^- r*. <—«
\D & s£> ^O
O O O O
O 0 C C
O 0 O O
O cn i-- u*i
O *T O PN
m *-s -^ •<-*
CM »n co
-CO — —
— CM m — i
in in •*& in
^- yD m m
\O in \c ^
~UT^ ~^~^
v£) O f*- PM
m — « CM cn
in
x^x-. x-. O
o o tn c
cc o r- m
"T c*"i o^ m
*T 00 CM --
CM — CM O^
a
oc u
c re
•H c
C > '*
r- c a c
i-* iz. (*- »— i
u; -^
i- • u id
M QJ N i-*
-H w --^ CC 0*
aj C u v- E
3: W EC U- ^
O O
co O r*-
.M m — <
<
e-
to
x -
Ol C
w o
to 4->
u to
•n u o)
s^^
Ol M «
z: CM
&0
c
-H
Jj
OJ
•H
X
01
o
iJ
73
ns correspoi
£
3
O
O
"a
i
u
w
U-
J=
60
X
t3
C
C3
"01
4>J
4_)
U
U
4_*
W
0>
sa
0)
jj
-H
C
<5
C-
o
c
N
>>
c-
Jj
OJ
3
OJ
s
OJ
w
o .
u- X
r-H
c aj
li -*
The arsenic emissions estimates s!u
and FBC arsenic emissions, respect:
87
-------
5.7 REFERENCES
1. U. S. Bureau of Mines. Zinc: In: Mineral Facts and Problems.
Preprint from Bulletin 671. Washington, D.C., U.S. Government Printing
Office, 1980.
2. U. S. Bureau of Mines. Occurrence and Recovery of Certain Minor Metals
in the Processing of Lead and Zinc. Information Circular 8790.
Washington, B.C., U.S. Government Printing Office, 1979.
3. U. S. Environmental Protection Agency. Background Information for New
Source Performance Standards: Primary Copper, Zinc, and Lead Smelters.
Research Triangle Park, N.C. Publication No. EPA-450/2-74-002a.
October 1974.
4. Telecon. Cardenas, R., ASARCO, with Keller, L., Radian Corporation.
January 26, 1982. Information concerning the ASARCO/Corpus Christi
electrolytic zinc plant.
5. Telecon. Janeck, T., St. Joe Minerals Corporation, with Keller, L.,
Radian Corporation. February 11, 1982. Information concerning the
St. Joe Minerals zinc smelter in Monaca, Pennsylvania.
6. Donovan, J. R., and P. J. Stuber. Sulfuric Acid Production from Ore
Roaster Gases. Journal of Metals. 19; 45-50. November 1967.
7. Telecon. Suresh, M., Pennsylvania Bureau of Air Quality Control, with
Keller, L., Radian Corporation. January 18, 1982. Stack parameters at
the St. Joe/Monaca zinc smelter.
8. Letter from Janeck, T., St. Joe Minerals Corporation, to Keller, L.,
Radian Corporation. March 10, 1982. Arsenic emissions from the
St. Joe/Monaca zinc smelter.
9. Telecon. Benbenek, J., Illinois Environmental Protection Agency, with
Keller, L., Radian Corporation. January 8, 1982. Stack parameters at
the AMAX/Sauget zinc plant.
10. Telecon. Canon, J., Tennessee Division of Air Pollution Control, with
Keller, L., Radian Corporation. January 29, 1982. Stack parameters at
the Jersey Miniere zinc plant.
11. Telecon. Drake, J., Oklahoma State Air Quality Control, with
Keller, L., Radian Corporation. February 10, 1982. Stack parameters
at the National Zinc/Bartlesville zinc plant.
12. U. S. Environmental Protection Agency. National Emissions Data System.
Zinc Smelting. January 7, 1982.
88
-------
13. Memo. Keller, L., Radian Corporation, to arsenic file. April 20,
1982. Estimating arsenic emissions from zinc roasting acid plants.
14. Telecon. Crocker, J., EPA:Region VI, with Keller, L., Radian
Corporation. April 21, 1982. Stack Parameters for the ASARCO/Corpus
Christi zinc plant.
15. Letter from Silvestris, M., New Jersey Zinc, to Keller, L., Radian
Corporation. May 11, 1982. Stack parameters and flow diagram of the
New Jersey zinc plant in Palmerton, Pennsylvania.
16. Memo. Keller, L., Radian Corporation, to arsenic file. May 17, 1982.
Estimating arsenic emissions from the New Jersey zinc plant in
Palmerton, Pennsylvania.
17. Letter from Richardson, J. B., ASARCO, to Fensterheim, R. J., Chemical
Manufacturers Association. May 17, 1982. ASARCO response to Radian
emission estimates.
18. U. S. Environmental Protection Agency. Arsenic Emissions from Primary
Copper Smelters - Background Information for Proposed Standards.
Preliminary Draft. Research Triangle Park, N.C. Februrary 1981.
89
-------
6.0 SECONDARY LEAD SMELTING
This chapter discusses the secondary lead industry and its associated
arsenic emissions. Section 6.1 presents a brief source category
description. Section 6.2 discusses the regulations currently covering the
industry and the resulting level of control and associated arsenic
emissions. Section 6.3 presents a discussion of the EEC options and
Section 6.4 discusses the impacts of applying these options. Section 6.5
presents the input data to be used in assessing the population exposure to
arsenic emissions from secondary lead smelting. The ease of standards
development is discussed in Section 6.6.
6.1 SOURCE DESCRIPTION
6.1.1 Industry Description
Lead produced by smelting lead scrap is called secondary lead and
accounts for approximately half of the lead produced in the United States.
Secondary lead production in 1980 was 676,000 Mg (744,000 tons). The
secondary lead industry consists of between 50 to 66 ' lead smelters, owned
by some 26 companies. The differences in the reported number of smelters
may be due to inclusion of some lead refiners who do not reuse lead from
scrap but just remelt bullion from blast and reverberatory furnaces and
remove some impurities. Secondary lead smelters are located in 26 states
and in all ten EPA regions. Table 6-1 lists the major secondary lead
smelters in the United States. The 66 plants listed in^the table which will
be considered in this study are distributed as follows:
1. 18 large plants [with production greater than 22,700 Mg/yr
(25,000 tons/yr)];
2. 13 medium plants [with production in the range of 13,000 to
22,700 Mg/yr (15,000 to 25,000 tons/yr)]; and
3. 35 small plants [with production less than 136,000 Mg/yr
(15,000 tons/yr)].
6.1.2 Process Description
The normal sequence of operations of a secondary lead smelter are scrap
receiving, charge preparation, furnace smelting and refining and alloying.
As shown in Table 6-2, battery plates constitute the majority of the
recycled scrap. In 1980 over 76 percent of the total lead recycled came
from this source.
Scrap batteries which contain up to 50 percent by weight of lead have
to be broken or sawn open to allow removal of the lead plates and lead
90
-------
TABLE 6-1. MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES'
Annual
Plant Location 10
Interstate Lead Co., Inc.
Leeds, AL
Sanders Lead Company
Troy, AL
RSR Corporation
City of Industry, CA
ALCO Mining, Inc.
Gardena, CA
South West Metal
San Bernardino, CA
ASARCO
San Francisco, CA
Gould, Inc.
Vernon, CA
National Smelting & Refining
Denver, CO
Refined Metals Corp.
Jacksonville, FL
Globe Union, Inc.
Tampa, FL
Gulf Coast Lead Co.
Tampa, FL
NL Industries, Inc.
Atlanta, GA
TARACORP
Atlanta, GA
Sanders Lead Co.
rir,.1tv Furnaces
-Capacity
Mg Blast Reverberatory Rotary
30 1 1
50 22
30 1
5 1
15 1
N/A 1
36 22
N/A
12 1 1
N/A
15 1
18 1
30 1 1
50 22
Cedartown, GA
91
-------
TABLE 6-1. (Continued) MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES'
Plant Location
Chloride Metals
Columbus, GA
TARACORP
Granite City, IL
TARACORP
McCook, IL
North Chicago Smelting
& Refining
North Cook, IL
Refined Metals Corp.
Beach Grove, IN
USS Lead Refining, Inc.
E. Chicago, IN
Northern Indiana Dock Co.
E. Chicago, IN
RSR-Quemetco Corp.
Indianapolis, IN
Mincon (Sold to Bergsoe)
Muncie , IN
ASARCO
Whiting, IN
Delco Remy Battery
Olathe, KS
Schuykill Metals Corp.
Baton Rouge, LA
General Battery Corp.
Heflin, LA
Sommerville Smelting
Furnaces
Annual _ Capacity -—-—————_—— ^_^_
10 Mg Blast Reverberatory Rotary
12 1
24 1
12 1
N/A
20 1
20 1
N/A 1
20 1
22 1 1
N/A 1 1
N/A 1
60 22
13 1
N/A 3
Sommerville, MA
92
-------
TABLE 6-1. (Continued) MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES1
Plant Location
New England Smelting Works
W. Springfield, MA
Federal Alloys
Detroit, MI
Industrial Smelting Co.
Detroit, MI
TARACORP Industries
St. Louis Park, MN
Gopher Smelting & Refining
St. Paul, MN
Chloride Metals, Inc.
Florence, MS
Schuykill Metals
Forest City, MO
Gould, Inc.
Omaha, NE
Delco Remy Battery
New Brunswick, NJ
Essex Metal Alloys
Newark, NJ
Aetna Alloys Division
Newark, NJ
NL Industries
Pedricktown, NJ
E. L. Beth
Perth Amboy, NJ
CAL-WEST
10 Mg Blast Reverberate ry Rotary
N/A
Shutdown
N/A
12 1
Co. 18 1 1
12 1
30 1
12.5 1
N/A 2
N/A
N/A
70 1
N/A lc
N/A
Socorro, NM
93
-------
TABLE 6-1. (Continued) MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES'
Annua 1
Plant Location 10
Republic Metals
Brooklyn, NY
Roth Bros. Smelting Corp.
Dewitt, NY
Willard Lead Co.
Charlotte, NC
Master Metals, Inc.
Cleveland, OH
River Smelting & Refining
Cleveland, OH
Bergsoe Metal Corporation
Portland, OR
Lancaster Battery Co.
Lancaster, PA
East Penn Manufacturing Co.
Lyons Station, PA
Tonolli Corp.
N es q ueho ning , PA
Gould, Inc.
Philadelphia, PA
Imperial Metal Co.
Philadelphia, PA
General Battery Corp.
Reading, PA
Marjol Battery Corp.
Throop, PA
General Smelting & Refining
Cauacif Furnaces
lidpclL.-!. t-y
Mg Blast Reverberatory Rotary
N/A 1
N/A . 1
10 1
24 1
N/A
27 b
N/A 1
15 1
40 4
N/A 1
N/A 1
60 22
N/A
N/A 1
College Grove, TN
94
-------
TABLE 6-1. (Continued) MAJOR SECONDARY LEAD SMELTERS IN THE UNITED STATES'
Plant Location
Refined Metals Corp.
Memphis, TN
Ross Metals, Inc.
Rossville, TN
General Battery Corp.
Annual- Capacity
ICT Mg
24
9
25
Blast
1
1
1
Furnaces
Reverberatory
1
1
1
Rotary
Dallas, TX
RSR Corporation
Dallas, TX
NL Industries, Inc.
Dallas, TX
Gould, Inc.
Frisco, TX
Houston Lead Co.
Houston, TX
Standard Industries
San Antonio, TX
Hyman Viener & Sons
Richmond, VA
RSR Corporation
Seattle, WA
Crown Metal
Milwaukee, WI
60
Shutdown
30
Shutdown
N/A
N/A
20
N/A
Information compiled from the Lead Industries Association, a NEDS listing,
State air pollution control agency contacts.
Bergsoe type furnace.
Q
Furnace is either under construction or not operating.
NA • Data not available, most likely to be less than 15,000 tons/year
production or refiners using pot furnaces.
95
-------
CO
OH
[V|
J22
M
ftt x" *•
W I *
iv* C
PS ,£-
00
O -H
2 0)
CO CO
pi CO
w o
>J 60
W
a •
co co
>• o
PQ 4-1
P-i U
CJ 4-1
co a)
§5
W
tJ O
rt"l
WJ
d ON
t—3 ^-4
O
Q H
<; co
3 H
W <
3 H
CO
[yt
O O
w
O M
H P
2 Cd
5 35
co H
O 53
CJ M
•
CM
1
W
PP
53
H
C
O
•H
4-1
ft
a
3
CO
C
o
*
o
H
ft
iH l-i
O 0
CO
ft
5 ca
01 s-i
52 o
CO
w
ft
1-1
o
01
es
CO .-i
Ai
1
co H
CO
CO
r-1
CJ
en ON CM en ^ •— i in o
en •— * <— i vo I-H t— i i 'H
•H « • tH CO C • CO
u-i • o^ ,0 c oi
01 •• i-t • l-i
l-i « T3 C 4J
. CO O CO T5
T3 T3 ^ £3 "X3 r~H C
C* *^3 'O ^0 ^^ £ C CO cQ
CO cd co 0) 1 O ca 4-1
0) 0) r-t >-, O 01 CO
w r-i IH )-i >-i a o>
>_i o) oi 'ct oi co
0>4JT3r-l4J 0)T3 0) CO
4-14-1 )-lra4->Xi-l ftO
^H O ^0 CO 03 *^ O ^*> J^
O W3 p^i CJ pQ *g^ ^/^ f-^ J^)
a
CO
o
o
0
m
CM
en
m
ON
*^-
VO
O
O
CM
co
\D
co
M
en
en
1-1
ON
»^-
r-^
co
en
ON
t— 4
ON
O
M
•^3"
r^.
.
•
,
.
.
•
*
•
•
•
•
•
i—!
(3
4J
o
H
en
O)
C
•H
S
M-l
O
3
TO
0)
^i
3
CO
• •
0)
o
i-l
3
O
CO
96
-------
oxide. This operation can be done manually or mechanically. Some secondary
lead smelters crush and grind whole batteries and then separate the
component parts using heavy media float-sink separators.
Prepared battery scrap is then combined with other furnace feed
materials and charged to the furnace. There are three types of smelting
furnaces commonly used at secondary lead smelters. These are blast,
reverberatory, and rotary (i.e., rotary reverberatory) furnaces. Table 6-3
compares the three types of furnaces. The following paragraphs will discuss
each furnace type in further detail.
A simplified flow diagram of a single blast furnace system is presented
in Figure 6-1. A blast furnace is a vertical unit and is charged through a
door at or near the top of the furnace. The blast furnace charge material
typically consists of: 82.5 percent reverberatory slag which consists of
oxidized lead and other metal oxides and reduction drosses from the refining
operation; 5.5 percent coke (for combustion); 3 percent limestone; ,
4.5 percent scrap iron; and 4.5 percent rerui} slag (from previous runs).
Air is "blasted" [using 24-32 Pa (8-12 oz/in^") gauge pressure] into the
furnace through tuyeres near the base. Oxygen enrichment is sometimes used
to increase metal production and recovery of antimony and arsenic values.
Oxygen enrichment also lowers fume production and coke usage.
As the charge material melts, the iron, silica and limestone form an
oxidant-retardant flux which floats to the top of the melt. Molten lead is
tapped almost continuously into a holding pot. When battery groups are
being charged approximately 70 percent of the charge material is tapped off
as hard (or antimonial) lead, which may contain as much as 12 percent
antimony. The molten lead is cast into large ingots called "buttons" or
"sows." Approximately 7 percent of the charge leaves the furnace as dust
and 18 percent of the remaining material is tapped as slag and matte.
Approximately 5 percent of the slag is recharged into the blast furnace. A
typical range for blast furnace production is 18 to 7_ Mg/day
(20-80 tons/day).
There is a wide range of temperatures found throughout the charge
column in a blast furnace. Temperatures of 1230°C (2240°F) or greater are
encountered in the slag zone and adjacent to the tuyeres. Lower
temperatures are encountered up the charge column and typical gas exit
temperatures at the top of the blast furnace are 430-540°C (800-1000°F).
The major sources of air pollutant emissions from a blast lurnace are
shown in Figure 6-1. Typically, furnace emissions are controlled by an
afterburner, a baghouse, and/or a scrubber. Knock-out boxes are frequently
used for collection of particulate which separates from the gas flow in the
ducts. Furnace emissions are discharged to the atmosphere through a stack.
Charging, slag tapping, and lead tapping are process fugitive emission
sources and may be hooded and ducted to the process baghouse or to a
separate sanitary baghouse for recovery of lead-containing particulate.
97
-------
TABLE 6-3. COMPARISON OF FURNACE CHARACTERISTICS
CHARACTERISTIC
BLAST
TYPE OF FURNACE
REVERBERATORY
ROTARY
Charge material
Fuel
Operation
Products
Reverberatory
slag
Drosses
Lead scrap
Antimonial ore
Iron scrap
Sand
Limestone
Recycled blast
furnace slag
Reducing agents:
Coke, sawdust,
ground battery
cases
Coke
Batch/semi-
continuous
Hard lead
Crushed batteries
or battery plates
Lead scrap
Flue dust
Carbon
Natural gas;
No. 2 fuel oil
Continuous
Semisoft lead
Batteries
Lead paste
Cast iron
borings
Anthracite
Limestone
£
Soda ash
Natural gas;
No. 2 fuel oil
Batch
Semisoft lead;
lead oxide
Typical air
pollution
device
Operating
temperature °C (°F)
Exit temperature
of furnace offgas
°C (°F)
Nominal furnace
capacity Mg/day
(tons/day)
Afterburner
Cooling device
Baghouse
430-1320
(800-2400 )
430-730
(800-1350)
36
(40)
Cooling device
Baghouse
1260-1320
(2300-2400)
1260-1320
(2300-2400)
45
(50)
Cooling device
Baghouse
1260-1320
(2300-2400)
1260
(2300)
36
(40)
Short rotary furnace only; does not include a long rotary kiln.
For typical furnace operations. Significant variations can occur from plant to
plant.
For soda matte process; removal of SO as slag.
Temperature varies throughout blast furnace column. Highest temperatures will
occur around or just above the blast zone.
98
-------
•—v
•H
00
U-i
oo
c
"to
u
03
CJ
J
00
3
00
c
c.
«3
*•*
00
to
c
JJ
0)
nj
y
00
c
a.
o
u
*n
to
—
CJ
^^
o
00
3
FH
F-t
•UJ
3C
O
•H
4J
O
-a
o
^
cu
C8
CU
13
C
O
a
§
u
CO
>,
CD.
a;
a
to
cn
eo
fO
a
•H
O.
>>
H
J-i
Sx _i as
O en M
99
-------
A simplified flow diagram of a single reverberatory furnace system is
presented in Figure 6-2. The reverberatory furnace uses gas- or oil-fired
burners. The charge material is heated by radiation from the flame and from
the furnace walls. As indicated in Figure 6-2, the reverberatory furnace
charge material typically includes: lead scrap, battery plates, lead oxides
and recycled flue dusts. Charge material is added as more of the solid
material in the furnace becomes liquid. The reverberatory furnace operates
at a temperature of about 1260°C (2300°F) and near atmospheric pressure.
Molten metal (i.e., semisoft lead containing 0.3 to 0.4 percent antimony) is
tapped into molds periodically as the level rises in the furnace.
Typically, a reverberatory furnace produces 45 Mg/day (50 tons/day).
Approximately 47 percent of input is recovered as lead, 46 percent is slag,
and 7 percent is lost as particulate and metal fume.
The rotary furnace, which is similar to the reverberatory furnace, is
used in Europe in most secondary lead plants. A major difference between
the rotary and reverberatory furnaces is that the rotary furnace rotates
slowly during heating of the charge material.
A reverberatory (or rotary) furnace can be used simultaneously to:
(a) reclaim lead from oxides and drosses, or (b) sweat lead from metallic
scrap. Sweating utilizes the differences in melting points to separate lead
from other metals.
Figure 6-2 indicates the major sources of air pollutant emissions from
a reverberatory furnace. Furnace (nonfugitive) emissions are typically
controlled by an exhaust gas settling chamber, a baghouse, and/or a
scrubber. Process fugitive emissions from charging, lead tapping, and slag
tapping may be hooded and ducted to a separate baghouse.
Following the smelting operation the hard and semisoft lead is further
refined and alloyed as necessary. Refining and alloying are done in pot
furnaces. The process is a batch operation and may take a few hours to
2-3 days depending upon the degree of purity or alloy type required. Pot
furnaces are gas- or oil-fired and range in capacity from 1 to 45 Mg
(1-50 tons). Pot furnaces operate at temperatures ranging from 320°C to
480°C (600°F to 900°F) and in addition are usually hooded. Emissions from
pot furnaces are therefore lower than from blast and reverberatory furnaces.
6.1.3 Arsenic Emissions from Secondary Lead Smelters
Emissions from secondary lead smelters occur from three groups of
sources. These are the process sources, the process fugitive sources and
the nonprocess or area fugitive sources. The predominant sources of process
emissions are the smelting furnaces and the refining pots. Process fugitive
emissions are generated during charge preparation, furnace operations and
refining operations. There are also several nonprocess fugitive emission
sources throughout the plant including the raw material storage piles, the
100
-------
ce
&
60
a
i> 00 jj u
•H 3 » (0
JJ u* td ^
3 CO DO CO
«-> C C O
••-^ *•< -^ -H
O. C- 00
GO C. C. W
C CO ffl 3
*•« *rf 4J «-4
60 -4
i-l QO -O *C
to re it a
6 3
c
o
•H
JJ
O
o
M
a.
*a
«
0)
CO
•9
O
O
a»
CD
o
U-l
I
4J
CO
0)
o
td
c
M
o
1-1
cu
0)
C8
o
CM
vO
to
•H
101
-------
haul roads, and the charge preparation area. Fugitive emission_sources
common to most secondary lead smelters are listed in Table 6-4.
Arsenic is present in many of the furnace feed materials and in all of
the furnace products. Figure 6-3 illustrates the many input and outputs
from typical blast and reverberatory furnaces. Table 6-5 gives ranges or
typical amounts for each feed material and the range of observed arsenic
contents. A detailed arsenic material balance cannot be constructed using
the data presented in Table 6-5 because of the wide ranges reported in both
weights of feed materials and arsenic contents.
During the smelting operations the arsenic containing raw materials are
subjected to high furnace temperatures and either oxidizing or reducing
conditions. In the reverberatory furnace, which produces an oxidizing
environment for most lead alloying constituents, some arsenic trioxide is
formed. This can vaporize and leave the furnace in the offgases.
Alternatively the arsenic can become complexed in the slag and leave the
furnace in this stream. Most reverberatory slags which are predominantly
composed of metal oxides are charged to blast furnaces to recover the lead
values. The blast furnace has a lower exhaust gas exit temperature than the
reverberatory furnace and in addition operates under a reducing environment.
Any arsenic in the charge materials either vaporizes and exits the furnace
in the blast furnace offgases or dissolves in the,product metal. Very
little arsenic remains in the blast furnace slag. '
Many factors can affect the emissions from the smelting operations,
including the amount of arsenic in the feed material, the operating
conditions of the furnace, the amount of chlorides in the feed material, and
the slag composition. Table 6-5 shows that the amount of arsenic present in
a given feed material can vary widely. In addition, the type of material
charged to any furnace varies widely from smelter to smelter. For example,
some plants charge whole batteries and others charge carefully segregated
plates and scrap. Arsenic emissions in terms of g/kg of lead produced could
therefore vary widely from plant to plant.
Furnace operating conditions can vary widely from smelter to smelter.
Reverberatory furnaces can be operated to conserve arsenic and antimony
values in the slag. Similarly blast furnaces can be operated to conserve
these metal values in the product metal, or can be operated to produce a
metal with lower arsenic and antimony contents. The amount of chlorides in
the furnace charge material could also affect the amount of arsenic
emissions. If any chlorides are present in the feed material, arsenic
trichloride, a very volatile compound may be formed and subsequently
liberated from the furnace. Secondary lead smelters who charge whole
batteries and groups of battery plates which contain polyvinylchloride plate
separators would be expected to have higher arsenic emissions than smelters
who charge only battery plates and battery mud.
10;
-------
TABLE 6-4. FUGITIVE EMISSION SOURCES AND CONTROL METHODS
Fugitive Emission Source
Control Methods'
I. PROCESS FUGITIVE EMISSIONS
A. Battery breaking
Battery/charge preparation
B. Furnace operations
1. Charging
2. Slag tapping
3. Lead tapping
C. Refining
Hoods and scrubber/FF
Washing down by hose
Hoods ducted to a separate
baghouse
Slag caster
Hoods ducted to a separate
control device
D. Casting
Hoods
II. NONPROCESS FUGITIVE EMISSIONS
A. Raw materials
1. Unloading of trucks/
railcars
2. Storage/wind erosion
3. Transfer to furnace area
Enclosure; separation; covering
with plastic sheets; wetting
down by hose
B. Product/scrap materials
1. Transfer to storage
2. Storage
3. Loading to trucks/
railcars
C. Furnace area floors
D. Alloying department floors
E. Paved/unpaved roads
1. Vehicular activity
2. Storage
3. Loading to trucks/railcars
Agglomeration of flue dust,
enclosure, separation,
wetting down by hose;
washing of trucks before
leaving plant property
Paving, vacuuming, wetting
down by hose
Oiled sawdust on floor —
vacuumed and charged to
blast furnace
Paving, vacuuming, wetting
down by hose
Observed in varying degrees at some or all
the MRI study. (Reference 2)
of the nine plants visited in
103
-------
FLUE GASES
PLATES SCRAP, ETC. 1
FLUX MATERIALS 2
FLUE DUSTS 3
REVERB SLAG 4
PLATES SCRAP, ETC. 12
FLUE DUSTS 13 —
REntTCIUG AGENTS 8 -^
> ,
7
BLAST
FURNACE
!'
BLAST FURNACE
SLAG
FLUE GASES
REVERB ERATORY
FURNACE
9
1 '
6
10
-^- HARD LEAD
REVERBERATOR! FURNACE
SLAG
SEMISOFT LEAD
Figure 6-3. Secondary lead material balance,
104
-------
TABLE 6-5. SECONDARY LEAD MATERIAL BALANCE
(Typical Reported Values) ' '
STREAM (From Fig. 6-3)
WEIGHT % of FEED
ARSENIC CONTENT Wt %
INPUTS
1. BLAST FURNACE FEED
(Plates, Scrap, etc.)
2. FLUX MATERIALS & COKE
3. AGGLOMERATED FLUE DUSTS
4. REVERBERATORY SLAG
TOTAL FEED MATERIALS
OUTPUTS
5. BLAST FURNACE SLAG
7. FLUE DUSTS
6. HARD LEAD FROM BLAST
FURNACE
BLAST FURNACE
0-80
5-20
0-15
0-80
100
28
7
65
0.05 - 0.2
0
0.3
0.003 - 2
0.001 - 1.0
0.3
1 - 3
INPUTS
12. BATTERY SCRAP, OXIDES
DROSSES
13. FINE DUSTS
8. REDUCING AGENTS (COAL
FINES, SAWDUST, ETC.)
REVERBERATORY FURNACE
70 - 95
0-10
0-5
TOTAL FEED MATERIALS
OUTPUTS
9. REVERB FURNACE SLAG
10. SEMISOFT LEAD FROM
REVERB
11. FLUE DUSTS
100
47
46
0.05 - 0.2
0.3
0
0.003 - 2
0.001 - 0.05
0.3
105
-------
Blast furnace slag composition can have an affect on arsenic emissions.
Arsenic forms complexes with iron and other metals and can be removed from
the furnace as part of the slag, thereby reducing arsenic levels in blast
furnace offgases.
There are three main sources of process fugitive arsenic emissions.
These are the charging operation, the slag tapping operation and the metal
pouring operation.
The magnitude of arsenic emissions from process fugitive sources would
be expected to vary with the arsenic content of the charge material.
Emissions from the charging operation will consist of fine particulates and
fumes. When the blast furnace is charged some of the furnace offgases will
exit the furnace through the charge hole. If the charge contains any fine
material this can be entrained and become a potential process fugitive
emission. The major source of fines in secondary lead smelter charge
material is the recycled flue dusts. As shown earlier these dusts contain
an appreciable amount of arsenic. Several smelters are currently using an
agglomeration furnace to process flue dusts prior to recycling them to the
process. The agglomeration step produces a material which is more easily
handled and less subject to dusting.
During the slag and metal tapping operations a considerable amount of
fumes are generated because of the high temperatures involved. If these
fumes are not captured and controlled they can constitute process fugitive
emissions. Most furnace tapping operations are currently equipped with
capture hoods and fabric filters. The arsenic content of the fumes
generated during slag tapping and metal pouring should reflect the pouring
temperature and the arsenic content of the slag and product metal
respectively. The arsenic contents of the slag and product metal in turn
should reflect the amount of arsenic in the charge material and the effects
of the furnace operating procedures.
Arsenic emissions from area sources will be influenced by the arsenic
content of the various fine materials being stored around the smelter. This
includes the non-agglomerated flue dusts and the dried battery mud. Other
factors which will influence the total amount of arsenic for this source
include the meterological conditions, especially the amount of wind and
rain, and the amount of activity around the plant site.
6.2 INORGANIC ARSENIC EMISSIONS UNDER EXISTING REGULATIONS
6.2.1 Regulatory Impacts
All of the point sources in an existing secondary lead smelter are
covered by the applicable SIP process weight regulation for particulate and
in some cases the NSPS for particulate emissions from secondary lead
smelters. The NSPS which requires particulate emissions from reverberatory
and blast furnaces to be less than 50 mg/m (0.022 gr/dscf) is the more
106
-------
stringent of the two regulations. Available data indicates that the
majority of secondary lead smelters are easily complying with the applicable
SIP and NSPS. In most cases the smelters are exceeding the NSPS
requirments. In addition to particulate regulations, some States and local
agencies are requiring control of SCL emissions from reverberatory and blast
furnaces.
Fugitive particulate emission sources are currently impacted by three
regulations. These are (1) the OSHA workplace standard for lead, (2) the
EPA national ambient air quality standard (NAAQS) for particulate and,
(3) the EPA NAAQS for lead. These three regulations are briefly discussed
below.
The Occupational Safety and Health Administration (OSHA) standard,
recently proposed, limits in-plant lead concentrations to 50 yg/m based on
8-hour worker exposure. This standard directly affects the control of
process and some nonprocess emission sources within a secondary lead plant.
Individual plants are being evaluated to determine work practices, control
equipment, and ventilation measures that will be needed to achieve the OSHA
standard. All of the existing fugitive emission control technology now used
in the secondary lead industry was instituted to meet the OSHA standard.
The EPA National Ambient Air Quality Standard (NAAQS) for particulat^
limits ambient concentration to 260 yg/m (primary standard) and 150 yg/m"
(secondary secondard) for a 24-hour period. No evidence was found that the
standard was violated at or around any of the secondary lead plants visited
by MRI, as part of the NSPS revision project for secondary lead smelters.
3
The EPA NAAQS for lead limits the ambient concentration to 1.5 yg/m
for a 3-month average. The standard, promulgated in 1978, requires the
States to revise State implementation plans (SIP) to pinpoint areas where
emission control is needed. Personnel at several plants visited by MRI
indicated that this standard was not being met at or round their plant.
6.2.2 Existing Controls and Inorganic Arsenic Emissions
6.2.2.1 Process Sources. As a result of complying with the applicable
SIPs and NSPS most secondary lead smelters have applied fabric filters to
control particulate emissions from their blast and reverberatory furnaces.
Typically these fabric filters are preceded by an afterburner and cooler
when used on a blast furnace and a cooler alone when used on a reverberatory
furnace. Several secondary lead smelters have also installed a scrubber
after the fabric filter to control SO. emissions. There are very little
arsenic emissions data available for secondary lead smelters. An
uncontrolled emission factor of 0.4 kg/Mg lead produced has been used in
several studies of the secondary lead industry, however the basis of this
emissions factor appears to be engineering judgement combined with analogy
to primary lead smelters. '
107
-------
The California Air Resources Board has conducted an arsenic emissions
test on the outlet of a fabric filter controlling a reverberatory furnace
and a blast furnace. Results of this test are not available at this time.
As discussed earlier, no precise material balance can be constructed
around a typical smelter or furnace because of the wide ranges in arsenic
contents of the feed materials and variability in operating practices.
However, particulate and lead emissions from furnaces at secondary lead
smelters have been well characterized, Table 6-6 presents some test results
for fabric filters and scrubbers applied to blast and reverberatory
furnaces. In general blast furnaces have slightly higher particulate and
lead emissions rates than reverberatory furnaces. For the tests listed in
Table 6-6 blast furnaces emitted 14 percent more particulates than a
reverberatory furnace on a g/kg lead produced basis.
Several studies have shown that lead and arsenic behave in a similar
manner when exposed to high temperature environments. These studies, which
are discussed below, will be used as the basis for drawing parallels between
lead and arsenic emissions from secondary lead smelters. In addition, the
studies are supportive of using control device efficiency measurements based
on lead to estimate arsenic removal efficiencies.
Two studies of the behavior of trace elements during coal combustion
have shown that lead and arsenic behave in a similar manner. Both are
preferentially enriched in the fly ash relative to the bottom asluancLboth
show increasing enrichment with decreasing fly ash particle size. ' This
behavior has been attributed to a volatilization/condensation mechanism
occurring during the passage of the flue gas through the coal-fired boiler.
The elements and their compounds are volatilized in the high temperature
combustion zone and then either condense or adsorb onto entrained particles
as the flue gases cool down as they pass through the boiler and control
devices. The mass deposited is greater per unit weight for the smallest
particles since these have the highest surface area. Table 6-7 shows the
enrichment ratios reported for lead and arsenic from several studies. The
table shows that lead and arsenic track each other fairly well.
The combustion zone of a coal-fired power plant is considerably hotter
than a reverberatory or blast furnace. In addition, the initial
concentrations of lead and arsenic in coal are considerably lower and at
different ratios than that found in secondary lead furnace charge materials.
However, secondary lead smelters operate at above 1000°C (1830°F) which is
sufficient to vaporize considerable quantities of both elements and their
oxides and chlorides. Figure 6-4 shows the vapor pressures of several
species which are present in both blast and reverberatory furnaces. Several
lead compounds and all arsenic compounds exhibit appreciable vapor pressures
above 1000°C (1830°F).
108
-------
TABLE 6-6. EMISSIONS TEST RESULTS FOR SECONDARY LEAD SMELTERS
PLANT
Sanders13
Lead
Totiolli14
SOURCE
Blast
Furnace
Rotary
Furnace
(Fabric Filter
& Scrubber in
Series)
General15
Battery
A16
B
C
D
E
F
G
H
I
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Reverb
Furnace
Reverb
Furnace
Reverb
Furnace
CONTROL
DEVICE
ABa/Fabric
Filter
AB/Fabric
Filter
Scrubber
AB/Fabric
Filter/
Scrubber
AB/Fabric
Filter/
Scrubber
AB/Fabric
Filter
AB/Fabric
Filter/
Scrubber
Scrubber
AB/Fabric
Filter
AB/Fabric
Filter
AB/Fabric
Filter
Fabric
Filter
Fabric
Filter
Fabric
Filter
AVERAGE FABRIC FILTER
AVERAGE FABRIC FILTER/
SCRUBBER COMBINATION
TEMPERATURE
"C (°F)
-
31
(87)
-
36
(96)
82
(178)
49
(121)
35
(95)
67
(152)
79
(175)
43
(110)
76
(168)
51
(124)
164
(327)
80
(176)
38
(101)
PARTICULATE
ug/Nm
(gr/dscf)
0.5
(0.0002)
14
(0.00616)
28.7
(0.01264)
54.5
(0.024)
45.5
(0.020)
22.7
(0.01)
5.9
(0.0026)
18
(0.0079)
32.5
(0.0143)
3.0
(0.0013)
13.4
(0.0059)
32.3
(0.0142)
8.0
(0.0035)
7.5
(0.0033)
5
(C.0022)
9.8
(0.0043)
22.7
(0.010)
LEAD3
Ug/Nm
(gr/dscf)
-
0.3
(0.00013)
0.5
(0.00024)
0.4
(0.00018)
0.2
(0.00007)
0.3 - 0.6
(0.00014 -
0.00025)
0.8
(0.00034)
0.3
(0.00012)
2.4
(0.00105)
1.4
(0.00061)
-
.
1.4
(0.00061)
0.9
(0.00038)
-
0.9
(0.000414)
0.4
(0.00017)
..FLOW
m /Mg Pb
(dscf/ton Pb)
14,853
(462,700)
15,736
(490,200)
15,810
(592,500)
10,320
(321,500)
31,300
(975,000)
6,565
(204,500)
3,450
(419,000)
17,662
(550,200)
23,565
(734,100)
AB - afterburner
109
-------
Q
Z
<
u
M
CO
OS
OS
o
fe
u
0
z
— ,f4
3
3 CO
O H
fa Z
OS <
U _]
PLI &4
u BS
u u
1— 1 S
> 0
W &4
O
o
hJ U
q PS
OS M
H fa
Z 1
0 ,J
°3
Q CJ
5s
O
en os
O fa
§z
clo
l—t
H co
Z CO
W M
s s:
S u
u
M O
OS <
z w
W _3
r^»
1
\Q
W
i_]
CQ
•<
H
ca
o
n
as
|
^
"3
a
>.
[b
M
o
H
2
H
z
UJ
3 "
S £
i-* E
a£ £
Z >-t
u u
o
o — —
c
-^ — ON
< m %r
pa \o
< — ^
o
-» \o
(8 -• —
o
G
r- o
JA ~* r*i
s
6Q
/-s.
0=
I
»f
•^ • fj
1
C^
CM
<
§
O
• — ff.
f» f^. f.
1
o
u
f4
C
O "O
W C5
U U
-C U
c
*>H
6 *£,
"^*- O
eo m
S
"S
c
OJ
60
^"- O
w w
g
u
_
E
u:
CJ
2
H
Z
o
u
u
z
2
1
u
a.
oi
u
ca
CO
2
ae
u
C/3
Q
Z
<
a£
O
S
G.
O
as
a.
u
t-
S
s
DC
J
w
00 —i
;i *-4
c
'tH
E ao
lo S
E
U
c
a
n
i_i
<
00
•7- o
BO n
3. —
0)
±J C
n IH
-3 4
I
*^. o o i
U r*i 00 I
=1 - fM I
OJ —I -^
110
-------
1)
M
01
CO
0)
M
Hi
h
o
D.
(d
0.5
AsJJ
2 3
2.0
10 X Reciprocal Absolute Temperature, °K
1500
1000 800
600 500 400
Temperature, (°C)
300
200
Figure 6-4. Vapor pressure of Pb, PbO, Sb , PbS, PbCl , As,
and
.
Ill
-------
Several studies of particle_size distribution at secondary lead
smelters have been performed. ' These studies have shown that the lead
content of the particulate matter increased with decreasing particle size
for hot emissions streams. These studies are supportive of the
volatilization/condenstation mechanism, and of the extrapolations between
trace element behavior in coal-fired power plants and lead and arsenic
behavior in secondary lead smelters.
The use of lead emission rates as indications of arsenic emission rates
from secondary lead smelters is further supported by studies of lead
contamination,of.urban soils by emissions from secondary lead
smelters. ' ' These studies showed similarities in the patterns of lead
and arsenic contamination around two secondary lead smelters. In addition
one of these studies proposed the use of_arsenic as a tracer for certain
kinds of industrial lead contamination.
Estimates of arsenic emission rates from secondary lead smelters were
developed using measured lead emissions rates as the basis in conjunction
with observed lead to arsenic contamination ratios in urban soils and
vegetation surrounding secondary lead smelters. The estimation procedure is
fully described in Appendix A-6. The major assumptions behind the
estimation procedure are given in Table 6-8. For the purpose of this study
the assumptions were chosen to maximize arsenic emissions. It should be
noted that the estimated arsenic emission rates are very sensitive to
changes in any of these assumptions. For example, a change in assumed lead
to arsenic ratios from the upper end of the range to the lower end would
decrease arsenic emissions by a factor of eight or more. Table 6-6 showed
that a fabric filter/scrubber combination has lower lead emissions than a
fabric filter alone. This is to be expected because the outlet temperature
of the scrubber is considerably lower than the outlet temperature of the
fabric filter alone and therefore more lead vapor would be condensed and
captured as particulate. Chapter 2 of this report discusses the effect of
temperature on arsenic removal efficiency and indicates that at lower
control device temperatures, lower outlet arsenic concentrations are
achievable. Arsenic and lead from combustion sources have been shown to be
removed at approximately the same efficiency by a given control device.
Results of a test on the arsenic and lead removal efficiency of a scrubber
and ESP are given in Table 6-7. The table shows that, in general, the
control devices are more efficient for arsenic removal than for lead
removal. The observed lead removal efficiencies for control devices on
secondary lead smelters were used to predict arsenic removal efficiencies.
6.2.2.2 Process Fugitive Emissions Sources. The major process
fugitive emissions sources are the furnace charging operation, the slag
tapping operations and the metal tapping operation. These operations are
controlled by a combination of hoods and enclosures coupled to fabric
filters at the majority of secondary lead smelters. Some secondary lead
smelters use a single fabric filter to control all process fugitive
emissions, and others use a separate fabric filter to control the slag
112
-------
TABLE 6-8. ESTIMATION OF ARSENIC EMISSIONS FROM PROCESS
SOURCES OF SECONDARY LEAD SMELTERS
ASSUMPTIONS FOR EXISTING EMISSIONS
- Lead to arsenic ratio in particulate emissions is 13:1
(range seen 110:1-13:1).
- Lead concentration in particulate matter is 42 percent
(range seen 18 percent - 52 weight percent).
- Particulate concentration in fabric filter outlet is 50 ug/m based
on NSPS limit (test data .5 - 23 ug/m ).
- All process fabric filters have the same outlet particulate
concentration.
ASSUMPTIONS FOR FABRIC FILTER FOLLOWED BY A SCRUBBER
- Arsenic emissions will be reduced at the same efficiency as lead
emissions.
- The incremental reduction of lead emissions by a fabric filter/scrubber
combination over a fabric filter alone is 59 percent.
RESULTS
- Arsenic emission rates-for fabric filters on blast and reverberatory
furnaces are 1.6 ug/m .
- Arsenic emission rates for fabric filler/scrubber combinations are
0.66 vg/m .
113
-------
tapping operation and couple all other hoods into a single fabric filter.
Fabric filters used to control process fugitive emissions are often called
sanitary filters.
The emissions collected and controlled by these sanitary filters are at
a much lower temperature than the emissions from the blast or reverberatory
furnace. This is because the hooding and capture system entrains large
volumes of ambient air as it collects the fugitive particulate emitted
during the charging or tapping operations.
No arsenic test data are available for any sanitary filters. There is
a limited amount of particulate and lead emissions data for these filters
which indicate that they can achieve comparable particulate emission
concentrations as the process fabric filters (3.6 - 24 vg/m ). ' In
addition, the lead concentration in the particulate is approximately the
same.
One study of a secondary lead smelter was performed which compared lead
emissions from process, process fugitive, and area sources. This study
showed that lead emissions from the sanitary or process fugitive fabric
filter were approximately 74 percent of the lead emissions from the fabric
filter controlling process emissions. Using this study as a basis, the
assumed arsenic emissions from controlled process fugitive sources were
calculated. Table 6-9 presents the calculated arsenic emissions factors for
process, process fugitive, and area sources for a small secondary lead
smelter. Details of the calculations used to derive these figures are given
in Appendix A-6.
6.2.2.3 Nonprocess or Area Fugitives. As discussed previously, there
are many nonprocess fugitive sources at a typical secondary lead smelter,
the majority of which are currently uncontrolled. Many States are
developing State implementation plans (SIPs) to meet the NAAQS for ambient
lead. These SIPs will require control of many of the area sources in
secondary lead smelters. The major source of fugitive lead and arsenic
emissions from a secondary lead smelter is the flue dust storage pile in the
charge preparation area. These dusts which are captured by the process and
sanitary filters are often stored out in the open until they can be
recharged to the reverberatory furnace. The material is fine and easily
entrained by the wind. A study has been made of lead fugitive sources
around a secondary lead smelter. Table 6-10 reports the results of this
study. The table shows that the charge preparation area is the largest
contributor to nonprocess fugitive lead emissions, and that the battery
breaking area is also an important source. No arsenic emissions test data
are available for nonprocess fugitive sources of secondary lead smelters.
Arsenic emissions factors for area sources were estimated in the same manner
as for process fugitive sources. A comparison of the uncontrolled area
sources to the controlled process sources showed that the nonprocess
fugitive lead emissions were 1.7 times larger than the controlled process
114
-------
TABLE 6-9. ARSENIC EMISSIONS FACTORS FOR A SMALL SECONDARY LEAD SMELTER
SOURCE
PROCESS
PROCESS FUGITIVE
AREA
Emission Point
Control Device
Blast & Reverberatory
Furnaces
Fabric Filter
Furnace Charging
& Tapping
Fabric Filter
Material Storage Areas
& Haul Roads
None
Emission Factor
g/kg lead produced 0.036
Ib/ton lead produced Q.012
0.028
0.054
0..06
0.12
115
-------
TABLE 6-10. NONPROCESS FUGITIVE LEAD EMISSIONS AT A
TYPICAL LARGE LEAD SMELTER
EMISSIONS SOURCE LEAD EMISSION RATE g/hr
Battery Storage Areas
Battery Breaking Yard
Charge Makeup Area
Slag Storage Area
Smelter Access Road (workday)
cl
Building Fugitives
106
260
400
235
41
228
8.3
20.5
31.5
18.5
3.2
18
TOTAL Nonprocess Fugitives 1,270 100%
116
-------
lead emissions. This factor was then used to estimate nonprocess fugitive
arsenic emission rates for a model smelter.
6.3 DEFINITION OF EEC OPTIONS
This section discusses the estimated best control technology (EEC) for
control of arsenic emissions from secondary lead smelters. The section will
be divided into three parts which will discuss the availability of EEC
options for process, process fugitive and nonprocess fugitive sources in
turn. Each section will address, where applicable, the arsenic removal
potential of the identified EEC option, the technical feasibility of EEC and
identify existing plants currently using the EEC option.
6.3.1 Process Sources
Blast and reverberatory furnaces are currently controlled by either
fabric filters or fabric filters followed by scrubbers. In Section 6.2 it
was shown that the fabric filter/scrubber combination had lower lead
emissions than the fabric filter alone. In addition, several studies were
discussed which showed that it was reasonable to assume that lead behavior
could be used as an indication of arsenic behavior. For the purpose of this
study the combination system of fabric filter followed by a scrubber will be
chosen as the EEC option for further study. Using the data presented in
Section 6.2 the estimated incremental arsenic removal efficiency of the
combination system over the fabric filter alone is 59 percent.
There are no technological problems associated with following a fabric
filter with a scrubber. Five secondary lead smelters currently use
scrubbers after their fabric filters to control SO- emissions from their
smelting furnaces. Retrofiting existing secondary lead smelters with a
scrubber or scrubbers may be difficult because of space limitations at the
plant. These problems are not insurmountable but may increase the costs of
applying a scrubber over the costs expected for control of new sources. Use
of a scrubber will result in a wastewater treatment problem in addition to a
solid disposal problem. Any scrubber liquor which is blown down will have
to be treated and disposed of. Existing plants currently have to treat any
acidic liquid waste derived from the battery crushing operation or battery
storage areas. However, the treatment system for these acidic wastes may
not be applicable to the scrubber blowdown.
Existing plants currently recycle all captured particulate matter to
the smelting furnaces. It may be possible to do the same with the scrubber
generated sludges if the material can be dewatered and if the scrubber is
not used for S0? control. If the scrubber is used for joint control of SO
and arsenic then large quantities of calcium sulfite and sulfate sludges
will be generated. These materials cannot be recycled to the smelting
furnaces, and must be disposed of. Several existing plants are currently
landfilling this material on site.
117
-------
6.3.2 Process Fugitives
All existing process fugitive sources of secondary lead smelters are
controlled by fabric filters operating at low temperatures [below 37°C
(100°F)]. Further cooling of these streams to increase arsenic removal
efficiency is not feasible. For this reason EEC is assumed to be in place.
6.3.3 Nonprocess Fugitives
The majority of nonprocess fugitive emission sources are currently
uncontrolled. However, because of the NAAQS for lead, most States are
revising their STPs for lead, and it is anticipated that these revisions
will require tight control of all nonprocess fugitive sources at: secondary
lead smelters. The controls which will have.to be implemented to achieve a
boundary line lead concentration of 1.5 yg/m will be EEC for lead and for
arsenic.
6.4 DETERMINATION OF THE INCREMENTAL IMPACTS OF EEC
The purpose of this section is to discuss the incremental impacts
associated with applying the EEC options identified in Section 6.3 to
typical smelters in secondary lead industry. The section is divided into
four parts. These are model plants, incremental impacts, population
exposure assessment, and economic impacts.
6.4.1 Model Plants
Three model secondary lead smelters have been developed in this study.
These are described in Tables 6-11 and 6-12. These model plants will be
used in the ambient modeling associated with the exposure analysis, and as
the basis for determining the potential impacts of applying EEC options to
any of the sources considered. Model plants were used in this study because
it was not feasible to individually analyze all 66 secondary lead smelters
in the U.S.
As can be seen from the tables, all three source types, process,
process fugitive, and area fugitives will be considered in the exposure
analysis. However, only the process sources will be considered for the
incremental impact analysis for EEC since the two fugitive sources will have
EEC in place when they come into compliance with all existing regulations.
The model plant parameters presented in the tables were derived from a
survey of the NEDS data, telephone contacts, and from trip reports made as
part of the NSPS revisions for secondary lead smelters. Arsenic emissions
estimates presented in Table 6-12 are based solely on calculations and
engineering judgement and therefore should be used with caution. Appendix A
to this report discusses the assumptions used to derive the emission factors
presented. As can be seen in the appendix, the emission rates used are
sensitive to any changes in the assumptions used. For example, changing the
118
-------
CJ
to
teJ
ft.
CJ
> u
ft! Z
Q M
pj ft!
C cn
OS <
H EC
CJ
O
z
M
cn
X
u
z
C fcl
1- CJ
cn os
cn o
£ C
ft. C
C M
f-
cn -c
3 t»
O &.
S O
,-,
"a.
z u
C JZ
t— ' ^.
u x
o
o
cs
a.
c /—
ft} C.
J u
u
p*j
_W3
s
£— 1
3
0.
J_,
XI ft.
a
U I-H
cn oo
ft. cn e
W-l
01
a
ft. ft. O
01 01
00 00
c c
cn cn
•o
CO .-1
ft. ft. 1-H
V4
CU CU iJ
--( r-i C
00 CO O
C C
U U M
a. c- <
0
O
o
sD
^
-<
•
""
0
o
o
o
^
0
o
»~*
E
cn
p
01
J2 ft.
^ fti
k. 01
CJ iH
cn co
~~ c -o
ft. ^H 01
ft. cn c
U-l
^3
ft. ft. O
ft. ft. Z
01 01
00 00
c c
•H ^H
cn cn
•o
01
ft. ft. rH
ft. ft. O
^1
CU CU 4J
00 00 O
C C CJ
•r4 *r4 C
cn cn 3
01
•H
cn x jj
cn cn -H
CU 01 00
u u a eg
o o ft. cu
V4 1-1 14
04 PU <
0
O
o
s£>
CO
c
<*>
Q
O
O
CN
O
00
•.
03
a
3
•H
"S
*-
cn
u.
cu ft.
^
3 0)
U 1-4
u ec
cn c
ft. cn
ft.
CN
ft.
ft.
cn o>
"ft* 00
ft. C
CN cn
ft.
ft.
cn cu
ft. OC
ft. c
CN cn
cu
-H
cn cn u
cn x -H
cu cu oc
CJ CJ 3
O O ft.
OH a.
o
o
o
V4.T
vo'
•
vO
/— .
O
o
o
0
o
VD
•
*
-------
CO
Ed
Ed
i~J
*H
|
S3
o
^/^
§
1-i-l
^-^
2
fX
Ed
c
o
•
i-H
1
w
PQ
^*
e/i
o
t— t
t/3
(— *
g
y
z
u
u
U3
M
01
a
<
c-
u
<
so
-
y.
u *~l
u
J=
00
l-l
0) "^
c ,0
0)
to w
Clj r*
cc --*
00
^^
tn
x 0-
4J U-l
U
0 0
i-4 0)
o en
> "^
«£
*J U
10 a
os ^
g 8
-4 10
en
e
E —
E- o
4J
**^ ^
X
u
u
0) x—
AJ *J
O u-i
Ea ^^
Q
JJ ^*
.C 4J
TH ^-^
E e
c
O cfl
O O o
m so
^xif.
O -H
CT O
CM CM
X™" X>-V
O O
§§
O 0 r-P
es u-i *J
O
-a- o
0^ en ,0
CN en
eM
E
m in -*ff
***• *
en
-.CM 0)
m o u
m en tj
1
CO
QJ
l-i
X~v **• * -O
§-H 03
u u
00 60 -H
fl 3 W
•° ^ oo
« tn 3
0 0)
U U 03
O O 0)
0. t- <
01
rH
CO
S
en
^
a
CO Csl
O —
*J"^ 1
3 -H W
O W OJ
j: -H >
00 00 H-t
CO 3 . X-N
000
800 ^
O O CM
sO sO O "-<
m m co
o
0
oo in
• "^
^-i i—i m ^-s
e
x~^ x~v x*^. O
co cc OD -a-
•>-' w en
m m CM o)
in m o u
m en en V-i
3
O
CO
OJ
/**. /»>. x~> ^
>T -a* in
en
CM CM in
m m o
oo oo m
•
CM CM *— «
X~N S~*.
— H CM
tn
0) 0) 0)
tn en >
3 3 -H X
O O U 0>
fQ CO 3 *J
00
cn tn to 3
to tn en <*-»
0) 0) QJ
O O O nj
O O O 0)
Cu CU D- <
00
cu
oc
1-1
CO
en
•
o
»
cc
o
^
a.
J3 t4
CU
T3
CD 01
LI 00
O 1-
i-H ffl
i-H J
o y
U-l
•o
M C
rH 0)
-H 00
U-l Cfl
0 O
t-t tn
U-4 O
CQ VM
w wo
0) -H -U
4J CC
^H OJ «H
M-i U OO
U 3 M
-H O O
M tn
jQ Q
CO W *J
y-t cn
(0 • O OJ
QJ o e
i-t (0 CU W
r-4 &, J-l TO
0 0
P e u-i cfl
4J -H 0)
c tn >
U rH O 4J
•H e o M-I
r-. O 0
0> U r - A
tn w a*
CO 2 W <
£
j_i
—
00
c
4J
flj
u
a.
0
o
o
o
^4
^
•^
u
3
•0
a.
Px
y)
O
0
o
0
««
1
c
to
p— t
cq
(u"
^
f*_
00
c
«H
U
M
QJ
Q.
O
0
O
O
.K
CO
*-»^
TJ
QJ
*T3
O
Cu
Ifi
C
u
O
O
o
0*
1
*J
TO
i—4
E
3
•H
t!)
JF
01
S
^
i
*
V
f.
00
g
•H
u
a
a
a.
o
tn
w
3
r~
DC
a
.0
w
0)
0
a.
5
^
V-<
•^
.00
c
u
«
0)
a.
O
o
o
o
!>-<
^
•u
o
3
•a
a
D-i
w
o
o
o
o
•.
r-H
cc
}-<
3
a.
QJ
QJ
4J
o
Q;
o
d.
3
^j
l_,
U-
120
-------
assumed lead to arsenic ratio from 13:1 to 110:1 will change the emissions
by a factor of eight.
Process fugitive emissions and nonprocess emissions are based on the
ratio of lead emissions from these sources to the amount of lead emitted
from process sources for a single plant. These arsenic emissions estimates
are therefore even more tenuous than the estimates of process emissions.
6.4.2 Incremental Impact of EEC Options for Process Emissions
As discussed previously the blast and reverberatory furnaces in a
typical lead smelter are currently controlled by an afterburner, cooler and
fabric filter or cooler and fabric filter to meet the applicable NSPS or SIP
particulate regulation. The EEC option to be evaluated involves following
the existing fabric filter by a scrubber. The incremental impacts of the
EEC option can therefore be determined by looking at the impacts associated
with the scrubber alone.
The capital and operating costs associated with the use of a scrubber
are shown in Table 6-13. The table shows that the capital costs will range
from $209,000 to $594,000 depending upon plant size, and the annualized
costs will range from $127,000 to $371,000. These estimates were derived
from cost data presented by MRI in their project conclusions report for the
secondary lead NSPS revision. Details of the derivation of the costs used
in this study are given in Appendix B-6. All costs have an accuracy of
±40 percent at best.
The energy and water consumption of the scrubber system is also given
in Table 6-13 for each model plant size. The table also shows the estimates
of the amounts of solid waste that would be generated by a water scrubber
used to control arsenic emissions. If the scrubber were installed for joint
S0» and arsenic control then the quantities of arsenir containing solid
waste would be much greater.
The nationwide impacts of applying the EEC option to all existing
smelters are shown in Table 6-14. The table shows that an arsenic emissions
reduction of 22.4 Mg (24.6 tons)/yr could be achieved by applying the EEC
options.
6.4.3 Economic Impact of EEC
An economic impact analysis was conducted for the secondary lead
smelting plants to estimate the potential impact of requiring EEC
technology. Using the cost data in Table 6-13 and the methodology given in
Appendix C, revenue and product price increases, required by smelters to
maintain the same net present values before and after the installation of
EEC, were estimated. To test the sensitivity of the economic impact results
to the various input data, a sensitivity analysis was conducted for the
secondary lead model plants, assuming simultaneous 15 percent decreases in
121
-------
TABLE 6-13. CAPITAL AND OPERATING COSTS FOR CONTROL OF ARSENIC
EMISSIONS FROM SECONDARY LEAD PROCESS SOURCES
PLANT SIZE
Flow m /sec (dscfm) 9
Total Installed Cost
Annualized Capital
Cost @ .131473
Operating Labor
General Maintenance
(3 11% of TIC
Electricity Costs for
fan and pump
Water
Overhead
Small
.64 (20,000)
$ 209,000
27,477
18,100
22,990
31,000
425
18,200
Property tax and insurance
(§ 4% 8,360
Total Annualized Cost
Arsenic Removal kg/yr
(pounds/year)
Energy Consumption Mj/yr
(10 kWh)
Water Consumption
10 M /yr (M gal/yr)
Solid Waste Generation
Mg/yr (TPY)
$127,000
191 (420)
2,808 (780)
6.4 (1.7)
20 (22)
Medium
17 (36,000)
$ 297,000
39,047
18,100
32,670
56,000
2,150
25,862
11,880
$186,000
327 (720)
5,011 (1,392)
11.7 (3.1)
34 (37.5)
Large
34 (72,000)
$ 594,000
78,093
36,200
65,340
112,000
4,300
51,724
23,760
$371,000
655 (1,440)
10,000 (2,778)
23.5 (6.2)
68 (75)
10% interest for 15 years.
122
-------
TABLE 6-14. NATIONWIDE IMPACTS OF APPLYING EBC TO ALL
PROCESS SOURCES3
Total Capital Cost $ $21,600,000
Total Annualized Cost $ 13,400,000
Arsenic Removal Mg/yr 22.4
Energy Consumption Mj/yr 338,000
Water 103M3/yr 790
Solid Waste Mg/yr 2,300
aBased upon 18 large plants, 12 medium plants and 35 small plants.
123
-------
baseline price and annual output and a 15 percent increase in the weighted
cost of capital.
The annual revenue increases required by the secondary lead smelter
model plants to maintain their profitability range from $127,000 in the
small model plant to $370,000 in the large model plant. These additional
revenues would require product price increases of between 10c and 15c per kg
(4.5 - 6.8c per Ib). Percentage product price increases ranging from 1.1 to
1.6 percent would be required to maintain profitability. The revenue and
price impacts for the secondary lead model plants are summarized in
Table 6-15.
Under the conditions of the sensitivity analysis the small model plant
would require a unit price increase of 17c per kg (7.7c per Ib), while the
medium and large plants would require a 12c per kg (5.5C per Ib) increase.
Percentage price increases in the secondary lead models rose to 2.2 percent
in the small model plant and 1.6 percent in the medium and large model
plants. The sensitivity analysis results are summarized in Table 6-16.
6.5 POPULATION EXPOSURE DATA
The data in Tables 6-12 and 6-17 were developed during this study for
use in an air pollution dispersion model to estimate the population exposure
to arsenic from secondary lead smelting plants. The Strategies and Air
Standards Division (SASD) of EPA is using these data in an ongoing effort to
estimate exposure to arsenic from this source category.
6.6 EASE OF STANDARDS DEVELOPMENT
The biggest impediment to developing standards to limit arsenic
emissions from secondary lead smelters is the lack of actual arsenic test
data for all of the sources at a typical smelter. As discussed in this
chapter, the arsenic emissions estimates are sensitive to changes in any of
the assumptions. In addition, arsenic emissions from any two smelters
cannot be expected to be the same because of the wide variations in the
arsenic contents of the feed materials and the variations in operating
practices. The only way to address this problem adequately would be to test
several smelters operating at different conditions and utilizing different
feed materials.
An additional problem involves the control of area fugitive sources
around secondary lead plants. Emissions from these sources are difficult to
quantify, the current status of control is difficult to determine and the
level of control which will be required by the revised SIPs for the lead
NAAQS is hard to assess.
124
-------
TABLE 6-15. REVENUE AND PRICE IMPACTS FOR SECONDARY LEAD
SMELTERS TO ACHIEVE EBC
MODEL REQUIRED REVENUE
PLANT INCREASE
(000 dollars)
REQUIRED UNIT
PRICE INCREASE
(dollars per kg)
REQUIRED PERCENT
UNIT PRICE INCREASE
Small
126.9
.014
1.6
Medium
Large
185.2
370.4
.010
.010
1.1
1.1
125
-------
TABLE 6-16. SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE
INCREASES FOR SECONDARY LEAD SMELTERS
MODEL REQUIRED REVENUE
PLANT INCREASE
(000 dollars)
REQUIRED UNIT
PRICE INCREASE
(dollars per kg)
REQUIRED PERCENT
UNIT PRICE INCREASE
Small
130.0
.017
2.2
Medium
Large
189.6
379.2
.012
.012
1.6
1.6
126
-------
CO
en
S
H
d
CO i-l
CO 4-> O
O -H CQ
O 60 W
S-< 3
O TTI
I-M HU
CJ
P3
W
CO
ceo
cu cu
o d
O i-l
M iH
P-i 0)
CO
CO
CO
xi
4-1
d
CO
i-i
(X, 0)
N
i— i *r4
cu en
T3
0
S
r-l
CO s~**
o B
•H X
4-> ^— '
p
0)
>
CO
0) r-l
*O +J Cfl
•H Cfl 4J
1-1 d C ^
O 1-1 OB
T3 N Jd
s i-i i-i ^
HO l-i
P O O
cj EC
0)
d
0
N
d
o
•H
4-1
Cfl
O
O
4J
S
CO
t— i
PH
XXXXXXX XXX
XXXXXXX XXX
XXX
XX X X X X X
B B B
3CV)0)r-li-lr-l3 1-130)
•H 60 60 r-l i— 1 i— 1 i-l r-l -H 60
*^3 M V^ CQ cO CO *O cfl ^ }-i
CDcflCflSSBJ S OJ IB
SJ^jencnenS cn g j
r^ io ov >^ CTI m u-\
• • • • • • •
r^r^ ^CN «3-
«• •• •••»
m\o ON^-I r— r-~ i— i cr\
»d"cri r».r^ oomio
u~im mst «q- m >* co
vO^O i— (•— i •— ( m r^ P^
r— li — t i— 1 i— 1 1— 4 f— t i — |r«l
o
d
M <
CJ
M *»^ ^**
*> ^ -<5 •
d * cj , CJ Jj ,J
CJCX t-i- « OO Ofa'
B 4J o r-i o « d oo
•o o dend cod o -H -e
CO CJ O 9 M 4J-H CJ 4J CDCUM
0) i-(-a--Hd,J
a) ofa
•M rJ < o^-Mcfl tni-i co d- d • S c -H
CO" Q.O C C 0)co co» (-1 -HOI Sea fa o coy-i(U TSW i=>co
COT3 )-!>> O>\ST3 O *C dd)> CU^ii O,
(Jd) a)O U4J S-i 42d CJC T3M OOdd dO 0)6
0)0) -a>j i-l Ocfl 4-itO OicO i-HCl) -H C1J .HCO J3CO
4->,j dH oscj cjcj sen <3M 3> -Muaa ^n OH
d(0cn_aocnoco a>--i
McneiS-^cn<3cjz neJcj
127
-------
CO
to
H
H
co
j
O
£H
J2
O
CJ
Q
^
to
Z
o
!___-[
H
CJ
3
H
J
cu
O
Z
H
CO
W
X
W
/_N
-a
CU
3
C
•H
4_)
C
o
CJ
pj
rH
1
W
rJ
H
co
4-1
- - |
*r*
o
eu
c
o
^4
co
W
•H
t!
14H
O
CO
3
4-1
Cfl
+J
CO
rH
0
^4
4-1
C
0
CJ
CJ
M
co
0)
•H 00
CO 4J C
0) i-l -H
>H 00 +•>
< 3 CO
PM *T"4
U
CU
co >
CO -H
CU P CJ
O -H CQ
O 00 W
rl 3
c^ r*j<
CJ
CO
w
co
coo
CU CU
0 C
O i-l
IH rH
OH CU
CO
CO
PQ
JJ
c
cfl
rH
PH CU
N
rH 1-1
CU CO
"o
X
cfl /-^
0 g
1-1 ^
4-1 >-'
cu
p,
CO
CU rH
•O 4-1 Cfl
•H Cfl 4->
rl C C ^-v
O 1-1 OB
•a N ^,
H O IH
» 0 O
cj a
CU
c
o
c
o
•H
4-1
CO
O
o
^J
c
CO
1— 1
OH
X X
X X
CU
rH
n
cfl
O
IH
i-H
o, x
Cu
4-1
O
25
rH rH
rH rH
CO Cfl
B B
to to
CN 00
• •
r> in
r-^ r-
vO ON
*^" **3"
-* CN
CN rH
CO ^*
PI ~a-
P*» in
rH rH
O
u
oo en 93
CM 01
IH X 1-1 •
4-1 lj ^
rH - /W rl
CU -M /CO Cfl
0 "rl 3 PH
CO O T3
rl C CO
rH 4J M 1-1
Cfl CU 3
•rl p OH O
4J O
CO O •
3 ^Cj *•*
*X3 CK CO
c <:
M H
X X
X X
X X
e
3 rH
•H l-H
TJ Cfl
cu e
X co
-3- rH
• •
H
CU IJH • IH O
^ O rH
O. OA CO rH b
O J=
0 CJ
XXX
XXX
XXX
e
CU 3 rH
00 -H i-H
M -a co
cfl cu e
KH X CO
CN rH
• • •
rH rH
r*^ oo
in <±
^T ^
-H CS|
m m
m \ cfl
CU 4-1 CQ
X i-l • W
CJ CJ Z >•>
rH C S
i-H 4-1 HH » CU
•H to tfl 06
^6 CU « rn
>. V- TJ cfl O
3 O rH B CJ
CJ O CU
to e> a
r— 3
rZ:
*
3
3
co
c
3
pa
3
cu
z
X
X
cu
cu
rH
Cfl
O
•H
rH
CX
, Ou
4-1
O
z
rH
rH
cfl
e
to
o
ON
o
m
^
oo
rH
r^
m
00
rH
co
0
rH
rH
rH J52
CO
4-1 *
CU ^
X K
CO
X 3
cu cu
CQ Z
CO
W
X
X
CU
1— 1
.0
cfl
CJ
•H
rH
Cu
Cu
^
4_)
O
z
i-H
i-H
CO
CO
C
O
•H
CO
•H
^
•H
Q
CO >~i
r^i Z
0
rH •
rH .^
-------
to
CO
£
IS
CO
nJ
o
04
EH
Z
O
CJ
Q
^
CO
Z
O
1-,-J
H
U
O
hJ
§
*^
53
CJ
Z
h-4
H
CO
M
X
W
•o
CU
3
£3
•H
4J
C
O
CJ
~**-s
^
1
a
H
CO
4J
.s
o
PL.
d
o
,, 1
*n
CO
CO
jj
<4-4
0
co
3
4-1
CO
4-1
CO
r-4
O
14
4J
C
o
CJ
o
ca
co w
(U
•H 00
Cfl 4-1 C
CU >H -H
(-1 00 4-1
<3 3 CO
Ctj *rH
X
w
0)
co >
CO i-l
CU 4J CJ
O i-l CO
O 00 W
0, ta
CJ
ca
u
co
co a
CU CU
o d
0 -H
VJ FH
Pi CU
CO
tO
pa
.a
d
tO
rH
0< CU
N
i-l -rt
CU CO
o
a
cr
CU
•a 4-
•H n
CJ T
S "£
H C
1-0 C
CJ
i— t
tO /^
•H
XJ ^^X
M
cu
^
to
4-1
o e
N ^
1-1 ^^
V-i
o
EC
cu
O
Plant Location
X
X
X
cu
00
(-1
CO
hJ
o
CJ
•«
to
cu
I-J
co
to •
O to
M-l C8
3
O
X X X X X
X X X X X
X X X X X
e e B e
3 3 3 %
CO -i • • cu 3 a co
*J tO tO 3 O 3 CU
(04-1 Oj 4-1 4-> (U ^3 P-i 4J
3d at G con Tie oi-H
•^a) O<0 (-(tO 1-13 OC
C--" CJr-l CUT? M i-l CJtO
1-44-1 «$4J "OCU OO -i
-------
i
CO
H
H
CO
o
Pi
H
Z
O
o
^
C/3
r<3
o
H
U
O
tJ
EH
CO -H
CU 4J
0 -H
0 00
r^ £3
ft... pTj
CO
o
CO
td
•4
J
00
c
•H
4-1
CO
•H
^
W
£_}
«
M
CQ
W
cou
cu
a
o
>-<
O-i
ja
C
,_(
Pi
rH
CU
O
2
rH
U
•H
4-1
VJ
CU
Cfl
CU i-H
T3 4-> CO
•H Cfl 4-1
M C C
O -H O
•0 N
H O l-i
*H5 O O
U 33
0)
C
•H
rH
CU
CO
CO
M
cu
N
•H
CO
^
S
'^^
x™s
5
^ S
cu
C
0
c
o
•H
4J
CO
U
O
4-1
C
cfl
rH
Pi
X X X X X X X
X X X X X X X
X X X X X X X
rH CU CU i— 1 T— 1 CU r— I
rH 00 00 rH rH 00 rH
cfl M S-< CO cfl 5-i cfl
g to co e s cfl e
CO rJ rJ CO CO i_4 tO
CD f") O CT*
• • • •
CN O ^O v CJ O
pj i— i cu i-> <3 o
CO Z I-l M CU CO nJ
C M O - 4J rH >,
tfl OCOO +-1 cO«V-i
•H" T-l4-> ZCfl 4->CUCU
T3O OrH Z rH PQtO CUOO ^ •<
COOOOT3M biJgS-^rJ
MC04JCUrH »>> OCO
a cuco O- oo 6"'-|Pi M«>
C-H SC COO) C CUCUrH C
M J3 CU tfl ^— ' -i-H -H pi X; -H C rH -H
0)O 3-H CO O*-1 -1-1 ^O cflrH
J3 CXT3 OC CjT-l OCO !^4-l IJIW
4J. ic 03 Pi,d OrH sea cucu
Vl W CrfM CS CO
•rH r-l rH
^> CU OO
rJ E C
cu e w
1 co 3
o cu
CO 3
^
2
*
13
rH
CU
•H
M-J
00
C
•H
^4
O.
CO
•
3
X
X
cu
rH
c^
tfl
o
•H
rH
CX
cx
^J
4-1
o
rH
rH
Cfl
CO
s~\
C
s
o
rrj
4J
3
rC
CO
**-^
CO
>»
o
<— 1
rH
^
rH
CO
^
cu
13
CU
fa
M
53
M
4^
•H
O
j^
4->
CU
O
130
-------
0
t/S
H
H
,
O
2
H
Z
O
CJ
Q
*2
CO
Zj
o
H
O
o
p..
ry|
^
rJ
IX
o
z
H
M
td
^^
T3
CU
3
C
•H
4-J
c
o
CJ
•
,_4
I
vO
Cd
13
5
H
CJ
co pa
4-1 CO Cd
c cu
O H 60
to Cd 4J C
0> iH iH
e r< 60 4-»
O < 3 CO
•H to i-l
CO X
CO Cd
, J
Cd CU
CO >
14H CO 1-1
O <1) 4-1 CJ
y 1-1 oa
CO O 60 Cd
3 IH 3
•M (X to
td
1 1
*•
c/i cj
CO
rH Cd
O CO
IH coy
4J CU dl
c y c
O O -H
U S-l rH
ex cu
co
cd
CO
J*
e
rH
f\ t cU
N
rH -H
CU CO
•a
o
s
eg /->
iH £&
4-i V-J
).(
cu
CO
H
C3 0 O
cu
c
o
N
c
o
•H
4-1
cd
y
o
4-1
§
1— 1
X X X X X X X
X X X X X X X
CU
rH 0)
43 •-*
cd -a
cj td
•H y
rH 1-1
a. ^
o. a.
x < x x x x 3-
^
4-1
o *•*
z o
z
g
rH rH rH rH rH 3 rH
rH i— 1 rH rH rH iH rH
cd td cd td cd T3 td
g e s e e aj e
C/3 C/3 U3 W C/3 5G W
r^. ^- -i
CO rH O CJ 3! 33
• rHSjCU OZ »OOOO
>> S td Z S co C
O Z 4^ C/3^-1 »rJ«»rH«lH»
43 4)« ztdcucd-d4-i'o
J2S "SJG • CU4-t4JCrHC
4->< o >-t(o*i-J4-1 cued cued
CU H>-l yrH O4J O grH E*-1
(O 43 COlH -i-l^ U4J T3rH CU C/JCU
*J CdO rH O M T-I MlH P > >
• H S y 43 O 3 tdcd CUCU l-i 01
r4CU IO 3>H J3CU -IJ= 4JrH CUrH
to >~3^3 CXSO 4JQ rHO COCJ >CJ
. < CU O *H Cd *H
Cd CJ PS pd 3 S Dd
X
X
X
rH
rH
cd
e
C/3
ro
«^-
m
*3-
»
*^-
in
•
vO
oo
oo
»-H
•
O
CJ
^1
f_l
0) to
cd •
BO I-l
a)
IH 4->
CU CO
4-> Cd
co y
cd C3
y cd
cd
j
X
X
X
rH
rH
cd
g
C/)
CM
O
oo
«^
*
*^i
O
•
m
CO
-3-
00
i-H
O
CJ
60
C
•H
IH
3 <:
4J p [
y
cd •
«w c:
3 0
C -H
23 td
4-1
C C/3
fl
cu co
to C
O
cO
Cd
X
X
X
cu
bO
cd
(_3
r— 1
•
CM
CM
m
*
-vf
vo
•
LO
Csl
-------
CO
en
ro
H
H
to
, **
O
eA
H
Z
o
CJ
Q
•^
r
z
o
!_„ j
H
a
o
,j
H
§
j i
Cu
O
z
H
V)
M
X
w
/^
•a
0)
3
rj*
-rH
4J
(3
o
u
v^
P-.
rH
1
vO
W
•J
H
co
4.1
C
o
e
o
•H
CO
en
., j
e
M
"4-4
o
en
3
4-1
CO
rH
O
H
4-1
f^
O
C_3
U
CO
CO Cd
cu
•H 00
CO 4-1 C
0) -H -H
U 00 4J
•< 3 CO
pti -H
X
td
0)
en >
CO "r^
0) 40 O
O 1-1 CQ
O 00 W
VJ 3
(^ F^
u
OQ
W
03
COU
cu cu
0 C
O -H
>-l rH
P-l 0)
en
ca
-Q
+J
C
CO
T— 1
a, cu
N
rH -H
0) t/1
o
x
CO x-\
u e
4-1 v— '
J-l
0)
^
CO
CU rH
-a 4-1 co
•H CO 4-1
PC c /-.
CJ T-l OS
T3 N Ji
S P "rH >«^
HOP
3 O O
U 33
0)
C
o
N
C3
O
•H
4-1
CO
O
o
rJ
4-1
C
CO
rH
X X X X X X X X ^^
-<
Xs^ s^1 s^ s*^ s^ ^ K^ s^i K>
^» >"f KS KS KS t'S ?*< ?^t ?S
0)
rH
X A
CO
a
•H
rH
a
a,
"^
XX XXXXVX
4-1
0
^
e e
<— 1 •— 1 CU 3 rHCUrH3(UrH
rH rH 00 *H rH 00 rH «k«k»Kf««k
^^ *^ *vf CO CO CO CO CO CO
vO CO \O **H ^\ »— 1 CO 1^* CJ\
• •• ••••••
CNI-HI — o^Ou-ioooor--
ON CTi •— i COVO>-OOCOO
^f *^ *sf LO P** CNI (""•• "^3 t*^
oo oo oo ^o uo ^c* *^r *^ *^
r-(rHrH i — 1 -H — 4 V-H rH rH
a • a
t-i a. z a. !H •
•^ • ^ O P ^ H POO
flij O p i CJ O O * c j j2
U CJ 00 " U O M
x ecu c z >> c
CO i— 1 CO !-J ^ i-l>CO MH^ O •>
•H cfl -H 0) 000 CU'rHrH cfl )-l (-1
CCU 0) CQoQ CO* COCO) Stl3 COi — 1 CQ** O" 4-*<*
M'GrH'O CCQ(X -HOD 'H 4-i-H CO CUCO COCO
CO COCO rH-H O rH(3CU "OJ2 CU> I— ICO PCO 3CO
" r-i -Hi— 1 COTJ rHO CO'Hi— 1 CUCX SCO COrH OrH T3rH
-O-H M 1-1 PCfl Ol-l MlMrH CS CO PrH O rH CrH
rHjs oij: 0)0) •'-i.e cuoio -HCU coo cuco co MCO
3CL, O.fXi CpS PH CceiCJ U-iS COBS CQ Pin Q
OSOJcOO) CUOO)C/3t—3
132
-------
CJ
cfl
C/3
H
H
C/3
O
OS
H
Z
O
O
Q
^
C/3
Z
o
M
H
CJ
O
rJ
H
^
53
O
Z
M
H
M
X
w
x""s
"O
01
d
•H
4-1
d
0
CJ
*~*
•
r*«.
t-^
1
vO
w
rJ
CO
•^
H
CO
P
d
•i-i
o
ex*
d
o
•H
CO
co
•H
£2
pdj
IP
O
co
3
cO
co
f— i
O
^
4-*
d
o
CJ
pa
w
CO
aw
> M)
•H d
cfl 4J -H
CU i-H 4J
r-4 00 CO
< 3 -H
Ctj X
W
cu
co >
co 1-1 cj
cu 4J pa
O iH W
O 00
V4 3
CJ
W
co
coo
0) CU
o d
O i-l
I-l M
(Xt Ol
CO
cfl
CQ
^jQ
4J
d
cfl
M
CXi CU
N
M l-(
CU C/3
T3
O
S
M
cC ^
0 e
Xj ^^
^
CU
^
CO
CU M
-0 4J Cfl
•H cfl 4J
M d d X-N
a 1-1 os
T3 N H
2-4 ^ *^ ^t*'
HO >-J
boo
CJ 32
Q)
d
o
N
d
0
•H
4-1
cfl
O
O
rJ
4-1
2
cfl
M
(X,
X X X X X X
X X X X X X
CU CU
t— 4 M
•Q .^
cfl cfl
O CJ
•H i-l
r-l r— i
cx cx
XX ex x X o.
"^ ^
4-1 4-1
O O
s
0) M r-l 3 CU M
00 M M -H 00 M
^* Cfl Cfl 13 ^1 Cfl
cfl 6 6 Ol co 6
tJ co co S3 >J co
M on vc
• • •
ON *^" ON
\O in ^O
vO — 4 CN
»• « *
en -^- in
'd
*j
m o fo r»-
• *^ • •
CN 4-) f>. QO
O3 CX5 -*
l*» £Z CN m
co
-a- oo o
»— 1 t-4 i—l
CO CO
cu d
•H X 0
M H en
O 4J d M
CJ CO • w8 xtj O S
X 3 O > -H -lS W4J SCO
•COOCOCfl XO4J S
T3-H 4Jg T3C CO OCO di-l
i-i>-i coo dco cfli-i ai 5-H
3ta 333 COW B^osi O^W OS
O 33 10 03 Cd 0
1-4
O
u
4J
d
o
CJ
0
o
•H TJ
4J CO
3 0)
r-4 M
r-4
0 rJ
a o
U-l
^1
•H CO
CO O*
,
•» Jj ^J
00 O
d 01 P
•H O CO
4J C f-i
CO Cfl CU
•H • -H ,0
r— 1 OO r— 1 r-l
cx cx cu
c/3 oo e >
Q Z 0 CU
u a tJ
Z r-l
O P r-l
CO O O
co cu
CX, r-4 P
d M <4-i co
O CO CU CO
•H S-4 r— 1
•P 00 rQ
cfl d -P
•i-l -H O O
a 4J d d
O • CO
co >, i-t CQ ••>
CO CO X 0) >>
, O
o) d 42 co
•r-l O 0) CO
VJ P T3 > CU
P o> -H a
co • in i-i P co
3 >%CN . i-l i-l d
3 CO •— 1 >v 3 00 I-l
d -a i « cr 3 3
M -~~ O -d cu IM tp.
to in -»> i-i
T3 d CO cfl p
CO O '•• C r-4 0) O
CU P ki O O I-i CX
r-3 >-t P VJ Cfl
m ~~~ 4J ••>
4) CN CO O C IJ 0)
X r— 1 C in O O r— 1
4J O O >4-l 42
S IJ M 14-4 r-l O
O >-, O X O O -H
»-i ^. O ^ p i-l
'f-l CO O CO r-l 4J CX
d • d o) d cx
-a o in o > o cfl
41 p CM P 41 CJ
r-l | i-l P
•r-l O O O 14-4 O
a,«ooo P o c
S CO O O O CO
OP • • • .C T-I Cfl
CO CN •— 1 T- 1 > P
d P co cu cfl
o d -H T-I P
•HO co
4-> u n ii ii cu oo
2 . „ d d M
a >. s i-i i-i o
HCJ OJ3»-lr-44J V-l
od OO-HT-I oi co P
<4-tcu I-ITJCO co 1-1 d
doococue co x o
MCO iJSco oa w cj
fl A CJ T3 0)
133
-------
6.7 REFERENCES
1. Rathjen, J. A. Lead. In: Bureau of Mines Yearbook, 3 volumes,
Washington, D. C., Superintendent of Documents. 1980. 27 p.
2. Midwest Research Institute. Development of New Source Performance
Standards, Project Conclusions Report for Secondary Lead Smelting
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-3059.
January 27, 1981. pp. 8-11.
3. Letter and attachment from Bitler, J. A., General Battery Corporation,
to Miles, A. J., Radian Corporation. January 26, 1982. 7 p.
Information on arsenic emissions data from smelter stacks.
4. Prengaman, R. D. Reverberatory Furnace - Blast Furnace Smelting of
Battery Scrap at RSR. In" Lead-Zinc-Tin. 1980, Cigan, J. M.,
T. S. Mackey, and T. J. O'Keefe, editors. New York, Metallurgical
Society of AIME. December 1979. pp. 985-1001.
5. U. S. Environmental Protection Agency. Background Information for
Proposed New Source Performance Standards: Asphalt Concrete Plants,
Petroleum Refineries, Storage Vessels, Secondary Lead Smelters and
Refineries, Brass or Bronze Ingot Production Plants, Iron and Steel
Plants, Sewage Treatment Plants; Volume 2: Appendix: Summaries of
Test Data. Research Triangle Park, N. C. Publication No.
EPA-APTC-1352b. June 1973. pp. 37-43.
6. Danielson, J. A., comp. (Los Angeles County Air Pollution Control
District). Air Pollution Engineering Manual. Second Edition.
(Prepared for U. S. Environmental Protection Agency.) Research
Triangle Park, N. C. Publication No. EPA-AP-40. May 1973.
pp. 299-304.
7. Murph, D. B., and J. L. Pinkston. Current Blast Furnace Practice at
Murph Metals Southern Lead Company Smelter. New York, The
Metallurgical Society of AIME. TMS Paper Selection No. A70-41. 1970.
13 p.
8. Reference 2, pp. 24-25.
9. Midwest Research Institute. Development of New Source Performance
Standards, Project Recommendation Report for Secondary Lead Smelting
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-3059.
June 12, 1980.
134
-------
10. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for U. S. Environmental Protection Agency.) Research
Triangle Park, N. C. EPA Contract No. 68-02-3173. May 1982. 171 p.
11. Burruss, Robert P., Jr., and D. H. Sargent. (Versar, Inc.) Technical
and Microeconotnic Analysis of Arsenic and Its Compounds. (Prepared for
U. S. Environmental Protection Agency.) Washington, D. C. Publication
No. EPA-560/6-76-016. April 1976. pp. 103-107.
12. Letter from Peters, W. D., EPArPAB, to Pantalone, J. A., California Air
Resources Board, March 16, 1982, Request for information on secondary
lead arsenic emissions data.
13. Trip report. Maxwell, C. M., Midwest Research Institute, to
Telander, J., EPArlSB. May 19, 1980. 8 p. Report of April 19, 1980
visit to Sanders Lead Company in Troy, Alabama.
14. Trip report. Maxwell, C. M., Midwest Research Institute, to
Telander, J., EPArlSB. May 6, 1980. 109 p. Report of April 22, 1980
visit to Tonolli Corporation in Nesquehoning, Pennsylvania.
15. Trip report. Medepalli, K. S., Midwest Research Institute, to
Telander, J., EPArlSB. May 5, 1980. 13 p. Report of April 23, 1980
visit to General Battery Corporation in Reading, Pennsylvania.
16. Reference 5.
17. Smith, R. D. The Trace Element Chemistry of Coal During Combustion and
the Emissions from Coal-Fired Plants. • Progress in Energy and
Combustion Science. 6^(1) :53-119. 1980.
18. Davison, R. L., et al. Trace elements in Fly Arh; Dependence of
Concentration on particle Size. Environmental Science & Technology.
13(13): 1107-1113. December 1974.
19. Kaakinen, J. W., et al. Trace Element Behavior in Coal-Fired Power
Plant. Environmental Science and Technology. 9_(9):862-869. September
1975.
20. Klein, D. H., et al. Pathways of Thirty-seven Trace Elements Through
Coal-Fired Power Plant. Environmental Science and Technology.
9(10):973-979. October 1975.
135
-------
21. Schwitzgebel, K., G. S. Gutm, and M. A. Capalongan. (Radian
Corporation.) Fugitive Emissions at the Secondary Lead Smelter
Operated by General Battery Corporation in Reading, Pennsylvania.
(Prepared for U. S. Environmental Protection Agency, Region III.)
Philadelphia, Pennsylvania. EPA Contract No. 68-02-3513.
December 1981. p. 3.
22. Schwitzgebel, K., and R. Vandervort. (Radian Corporation.) Emissions
and Emission Controls at a Secondary Lead Smelter. (Prepared for U. S.
Environmental Protection Agency and U. S. Department of Health and
Human Services.) Cincinnati, Ohio. EPA Contract No. 68-03-2807.
January 9, 1981. p. 5.
23. Temple, P. G. , S. N. Linzon, and B. L. Chai. Contamination of
Vegetation and Soil by Arsenic Emissions from Secondary Lead Smelters.
Environmental Pollution. 12_(4) :311-320. April 1, 1977.
24. Roberts, T. M., et al. Lead Contamination around Secondary Smelters:
Estimation of Dispersal and Accumulation by Humans. Science.
_186_: 1120-1124. December 20, 1974.
25. Linzon, S. N., et al. Lead Contamination of Urban Soils and Vegetation
by Emissions from Secondary Lead Industries. Ontario Ministry of the
Environment, Toronto, Ontario, Canada. (Presented at the 68th Annual
Meeting of the Air Pollution Control Association. Boston.
June 15-20, 1975.) #75-18.2. 13 p.
26. Reference 2, p. 18.
27. Trip report. Mappes, Thomas E., PEDCo Environmental Inc., to
Beck, L. E., EPA:ISB. March 30, 1981. Report of March 18, 1981 visit
to Refined Metals Corporation in Memphis, Tennessee.
136
-------
7.0 COTTON GINS
This chapter discusses the cotton ginning industry and its associated
arsenic emissions. Section 7.1 presents a brief source category
description. Section 7.2 discusses the regulations currently covering the
industry and the resulting level of control and associated arsenic
emissions. Section 7.3 presents a discussion of the estimated best control
(EEC) options and Section 7.4 discusses the impacts of applying these
options. Section 7.5 presents the input data to be used in assessing the
population exposure to arsenic emissions from cotton ginning. The ease of
standards development is discussed in Section 7.6.
7.1 SOURCE DESCRIPTION
7.1.1 Arsenic Acid Desiccation
The only cotton gin operations of concern in this study are those which
handle and gin cotton that has been desiccated with orthoarsenic (arsenic)
acid (H-AsO,). Approximately 98 percent of all cotton that is treated with
arsenic acid is grown in Texas, with the remainder grown in Oklahoma. ' As
a result, certain cotton gins in these two States have the potential to emit
inorganic arsenic during^normal operations. For both States combined, an
average of 8.5 billion m_ (2.1 million acres) of cotton are treated with
arsenic acid every year. ' ' Estimates of the number of,cotton gins
processing the desiccated cotton range from 285 to 320. ' In Texas several
designated areas, including the Blacklands (which is the most significant
source), the Coastal Bend, and the Lower or Rolling Plains constitute the
major-sources of arsenic acid-desiccated cotton and arsenic emitting cotton
gins. ' ' ' The High Plains area of Texas generally relies on winter frost
freezes to perform the function of an arsenic acid., desiccant, and uses
comparatively minor amounts of arsenic acid. » » » » in some years a
killing frost may desiccate cotton in areas outside of the High Plains, but
this is not common and cannot be relied upon. The southwest corner of
Oklahoma is the primary desiccated cotton area for that state. '
Figure 7-1 depicts the main areas of arsenic acid usage.
The use of arsenic acid as a cotton desiccant was first begun in 1956.
It replaced pentachlorophenol (PCP) as the standard cotton desiccant. The
use of PCP declined because PCP manufacturers discovered that they had a
larger and more profitable market available in the area of wood preserving.
The shift of the PCP supply to the wood preservers, combined with the lower
cost and equivalent effectiveness of arsenic acid tended to gradually phase
out PCP use as a desiccant.
137
-------
A
8
C
D
Black!ands
Coastal Bend
Lower or Rolling Plains
High Plains
Figure 7-1. Primary areas of arsenic acid desiccation.
138
-------
In the non-irrigated areas of Texas and Oklahoma, a concept of cotton
production known as short-season cotton has evolved. Consistently low
cotton yields in these areas necessitated the development of a production
system that uses specially-bred, short-season varieties of cotton in which
the bolls mature early and at the same time so that once-over harvesting is
possible. The use of once-over harvesting required that an efficient
harvesting method known as mechanical stripping be developed. A mechanical
stripping operation essentially removes all parts of the cotton plant
including the leaves, burs, bolls, and side branches and leaves only a bare
stalk.
The quantity of trash generated by mechanical stripping has been
reported to range from 325 to 545 kg (715 to 1199 Ib) per bale of lint
cotton produced, with the average trash content being 360 kg (792 Ib). Of
the total amount of seed cotton harvested by stripping, approximately 35 to
40 percent will be trash. » » » > ^ range of trash content exists
because the level of trash and foreign matter in the seed cotton increases
as the harvest season progresses. Because the arsenic acid is associated
with the plant material, the more trash in the seed cotton, the greater the
potential level of inorganic arsenic emissions.
Today's varieties of short-season cotton have been specifically adapted
for stripper harvesting. However, a desiccant must be applied to
short-season cotton prior to stripper harvesting to dry out green plant
leaves to prevent fiber staining and unacceptable levels of fiber moisture
content. The addition of green leaves to the seed cotton increases fiber
moisture, thereby creating a condition in which the.cptton will heat during
storage and be lowered in quality prior to ginning. ' It is recommended
that the fiber moisture^e less than 12 percent, with an optimum range being
6.5 to 9.5 percent. ' ', Overdrying must be avoided, however, because lint
quality can be lowered.
7.1.2 Methods of Desiccant Application
Arsenic acid for cotton desiccation is always applied as a spray.
About 20 to 30 percent of the material is applied by aircraft and the rest
by ground sprayers. Both self-propelled, high-clearance machines and
tractor-mounted sprayers are used for ground application. Generally,
1.4 liters (3 pints) of arsenic acid are diluted to a final volume of about
37.9 liters (10 gal) of spray solution per acre. Where aircraft are used,
1.4 liters (3 pints) are applied in a total spray volume of 11.4 to
18.9 liters (3 to 5 gal) per acre. The active ingredient of the spray
solution does not dry out prior to leaf penetration because the droplets are
hygroscopic in nature. The arsenic acid solution desiccates the cotton
plant by disrupting the cell membranes such that natural drying of leaf and
stem tissues occurs.
139
-------
7.1.3 Factors Affecting Potential Inorganic Arsenic Emissions Prior to
Ginning
7.1.3.1 Occurrence of a Killing Frost. Several factors dictate the
need to desiccate stripper cotton. If the cotton growing regions receive a
killing frost [-2°C (28°F) or less] at the correct point in the harvest
season, no arsenic acid desiccation is needed. Unfortunately this situation
does not occur in the majority of stripper areas, and therefore, cannot be
relied on for desiccation purposes. The sales records of one arsenic acid
manufacturer to cotton harvesters showed a range of consumption from 2.5 to
6.2 million kg (5.5 to 13.7 million Ibs) during the period from 1964 to
1977. ' However, during this period sales to the Blacklands and Coastal
Bend regions showed little variability indicating that the Texas Plains and
Oklahoma growing regions experience the most killing frosts and account for
the considerable fluctuation witnessed in arsenic acid consumption over the
196A to 1977 period.
7.1.3.2 Arsenic Acid Shortages. Potential arsenic acid emissions from
cotton ginning may be reduced because of a shortage of the acid for
desiccation purposes in some or all of the cotton growing regions. The
shortage can be caused by the inability of the arsenic acid suppliers to
produce enough arsenic acid or by the suppliers preferential sale of the
acid to other consumers, primarily wood preservative manufacturers. An
example of the first type of shortage occurred in 1977 due to a labor strike
at the sole U.S. plant producing arsenic trioxide, the chief raw material
for arsenic acid. When this situation occurred cotton growers were forced
into using the only other registered desiccant, paraquat. The second type
of shortage occurs because arsenic acid manufacturers prefer to sell most of
their product to the wood preservative manufacturing industry. ' ~ They
prefer the wood preservative business because it represents a constant
demand unlike the cotton desiccation use which is seasonal in nature. One
arsenic acid manufacturer indicated that cotton growers actually only „
receive the product that cannot be sold to wood preservative companies.
7.1.3.3 Meteorological Conditions. The growth rate of the cotton
plant is the most important variable in determining how much arsenic acid is
needed for desiccation, and consequently may be emitted during ginning.
Plant growth is directly related to soil and meteorological conditions, such
as soil moisture, air temperature, and quantity of rainfall. If
temperatures have been warm and rainfall plentiful, the cotton plant will
grow to be very sturdy. As a result, desiccation is more difficult and
greater quantities of arsenic acid are required for the job. Multiple
applications may even be required, particularly when second-growth or
regrowth leaves appear on the plant. ' '
The ability of rainfall to remove arsenic acid residue from the cotton
plant after desiccation and prior to ginning,tends.,to greatly reduce
potential arsenic emissions from the gin. ' ' ' Laboratory analyses of
gin trash indicated that approximately 74 percent of the arsenic content was
140
-------
28
water extractable. Substantial work done by Miller, C.S., et al. (1975)
indicated that there is no relationship between the amount of desiccant
applied and the arsenic contents of gin trash. Through a thorough study of
13 gins over the length of a ginning season, no positive relationship was
indicated between the quantity of desiccant applied and the amount of
arsenic acid residue in the gin trash. The level of arsenic residue in
the gin trash was correlated to the amount of rainfall received, growing
conditions, and the type harvester used, and not by the total amount of
desiccant applied. '
7.1.3.4 Efficiency of Desiccant Application. The efficiency with
which the arsenic acid desiccant is applied to the cotton field directly
influences how much arsenic is left on the plants and potentially may be
emitted by the ginning process. The primary application technique, ground
spraying, has been estimated to be about 5 percent efficient. '
Application efficiency is defined to be the amount of arsenic acid that
eventually reaches or contacts plant leaves, stems etc. divided by the
amount directed at the plant.
29
The poor correlations determined in past studies , between the total
amount of arsenic acid sold and applied as a desiccant to the total amount
of arsenic found in the gin trash, can partially be explained by the
5 percent application efficiency estimation. By using the 5 percent
application efficiency assumption and the calculations presented in Appendix
A-7, the difference between the total arsenic (elemental) quantity actually
contacting the plant and the quantity found in the gin trash can be
estimated to vary by less than 1 percent. The quantity of material not
contacting the plant falls on the soil and can be dispersed in the ambient
air as windblown dust. As a result any ambient air sampling for arsenic
attributable to cotton gins could be biased high due to windblown dust from
the arsenic-contaminated soil in surrounding fields.
7.1.3.5 Translocation and Transformation of Arsenic Acid. Potential
inorganic arsenic emissions from a cotton gin (handling arsenic
acid-desiccated cotton) may be affected by the translocation of arsenic acid
into the main stalk and roots of the plant, and by the transformation of
inorganic arsenic acid to organic arsenic compounds. Studies on cotton
plants have demonstrated that once arsenic acid is sprayed on the leaves of
a plantgand absorbed, it can be translocated to other parts of the
plant. ' Some parts of the plant, primarily the stalk or main stem, are
usually left in the field after stripper harvesting. Arsenic acid which has
been translocated into these parts would not find its way to the gin where
it could be emitted. Work performed by Miller and Aboul-Ela (1965)
indicated that arsenic acid applied to the cotyledonary leaves of a cotton
plant moved into the petioles^!leaf stem) within one hour and into the main
plant stalk within two hours. Once in the main stalk, movement was
towards the plant root with no upward movement recorded. The application of
arsenic acid to the first foliar leaves of cotton plants resulted in little
or no translocation out of the leaf. Further laboratory work by Miller
141
-------
(1974) indicated that when arsenic acid was applied to the stemqof cotton
regrowth branches, 71 percent of the arsenic was translocated. However,
because no translocation data on commercial field cotton crops have been
determined, it is unreasonable to speculate on the extent that translocation
may be occurring in desiccated field plants.
Although definitive work has not been done for cotton plants, evidence
suggests that some percentage of the arsenic acid absorbed by a cotton plant
is converted to organic arsenic compounds. This type of natural
transformation has been demonstrated in grass plants. Twelve hours after
the application of a solution of sodium arsenite, trimethylarsine was
detected in grass samples. The cycle of inorganic arsenic transformation
to organic arsenic compounds-has-been shown to be a natural arsenic
detoxification mechanism. * ' In addition to trimethylarsine,
methylarsine, dimethylarsine, and methylarsonic and methylarsinic acids have
been observed as a result of natural detoxification. Figure 7-2 shows a
simplified version of the natural arsenic transformation cycle in the
environment.
No quantification has been made on the rate or extent of the
transformation in cotton plants. It is probable that not all of the arsenic
content measured in tests of gin trash (Durrenberger, Miller) was inorganic
arsenic. Additional sampling and analysis work needs to be done_on_gin,
trash samples to speciate between inorganic and organic arsenic. ' "
7.1.4 Inorganic Arsenic Content of Gin Trash
There have been two scientific studies that attempted to estimate the
arsenic content of gin trash that was produced from desiccated stripper
cotton. The first study by Durrenberger (1974) sampled trash from seven
gins and determined that the average arsenic content (as total elemental
arsenic) was 2000 ppm or 0.20 percent. The second study by Miller, et al.
(1975) measured the arsenic levels in gin trash at 13 gins and found that
the arsenic levels (as total arsenic) ranged from 50 to 450 ppm, and
averaged about 225 ppm (0.0225 percent). The most recent work in this
area by the U.S. EPA and the U.S. Department^ of Agriculture (USDA) concluded
that the Miller, et al. study used more precise analytical detection methods
to determine the average arsenic.,content in gin trash, and therefore was
more accurate than Durrenberger. Based on the recommendations made in the
EPA-USDA report, 225 ppm was used as the arsenic content of gin trash from
desiccated stripper cotton for the analyses in this document.
As stated, the arsenic values for both studies are expressed as total
elemental arsenic. No measurements were made on the relative split between
inorganic and organic arsenic in the trash. The quantity of inorganic
arsenic in the total arsenic measurement may be lowered due to the effects
of transformation as discussed in Section 7.1.3.5. Arsenic content results
expressed as total arsenic may be biased high by the arsenic contribution of
the organic arsenical herbicides DSMA and MSMA which are frequently applied
142
-------
CH3AsH2 ^ CH3As03H2
I
methylation
demethylation
H3As04
II
As-containing sugars;
other, yet unidentified
arsenic compounds;
As-containing lipids
(CH-),,As09H
O £ Cm
S"
V
II
(CH3)3As
[(CH3)3AsCH2CH2OH]V—
(CH3)3AsCH2COO"-<-
Figure 7-2. The transformation cycle of arsenic compounds in the environment
33
143
-------
to cotton. The potential for this type of bias reinforces the need for
speciation when arsenic analytical tests are conducted.
7.1.5 Alternatives to Arsenic Acid Desiccation
7.1.5.1 Paraquat. Besides arsenic acid, paraquat is the only other
desiccant registered for use on cotton. It too, however, is proposed for
Rebuttable Presumptions Against Registration.fRPAR) by EPA. Paraquat is not
as effective a desiccant as arsenic acid. ' In tests between the two,
even a 50 percent greater per acre use of paraquat was not as effective in
the desiccation of regrowth leaves as arsenic acid. Researchers have also
attempted to combine paraquat with other substances (primarily additives or
defoliants like endothall) to obtain desiccants equivalent to arsenic acid
in effectiveness. This combination, while an adequate desiccant, was not as
effective as arsenic acid alone. Endothall has also been blended with
arsenic acid in hopes of achieving synergistic desiccation effects, however,
the blend did not work as well as arsenic acid alone.
7.1.5.2 Use of Defoliants. Defoliants, wiltants, and_regrowth
inhibitors cannot be used as replacements for desiccants. ' ' Sodiunw
chlorate, cacodylic acid, Folex^ (tributyl phosphorotrithioite) and DEF^
(S,S,S-tributyl phosphorotrithidate) are the primary commercial defoliants
recommended for cotton. All of these substances with the exception of
sodium chlorate are candidates for RPAR. Defoliants will not perform the
function of desiccants, particularly when there are second-growth leaves
present at the time of application. Defoliation only removes the leaves,
it does not dry out or kill the plant.
7.1.5.3 Heat Treatments. Experimental work has been conducted using
intense heat treatments to desiccate cotton plants in preparation for
stripping. The heat treatments consumed approximately 37.9 liters (10 gal)
of liquid propane gas (LPGl per acre treated. Leaf desiccation resulted
from 2.5 J/cm (2.2 Btu/ft ) of heat. Ideal time-temperature exposure for
defoliation response was 850 degree seconds above 54°C (130°F). Presently,
no major cotton equipment manufacturers are producing such units.
7.1.5.4 Conclusion on the Alternatives to Arsenic Acid. Presently,
there are no chemical replacement or new desiccation techniques that perform
with the effectiveness of arsenic acid in preparing short-season cotton for
mechanical stripper harvesting. '
7.1.6 Characterization of Cotton Gins Handling Arsenic Acid-Desiccated
Cotton
7.1.6.1 Model Cotton Gins. As discussed in Section 7.1.1, about
320 separate cotton gins process arsenic acid-desiccated cotton. It was not
possible in this study to determine the process parameters and configuration
of every gin that may potentially be emitting inorganic arsenic. As a
result, to obtain a reasonable estimate of the inorganic arsenic emissions
144
-------
from the source category, model cotton gins were designed that are
representative of the operations and emissions of the gin population.
Four model cotton gins were developed based on the average number of
bales of lint cotton that could be produced on an hourly basis. The four
model gins produce 4, 7, 12, and 20 bales/hr of product lint cotton. A bale
of lint cotton was assumed to weigh 227 kg (500 Ib). Table 7-1 presents
the production data associated with each model gin. Of the approximately
320 gins handling arsenic acid-desiccated cotton, 10 percent or 32 gins gin
A bales/hr, 30 percent or 96 gins gin 7 bales/hr, 50 percent_or 160 gins gin
12 bales/hr, and 10 percent or 32 gins gin 20 bales/hr. ' The model
gins were specifically designed to process stripper-harvested cotton that
contains an average of 386 kg (850 Ib) of trash material per bale of lint
cotton produced. A flow diagram of the model gin equipment configuration is
given in Figure 7-3. ' ' All four model gins contain the same basic
cotton cleaning equipment.
The use of cyclones and particulate filter media for trash emissions
control, particularly at gins handling stripper-harvested cotton, is
accepted as a necessary and a routine part of gin operation due to the.large
quantity of particulate matter generated from input seed cotton trash. '
The particulate emissions containing inorganic arsenic from model gin
emission points 1 to 5, 9, and 10 (Figure 7-3) are collected by cyclones and
are routed to a bur hopper or dust house for storage. Emissions from model
gin emission points 6, 7, and 8 are controlled by covering the-condenser
drums at each of these points with a screen filter media. ' ' The sources
of arsenic-containing particulate emissions from the model gin are
characterized in the following section (7.1.6.2). A detailed discussion of
the model gin particulate emission collection and control equipment is given
in Section 7.2.2.
7.1.6.2 Potential Inorganic Arsenic Emission Points in the Cotton
Ginning Process. Because of the spraying method used to apply arsenic acid,
it appears that the entire cotton plant would be contaminated with arsenic.
However, the available data do not support this implication. As discussed
in Section 7.1.4, analytical tests conducted on gin trash, lint cotton, and
cotton seed from arsenic acid-desiccated cotton demonstrated that gin trash
was the predominant source of arsenic. Gin trash consists of leaves,
burs, sticks, and hulls. Arsenic?was not detected in the majority of tests
on lint cotton and cotton seed. ' ' '
The seed cotton which has been stripper-harvested in the field comes
into the gin containing lint cotton, cotton seed, and arsenic-contaminated
trash. Process steps at the gin clean and extract trash components from the
seed cotton, separate the lint cotton from the cotton seed (ginning), and
clean the separated lint cotton. These steps are performed in one of two
parts of the gin known as the "high pressure" and "low pressure" sections.
The sections are denoted as high and low pressure because of the relative
pressure against which each system must operate. For the high pressure
145
-------
co
H
,J
W
O
O
s
2
I— i
O
ss
o
H
O
CJ
w
35
H
S5
O
^J
§— I
^4
(—5
O
w
H
<3
h-3
W
p4
i
o
1— 1
^
o
Q
O
OH
•
r— 1
1
'""'
w
h-3
pa
H
4J
3
P. C
4= 0
60 4J ^-s
3 4-1 en
O O c
n o o
4—< 4J
H "G *— '
01
>> 01 00
r-( CO «g
^4
3 u-i
O O
33
^
4J
•HCO
a e
ctj O
P.-H
CO 4-1
CJ CO
N
01 -H
60 rH
cfl -H
^•J 4«J
01 p
_>
CO
C
O
U-l CO
O CO
0)
M CO
M
3 l-i
O 0)
K P.
0) C
60 O
cfl >rl
01 CO
^ ^
< QJ
a,
0
5
o> c
tvl O
•H 4J
CO 4-1
O
C 0
O 4-1
c
. • x-s
• • m CN
-------
c c
o o
o o
0 CJ
3) T3 T3 W
CO d)
-------
section this pressure ranges from 25.4 to 50.8 cm (10 to 20 in) of water,
while in the.low pressure section pressures of 2.5 to 12.7 cm (1 to 5 in)
are common. ' Potential inorganic arsenic emissions (from arsenic acid
desiccation) from both the high and low pressure systems will be
characterized in the following sections.
7.1.6.2.1 Cotton gin high pressure section. The high pressure section
of a cotton gin includes primarily the air discharges or exhausts associated
with the seed cotton cleaning equipment. ' ' Basically this includes
the gin cleaning equipment in place at the point where the seed cotton
enters the gin system through the point where ginning occurs. In gins
processing stripper-harvested cotton, cleaning equipment such as the
unloading separator, inclined cleaners, bur and stick machines, and
extractor feeders would be present in the gin high pressure section. The
arsenic-contaminated trash which these cleaning devices remove from the seed
cotton is pneumatically exhausted from the gin, thus making each of the seed
cotton cleaning stations a potential inorganic arsenic emissions source.
The location of these emission points within a typical gin structure are
shown in Figure 7-3. The high pressure section of a gin emits burs, leaf
and stick pieces, hulls, and large soil particles, and therefore, the
majority of arsenic that is emitted from the ginning process.
7.1.6.2.2 Cotton gin low pressure section. The cleaning equipment
included in the low pressure section of a gin consists primarily of the ,,
first and second lint cotton cleaners and the lint cotton condensers. '
The lint cleaners and condensers are sources of particulatg7and inorganic
arsenic emissions in the form of lint fly and cotton dust. The low
pressure system, however, is not a large source of arsenic emissions as
evidenced by tests on lint cotton and cotton seed for arsenic. For this
reason, control of low pressure section emissions for arsenic removal could
be very costly as illustrated in Section 7.3.1.2.
7.1.6.3 Sources of Fugitive Inorganic Arsenic Emissions from a Cotton
Gin. Fugitive particulate emissions, potentially containing inorganic
arsenic, are emitted from a cotton gin in addition to the gin process
emissions from the high and low pressure systems. Cotton gin fugitive
emissions may be in the forms of fine-leaf trash, bur material, lint fly, or
cotton dust depending on the source. The predominant sources of fugitives
are building and piping leaks, equipment leaks (e.g., cyclones), bur hopper
dumping, and wind blowing of open bur piles.
The level of fugitive emissions from cotton ginning has not been
quantified. No test data are available to accurately estimate the absolute
mass emission of fugitive particulate or the relative split between process
and fugitive gin emissions. Experts in the cotton ginning field have
estimated, however, that approximately 50 percent of the tgtgl particulate
emissions resulting from a gin are from fugitive sources. ' As the
collection system for gin process emissions becomes more efficient, the
relative percentage contribution of fugitive emissions becomes greater.
148
-------
Bur hopper dumping is generally the largest source contributing to gin
fugitive emissions. Most gins have the bur hopper elevated such that a
truck or trailer can be placed under the hopper doors. The doors can then
be opened as needed and the gin trash collected by the cyclones released
into the waiting trucks. The emissions from hopper dumping are intermittent
and the emission periods short in duration. The dumping of these large
quantities of particulate matter (1500 to 7500 kg/hr) into the haul truck or
trailer and their subsequent displacement causes fugitive emissions. Any
wind during dumping worsens the fugitive emissions problem. The fugitives
from the bur hopper dumping operation are mainly trash from the high
pressure section of the gin. Even though it has not been measured, the
arsenic content of the fugitive emissions is, for the purposes of this
study, assumed to be the same as the process emissions, 225 ppm by weight.
Other fugitive emission sources not directly related to the ginning
operation may contribute to the ambient concentrations of arsenic found in
the atmosphere around a gin. These sources are the mechanical stripper and
open cotton fields that have been harvested. When the mechanical strippers
are in operation they displace dust that may contain arsenic. Based on
the poor application efficiency of the arsenic acid desiccants
(Section 7.1.3.4), open, harvested fields could be significant sources of
windblown soil containing arsenic.
7.2 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING REGULATIONS
7.2.1 Regulatory Impacts
The only environmental regulations having an impact on inorganic
arsenic emissions from cotton gins are State particulate limitations.
However, because cotton gins are classified as agricultural sources they are
exempted from the general particulate regulations, and are usually given
separate, less stringent process weight-based limitations to follow. ' '
Gins in Oklahoma are also regulated under,. that State's general opacity laws
which limit source opacity to 20 percent.
Based on the wording of particulate regulations in both Texas and
Oklahoma and the comments of air quality officials in these states, it
appears that gin regulation is based primarily ..on.. "equipment standard"
and not on some process mass emission rate. ' ' Although both states do
periodically inspect gins, they do not perform mass emission tests to
determine compliance with the agricultural process weight regulation. Gins
in Oklahoma are inspected once a year, while Texas gins^are generally only
inspected when a public nuisance complaint is received. '
In terms of an equipment standard, both Texas and Oklahoma attempt to
get cotton gins to install various types of collection and control equipment
to reduce particulate emissions from high and low pressure gin exhausts. If
the gins install the necessary equipment, n then the State is basically
satisfied that the gin is in compliance. ' For example, in Oklahoma
149
-------
small diameter, high efficiency cyclones are required on the high pressure
section emission points and 100 mesh screen covering is required on the
condenser drums of low pressure section emission points. The level of
existing arsenic emissions is dictated by the performance of these required
control devices and not some process weight-based particulate limitation.
The cotton gin model plants in Section 7.2.2 were assumed to be equipped
with the devices required by Texas and Oklahoma.
Gins are not viewed as serious emission sources of either particulate
or arsenic matter by Texas or Oklahoma. ' Neither State pursues gins as
violators of the NAAQS for total suspended particulate matter even though
measured ambient concentrations around gins is frequently above the primary
and secondary standards. ' ' ' The States believe that the
agricultural status and seasonalitv of the gin sources does not warrant
consideration under the NAAQS. '
In the past, arsenic could be emitted from gin sources through the
process of burning gin trash for disposal. However, today neither Texas nor
Oklahoma allow the burning of gin trash, therefore, this source for arsenic
emissions no longer exists.
Occupational exposure regulations issued by OSHA do not affect the
control or reduction of inorganic arsenic emissions from cotton gins. The
OSHA standard for airborne inorganic arsenic, at 29 CFR 1910.1018
specifically excludes gins because they are classified as agricultural
sources. Standards that have been issued for cotton dust, which could
indirectly affect arsenic levels, either do not apply to cotton gins (29 CFR
1910.1043) or,.have been deleted from the books by the courts following suits
against OSHA.
The majority of gins handling arsenic acid-desiccated cotton in Texas
and Oklahoma are in compliance with all existing regulations as applied and
enforced in these States. Therefore, the existing degree of control is
equivalent to the baseline.
7.2.2 Baseline Inorganic Arsenic Emissions and Controls
The control devices used at a cotton gin to remove arsenic particulate
matter from various exhaust streams is the same equipment that is used to
remove any other gin trash because any arsenic leaving the gin will be
contained on trash material. By controlling gin particulate emissions,
arsenic emissions are also controlled. The types of particulate control
devices used in a cotton gin are different for high and low pressure section
emissions. Different devices are applied because the high pressure
section is emitting mostly large trash (burs, sticks, leaves, etc.)
potentially containing arsenic, while the low pressure section emits
primarily small particle size lint fly and cotton dust. The control
devices and emissions characteristic of each gin system will be discussed in
the following two sections.
150
-------
7.2.2.1 High Pressure Section Controls and Inorganic Arsenic
Emissions. The standard collection,or_cpntrol device for gin high pressure
section emissions is a cyclone. ' ' ' Cyclones are predominantly used
because they are effective, inexpensive, and require little maintenance.
The type of cyclone used at cotton gins, which is illustrated in Figure 7-4,
was developed by the Atomic Energy Commission (AEC) and is known as a small
diameter, high efficiency cyclone or a 2D-2D cyclone. The 2D-2D cyclone
is equal to baseline control for the high pressure section emissions.
Small diameter, 2D-2D cyclones have been shown to be very efficient on
the sizes of particles predominantly found at a cotton gin. Testing by
Baker and Stedronsky (1967) indicated that 99.2 percent of stripper cotton
trash had a particle size diameter greater than 74 microns. Tests
performed on the 2D-2D cyclone have shown it capable of reducing
99.9 percent,of all particulate emissions larger than 20 to 30 microns in
diameter. ' To allow for the fact that not all gins have properly sized
and operated cyclones, the efficiency of the 2D-2D cyclones used on the
model cotton gins (Section 7.1.6.1) was assumed to be 99 percent. The 2D-2D
cyclones are used to collect particulate matter exhausted from emission
points 1-5, 9, and 10 in Figure 7-3, which are basically the high pressure
emission points.
The baseline arsenic process emissions as estimated from the model gins
are presented in Table 7-2. An example calculation of how the process
emissions were calculated is given in Appendix A-7. The total arsenic (as
elemental arsenic) process emissions from the source category were estimated
to be 1.6 Mg (1.8 tons)/yr.
7.2.2.2 Low Pressure Section Controls and Inorganic Arsenic Emissions.
As discussed in Section 7.1.6.2.2, the gin low pressure section emissions
come from the primary and secondary lint cleaners and the battery lint
condenser-. Eighty percent or more of the gins in thi° source category use
coverings over the condenser drums atQeagh,of these three points to reduce
lint fly and cotton dust emissions. ' ' ' The condenser drum coverings
consist of either fine-mesh, stainless-steel screen wire or fine perforated
metal. As the air stream containing lint passes through the drum coverings
the lint is caught and eventually collected. A schematic diagram of a
typical covered condenser drum is given in Figure 7-5. For the lint fly
type particulate matter being emitted in this section at the gin, covered
condenser drums are about 40 percent efficient.
The other 20 percent of gins use in-line filters to control low
pressure systeiruparticulate emissions, however, the use of in-line filters
is declining. ' * The three most prevalent types of in-line filters are
depicted in Figure 7-6. Tests performed by the U. S. Department of
Agriculture (USDA) indicated that the filter types are equally efficient.
The average collection efficiency for an in-line filter was determined to be
approximately 81 percent.
151
-------
CYCLONE DESIGN PROPORTIONS
BC * DC/4
Hr = Dr/2
u w
DE - Dc/2
LC ' 2DC
Sc = Dc/8
ZC = 2DC
Jc = 0.3 m MINIMUM
3,
DC
Figure 7-4. Relative dimensions for a 2D-2D, small diameter design cyclone
152
-------
0}
CA
H
Z
PH
h-3
w
o
o
s
5Z
O
z;
o
E"^
o
i
CO
52
o
M
CO
CO
H
§
^^
2:
[d
to
as1
fT3
f\^
M
w
CO
CO
CM
I
I
r—
U
_]
PQ
H
W
o
rH
M
05
i
a
H <
H
U
£
z
o
EMISSI
en
en
«
flu
C
O
•H
U
w
0)
s
CO
CO
01
a.
I
~
S
o
w
U
tt
tn
a
3
n
91
I*
a.
00
23
CQ
C
O
y_,
O
14
«
•8
s
V
en
6
•H
O
V
i
n
u t*
C X
a. ^
rH W
rH so
<: ji
<0
4J
5 £.
rH — ,
O. /—
HI r-l
rH — '
00 00
C Jf
•H
l/l
CO
C >s
(d *"^
fH ^^
p- g
r-l SO
•< M
to
C I"
C f
»-t *^.
Co"
SI ~t
00 00
c ^
en
CO
U Ui
c ,x
rt --
Bu jC
*H ^^
•H 00
< ^
CO
4_1
C M
a ^:
o. <~
O r-l
rH *— '
00 00
5 *
s
0
a
3
CU
O
Bu
J=
•"
e
rH
^
j:
u
c
•H
rH
CD
U
rH
a
P
r-
.
ol
tn
/~*
in
O
O
O
'*"'
o
o
o
&
-a-
d
(N
o
.^
§
S
o
GO
§
o
o
d
^
^
r*.
i^
u-t
CNJ
m
^^
^o
If}
§
o
in
(Nl
0
o
o
n
^f
O
r*
in
ON
in
(M
ON
1
O
•w
I
o
00
CO
m
r-
-i
r^
§
§
d
.00003
o
t;?
in
^^
ON
in
fsl
^^
ON
o
o
o
m
o
O
o
vO
ON
r-.
/-s
(N
^^ '
00
ON
ON
NO
O
d
s— '
o
O
d
09
^p
—
r^
NC
CM
O
o
o
d
o
o
§
c
2
^
s^
„_,
CO
^^
O^
vO
o
d
f^,
§
d
o
CM
-
CM
p*.
r*-
(T>
^
CM
CO
CM
O
d
""*
CO
CM
o
o
1?
OS
1
*^
CM
CM
ON
-H
O
d
.00009
o
£
IN
1^*
r**
CM
m
^^
CM
00
o
d
03
CM
O
d
CM
o
(N
•
U
•H
S
a
to
a
rH
a
o
HI
rH
01
ca
u
o
1C
a
•a
CO
to
a.
X
eg
CO
C
o
CO
CO
e
U
•H
e
CO
U
153
-------
Fine screen wire
or
Standard drum
covering
Fine perforated
metal
Figure 7-5.
Condenser drum with screen wire and fine
perforated metal coverings.62
-------
REVOLVING DRUM
a. Revolving drum in-line filter
STATIONARY SCREEN
b. Stationary screen in-line filter
OUST a LINT
LAOEN Alfl
CLEAN Afft
CLEAN AIR
CLEAN AIR
WIPING ARM
FILTERING
SC3ESN
OUST a LINT
ROUND AIR FILTER
c. Horizontal, round in-line filter
Figure 7-6. Schematic diagrams of the most prevalent in-line filters.
64
155
-------
Despite being half as efficient as in-line filters, covered condenser
drums are used more predominantly because they are less expensive to install
and maintain. The installation of a typical in-line filter costs about
seven times more than the simple covering of a condenser drum. Because of
its predominance in the ginning industry the model gins presented in this
document are assumed to be equipped with covered condenser drums as baseline
control.
It is difficult to assess the level of arsenic emissions from the low
pressure section of the gin because tests on these lint fly and cotton dust
particulates indicated that virtually no or almost undetectable amounts of
arsenic were present (Sections 7.1.6.2.1 and 7.1.6.2.2). However, for the
purposes of this study the arsenic level (total elemental arsenic) of the
low pressure section emissions is assumed to be 1.33 ppm by weight of the
total particulate matter emitted. The 1.33 ppm value was selected as the
arsenic level because it represents the highest amount of arsenic measured
in cotton seed samples. Therefore, it can serve as an upper bound for the
arsenic emissions estimate. Arsenic emissions from the low pressure section
of the model gins are presented in Figure 7-2. An example calculation for
estimating low pressure section arsenic emissions is given in Appendix A-7.
Annual arsenic emissions from the low pressure sections of the total source
category are estimated to be 10.9 kg (24.1 Ib) .
7.2.2.3 Cotton Gin Model Plant Fugitive Inorganic Arsenic Emissions.
As summarized in Section 7.1.6.3, fugitive arsenic emissions from cotton
gins handling arsenic acid-desiccated cotton are estimated to be equal to
process arsenic emissions from the gin. Table 7-2 presents the estimated
fugitive arsenic emissions for all cotton gin model plants. Only gin high
pressure emissions are included because they constitute the emissions that
are collected by cyclones, sent to a bur hopper, and potentially emitted as
fugitives during hopper dumping. Bur hopper dumping is important because it
is the largest single source of gin fugitive arsenic emissions. Other
sources of gin fugitive emissions include open trash piles and leaks from
gin equipment and piping.
Several techniques to control fugitive gin emissions are available
including:
- the use of mechanical conveyors to move trash into the bur hopper,
- the use of enclosures at hopper unloading areas,
- the use of windscreens around bur piles, and
- the use of wet suppression.
All of these techniques have been applied at various gins with varying
degrees of success. However, the majority of the gin population either does
not use these techniques or does not properly apply them when used.
Cotton specialists report that the use of mechanical conveyors is just
beginning to gain popularity in the ginning field due to efforts to reduce
fan usage and electricity consumption. Conveyor usage is not widespread.
156
-------
The use of enclosed hopper unloading areas has not been greatly successful
because ginners have indicated that the haul truck engines become so clogged
with particulate material that they fail to operate. As a result, ginners
open the doors to the enclosure during dumping or they poorly maintain the
structure such that leaks are allowed to develop and fugitive emissions once
again occur. The majority of ginners do not use wet suppression because it
disrupts the operation of the equipment used to spread the gin wastes on the
fields. When the gin trash gets wet, the spreading equipment can become
clogged and malfunction. However, if the gin wastes are to be mass-dumped
on the fields, then wet suppression causes no problems. Despite the
difficulties that have been incurred in attempting to control cotton gin
fugitive emissions, several facilities have successfully applied the
techniques discussed above, particularly the use of mechanical trash
conveying and bur hopper enclosures.
The type and extent of fugitive emission control required cannot be
determined without an examination of each individual cotton gin handling
arsenic acid-desiccated cotton. For this reason it is not possible, in a
screening study of this type, to recommend an EEC system or an EEC level
that could realistically be applied to every gin in this source category.
The lack of accurate definition of the amount of inorganic arsenic contained
in the fugitive emissions also raises questions as to the need for an EEC
level for the purpose of inorganic arsenic control. Because of these many
unknown and unquantified variables concerning cotton gin fugitive arsenic
emissions, only estimated existing fugitive arsenic emissions will be
addressed in this study for the purposes of the population exposure
assessment in Section 7.5.
7.3 DEFINITION OF EEC OPTIONS
7.3.1 EEC Options for Process Emissions
Classical particulate matter emission controls such as fabric filters
and scrubbers were not investigated in this study as arsenic EEC options
because they have been,shpwn in past studies to be too costly for gin
particulate control. ' '
Because gins are not stringently regulated in Texas and Oklahoma, a
situation has developed in which there is little or no incentive for ginners
to install highly effective particulate control devices beyond the basic
required cyclone and filter setup. As a result, it is difficult to
determine potential EEC by examining existing gins in these two States.
However, an examination of States such as California and Arizona, where
cotton gin regulation and enforcement is stricter- yielded two particulate
control systems that were considered for EEC. ' -. The first device,
depicted in Figure 7-7, is known as an Intr-A-Vac. The Intr-A-Vac® is
intended to remove lint and dust particles from cotton gin and textile plant
exhaust air streams. It is not a bulk trash handling device, and therefore,
must be preceeded by a cyclone bank or separator chamber if it is to be used
157
-------
CLEAN AIR FILTERS THROUGH ,
tlOM OF FILTERINO SECTIONS
CONTAMINATED
AWMkET
ROTATING SUCTION MANIFOLD
MANIFOLD BEARING
* AIR SEAL
FILTERED PRODUCT TO
WASTE COLLECTOR
WITH ENTRANCE ENCLOSURE
WITH FULL ENCLOSURE
Figure 7-7. Intr-A-Vac-^particulate matter control system.
153
-------
on high pressure exhausts. As shown in Figure 7-7, dust-laden air enters
the device and is passed through a stationary screen drum lined with a
fur-like filter media. As the air is passed through the filter media the
dust particles are caught and eventually vacuumed away by a rotating suction
nozzle. The cleaned air is then exhausted to the atmosphere or redirected
to other areas.
The second control device for cotton gin EEC consideration is a new
type of cyclone known as the "1D-3D" or "long-cone" cyclone, which is
illustrated in Figure 7-8. The long-cone cyclone has been demonstrated to
be much more efficient than the older 2D-2D or AEC7tyne cyclones,
particularly on low pressure system trash. ' ' ' ' ' The long-cone
cyclones have been demonstrated to be capable of achieving an average
particulate removal efficiency of 94 percent on low pressure
exhausts. ' '
®
The long-cone cyclones were chosen over the Intr-A-Vac system for
further consideration as an EEC option for process emissions^-because the
particulate removal performance efficiency of the Intr-A=Vac system has not
been demonstrated by any type of testing, and Intr-A-Vac systems have been
used primarily on gins handling relatively cleaner picked cotton and not on
the dirtier, stripped cotton of Texas and Oklahoma. In addition, according
to local cotton specialists the use of new types of cyclones is more likely
to be accepted by ginn.grs than a change to a totally new, unproven system
such as the Intr-A-Vac^. The use of long-cone cyclones as an EEC option for
high and low pressure system exhausts is discussed in the following two
sections.
7.3.1.1 High Pressure Section Exhausts. As discussed in
Section 7.1.6.1, the majority of cotton gins in this source category have
cyclones in place for high pressure particulate emissions control as a
routine part of gin equipment. These primary control cyclones would remain
in place under the EEC option. Therefore, under this baseline control
condition long-cone cyclones would function as secondary control devices for
high pressure section emissions. Long-cone cyclones cannot be installed as
the primary control device on most of the high pressure exhausts due to the
small diameter design of the cyclone's inlet duct. The large size trash
emitted from the high pressure points can clog the small inlet duct and
disrupt airflow and performance of the cyclone. However, long-cone cyclones
can be installed as the primary cyclone on the suction unloading fan exhaust
of the high.,pressure section because the trash is not large enough to clog
the inlet. As an EEC option in this study, long-cone cyclones were used
as secondary collection devices on all high pressure section exhausts.
As a particulate and arsenic control device, the long-cone cyclones are
used to collect particles from the exhaust stream of the existing 2D-2D
cyclone bank. Ninety-nine percent of the exhaust from the baseline 2D-2D
cyclones is less than 20 microns in,size and is controlled at a 50 percent
efficiency by the long-cone system. ' ' The estimated levels of arsenic
159
-------
CVCICNE DESIGK PROPORTIONS
Be « De/8
He » DC
Oe • Oc/2
Lc « DC
Sc - Oc/8
Zc « 3Dc
Jc - Dc/4
Figure 7-8. Relative dimensions of the long-cone cyclone
70
160
-------
emissions resulting from the implementation of the long-cone EBC option are
presented in Table 7-3. The total arsenic emissions reduction for the
source category, after the implementation of EEC, is estimated to be 804 kg
(1800 lb)/year.
7.3.1.2 Low Pressure Section Exhausts. As an EBC option in the low
pressure section of the gin, long-cone cyclones function as the primary
exhaust cleaning device. The lint fly and cotton dust emissions to be
removed will not clog the inlet duct of the long-cone cyclone, therefore, no
problems are incurred using the long-cone system as a primary device.
Studies performed by the USDA have demonstrated that long-cone cyclones are
an average of 94 percent efficient as the primary control device for low
pressure section particulate emissions. ' Even though the long-cone
devices are technically feasible and effective at controlling these
particulate emissions, they are costly. Analytical tests on cotton lint
have not found arsenic to be present, thus implying that lint fly emissions
would also be devoid of arsenic.
Even if it is assumed that the low pressure section arsenic emissions
are equal to these given in Table 7-2, arsenic EBC options may not be
justified because of the high arsenic removal cost incurred. To illustrate
the cost of arsenic control, the low pressure emissions of the 20 bale/hr
model gin will be analyzed. The 20 bale/hr model was selected for analysis
because it represents the model plant most likely to be able to afford
controls due to the effects of economies of scale. The control cost of
long-cone cyclones on low pressure exhausts would be about $48,500/kg
($23,000/lb) or $48.5 million/Mg ($46 million/ton) of arsenic removed from
the air. This cost estimate only includes the annualized capital equipment
cost, and no other annualized costs such as maintenance and labor,
utilities, taxes, and insurance. The equipment cost data used to calculate
the arsenic control cost are given in Appendix B-7. Potential arsenic
emissions from gin low pressure exhausts will not be 'iscussed further
because no inexpensive EBC option appears available.
7.3.2 EBC Option for Fugitive Emissions
As discussed in Section 7.2.2.3, no EBC option could accurately be
defined and quantified for fugitive inorganic arsenic emissions from cotton
gins. EBC was not quantifiable because neither fugitive particulate nor
fugitive arsenic emissions from gins have been measured. The level of
fugitive arsenic emissions from a particular gin is highly dependent on the
gin's own maintenance practices. An across-the-board EBC option may not be
applicable to all fugitive emission situations at all gins. The fugitive
emission control techniques of bur hopper enclosures, wet suppression,
mechanical conveying, and windscreens are useful in reducing fugitives,
however, the effectiveness of the techniques in reducing inorganic arsenic
emissions has not been quantified. Because of the lack of adequate fugitive
emissions data and the variability of the fugitives problem from gin-to-gin,
no EBC option for fugitive emissions could be determined. The arsenic
161
-------
(0
CO
2
O
H
PM
O
CQ
Cd
O
H
W
CQ
H
=3
CO
OS
g
CO
Z
0
H
a
w
oc
o
t— ;
CO
CO
33
0
1-1
§
CO
Crf
^c
C^l
r^.
u
pa
£•4
C
O
•^ *— *
(j o J=
3 C ~-
•O -rt JZ
01 f-l IH
e£ u ^
to
a a
C b
ft) W V-*
s > .e
0> O •-.
tj 1^
c
HH
to
§ ~
•H t-l
w X
cn •>•»
i™* J3
E 3
u
c
01
to u
11 J=
60
tQ
e
0
•rt
I
£
u
to
e.
O ^-v
•rt tJ
to s
m ^*
•H J3
E -H
U — •
U
c
41
to
^4
41 b
C IM
r-4 00
41 ^
10
CO
pa
41 0)
B i-H
Tl O
i-^ b
0) u
to c
^
CO £
e B
0-
(0
-4 01
01 i-l
"^ <0
o ^
^^
CO
e*j
§
O
m
§
o
00
esj
§
d
§
o
oi cr.
C 41
o e
u o
on u
3*
/_>
vO
O
O
Q
^^
in
o
o
C3
to
01
c
o
O 1-1
*
a u
JM
«^
ffm^
O
m
c
o
o
f>
CM
§
O
o
m
§
o
1
o
41 O
c a
0 C
U 0
1 p-l
oo u
^^
o^
§
o
in
O
O
o
to
01
c
o
a IH
CM 0
CM
r^.
^>^
m
§
O
d
c^
o
o
d
m
00
o
0
o
o
oi a
§01
c
u o
1 i-4
tO U •
^
ON
o
Q
— '
^
f^.
§
c
to
41
e
o
O i-l
CM U
AS
CM
CM
^^
,•4
•»
O
o
*~*
^
kO
o
o
o
£•
o
d
-3-
O
o
o
01 <0
B 01
O B
U 0
1 l-l
00 U
B >•>
0 U
,_
CN
00
CM
O
d
00
CM
O
d
CO
41
B
O
1 ><
P U
CM
O
CM
•
i %
Li 4)
C CO
01 h<
CQ <0
41
O. <0
oi e
K i
0)
to •-:
to
3 i-l
eg s 10
rl
-------
emissions from area fugitives, presented in Table 7-2, are assumed to remain
unchanged for baseline and EBC conditions.
7.4 DETERMINATION OF THE INCREMENTAL IMPACTS OF THE EBC OPTIONS
Only the impacts of the EBC options for gin high pressure section
arsenic emissions will be presented because, as discussed in Section 7.3.2,
low pressure section arsenic emissions are virtually nonexistent or so low
that EBC is costly for most ginners. The application of EBC to cotton gins
involves the addition of a control device and not the substitution of one
control for another. Therefore, the incremental impacts of the EBC option
are equal to the actual impacts of the added device.
7.4.1 Environmental Impacts
The application of EBC to the model cotton gins results in an increase
in the total solid waste load of the gin. The quantity of solid waste is
increased because the EBC options collect additional gin trash from the
exhaust stream of the baseline controls. The increase in the overall solid
waste collected from total gin, either as total particulate matter or
arsenic, is shown in Table 7-4. The collected solid waste is generally
stored for as long as necessary, and then is put back on the crop fields,and
incorporated into the soil as a compost material for soil enrichment. '
The trash solid waste can be returned to the soil because it is not _
hazardous under the Resource Conservation and Recovery Act statutes.
No wastewater or other environmental impacts are produced with the use
of the long-cone cyclone EBC option.
7.4.2 Energy Impacts
The use of EBC results in increased electrical energy requirements to
operate the fan motors of the long-cone cyclone system. Additional fans are
required with the long-cone cyclones to overcome the added pressure drop
they cause. The incremental energy impact of EBC is represented by the
absolute energy impact of the EBC system and is shown in Table 7-5.
7.4.3 Control Costs Impacts
This section presents the capital and annualized costs associated with
the application of the long-cone cyclone EBC option to the cotton gin model
plants. The actual costs of EBC represent the incremental cost that would
be incurred by a ginner with the application of the designated long-cone
cyclone EBC system. The capital and annualized costs of EBC are given in
Table 7-6. The bases for the costs in Table 7-6 are given in Appendix B-7.
7.4.4 Economic Impact Resulting from EBC
An economic impact analysis was conducted for the cotton ginning model
plants to estimate the potential impact of requiring EBC. Using the cost
163
-------
TABLE 7-4. SOLID WASTE IMPACTS FROM COTTON GIN MODEL PLANT CONTROL
Model Plant Size Incremental Solid Waste Load from
(bales lint cotton/hr) EEC Option, Mg(ton)/yr
Arsenic Contained Total Solid
in Solid Waste3 Waste
4 neg.b 0.9 (1)
7 - neg.b 2.7 (3)
12 neg.b 7.3 (8)
20 neg.b 11.8 (13)
Q
Expressed as total elemental arsenic.
Negligible incremental difference with the number of significant figures
used.
164
-------
TABLE 7-5. ENERGY IMPACTS FROM COTTON GIN MODEL PLANT CONTROL
Model Gin Size Incremental Consumption
(bales lint cotton/hr) for EEC Option,
Gj(kWh)/yr
4 24.8 (6,900)
7 68.4 (19,000)
12 166 (46,100)
20 260.0 (72,400)
165
-------
CiO
z
0
H
a-
O
pa
H
•J
P-.
>-3
to3
C
i
M
O
z
o
g
u
Cb
o
H
C/3
0
u
a
u
Cvl
M
^5
S3
Z
Z
a
55
,J
<2
^1
Pu
0
^O
r—
U
pa
H
4J
00
o
u
TJ
0)
N
•H
f-t
s
c
c
to
o
H
t-4
a
•H
a.
n
u
a
4J
o
c
o
•••4
4-1
0
u
•o
01
N
•H
*-4
to
3
C
C
CO
iH
<0
o
01
t-
Q
to
U
eo
1-4
"o
T3
CO
0\
—4
1-1
6)
la
10
a
er
a
(—4
C
01
u
to
n
4J
01
O
U
1— 1
^
jO
166
-------
data in Table 7-6 and the methodology given in Appendix C, revenue and
product price increases, required by gins to maintain the same net present
values before and after the installation of EEC, were estimated.
The annual revenue increases required by the four model cotton gins
range from $3,800 per year in the smallest model plant up to $15,900 in the
largest model plant. These increases translate into price increases that
range from less than Ic per kg (0.45C per Ib) of baled cotton in the largest
model up to 3.5c per kg (1.6<: per Ib) in the smallest model. These
increases are equivalent to 0.5 percent and 2.1 percent product price
increases, respectively. Table 7-7 presents a summary of the revenue and
price increases for the cotton gin model plants.
To test the sensitivity of the economic impact results to the various
input data, a sensitivity analysis was conducted for the cotton ginning
model plants, assuming simultaneous 15 percent decreases in baseline price
and annual output and a 15 percent increase in the weighted cost of capital.
The altered assumptions of the sensitivity analysis caused product
price increases to rise to l.lc per kg (0.5C per Ib) of baled cotton in the
largest model plant up to 4.4c per kg (2c per Ib) in the smallest model
plant. These increases are equivalent to 0.8 percent and 3.0 percent
product price increases, respectively. The revenue and price impact
sensitivity analysis for cotton gin model plants is summarized in Table 7-8.
7.5 POPULATION EXPOSURE DATA
The data in Tables 7-9 and 7-10 were developed during this study for
use in an air pollution dispersion model to estimate the population exposure
to arsenic from cotton ginning operations. The Strategies and Air Standards
Division (SASD) of EPA is using these data in an ongoing effort to estimate
exposure to arsenic from this source category.
7.6 EASE OF STANDARDS DEVELOPMENT
Technically there are no impediments to applying the long—cone cyclone
EEC option to control arsenic emissions from either new or old ginning
facilities. The long-cone cyclone technology has been demonstrated and
would be applicable to the emissions of any gin in this source category.
Ginners would not have a problem obtaining the long-cone cyclones because
they are manufactured by several companies that are located in the cotton
belt States. In addition, the cyclone design was originated by engineers at
Texas A&M University, such that information on the system is readily
available.
An important factor to consider before attempting to regulate inorganic
arsenic emissions from cotton gins is whether all of the arsenic that is
emitted is in the inorganic form. As discussed in Section 7.1.3.6, organic
arsenic, which is not listed as a hazardous air pollutant, may be present in
167
-------
TABLE 7-7. REVENUE AND PRICE IMPACTS FOR COTTON GINS TO ACHIEVE EEC
MODEL REQUIRED REVENUE
PLANT INCREASE
(dollars)
REQUIRED UNIT
PRICE INCREASE
(dollars per kg)
REQUIRED PERCENT
UNIT PRICE INCREASE
3,804
.035
2.1
6,214
.014
0.8
12,117
15,898
.011
.009
0.7
0.5
168
-------
TABLE 7-8. SENSITIVITY ANALYSIS OF THE REVENUE AND PRICE
IMPACTS FOR COTTON GINS
MODEL REQUIRED REVENUE
PLANT INCREASE
(dollars)
REQUIRED UNIT
PRICE INCREASE
(dollars per kg)
REQUIRED PERCENT
UNIT PRICE INCREASE
4,027
.044
3.0
6,574
13,954
16,837
.018
.015
.011
1.2
1.0
0.8
169
-------
TABLE 7-9. COTTON GIN MODEL PLANTS' ARSENIC EMISSIONS DATA
Model Gin
Size
(bales/hr)
Emission
Points
Baseline Arsenic
Emissions,
kg/hr (Ib/hr)
EEC Arsenic
Emissions
kg/hr (Ib/hr)
Process
0.0025 (0.0056) 0.0013 (0.0028)
Area Fugitives 0.0025 (0.0056) 0.0025 (0.0056)
a
12
20
Process
Area Fugitives
Process
Area Fugitives
Process
Area Fugitives
0.0045 (0.0099)
0.0045 (0.0099)
0.0077 (0.0169)
0.0077 (0.0169)
0.0128 (0.0282)
0.0128 (0.0282)
0.0023 (0.0050)
0.0045 (0.0099)'
0.0039 (0.0085)
0.0077 (0.0169)'
0.0064 (0:0141)
0.0128 (0.0282);
No EBC for fugitives could be defined. Fugitive emissions are
equivalent in the baseline and EBC cases.
170
-------
0
2
O
H
H
O
o
"T
r^
W
ca
H
C/3
&4
W
H
W
2
CM
^J
PM
^
u
H
C/2
•rl
CJ
O
i— 1
cu
5
o
I— 1
Cn
0)
4J
CO
OS
JJ
o
I-l
]*T)
•
p.
e
0)
H
4J
•H
X
w
1-1
01
4J
CJ"
r- f^- r-
\Q ^O **O
S™^ . \^^ \^/
oj cd oj
**3" <3" -^"
• • •
000
CN CN CN
/•"•N ^"^ X" N
O O O
in O in
r^ co vO
«« * M
-a- in o
CN -
r* ^ ^
•H -H -H
cn 4J cn 4J cn 4-1
eo -H cn -ri cn TH
Ol 60 01 60 CU 60
0 3 03 03
O Cn O [x-i O Pn
IH IH l-i
CM CO PH cfl CM cfl
01 Ol 01
IH IH IH
^C ^ "^
r~- CN o
<-> CN
60
C
1-1
CU 01
oi e TJ
42 3 6 -H
4_) ^ 3
o
C M-l • ^
CU O CN 4-1
(^ Ll^
3 4J TJ
(3 4J C O
4J TH C Cfl rH
1— 1 O CU '
3 (X TH 60
cn 42 C S
Ol Ol 0 O
i-l 42 cfl iH O
H
cn 4J /— \ co
CU Cfl 4J
^ • (4-1 TJ
•rl TJ CU (2
4-1 C M O Cfl
•H T-l CO CN
60 S ^ 60
3 TJ C
MH >, CU S O
43 en rH
Cfl CO rH
01 TJ CU • /— \
IH Ol i— 1 v£> U
Cfl CO CU MH
rl M ^>
CO Ol i-H O
JZ Cu 01 Ol CO
H Cfl IH 4-1 •— •
•H CO CO
TJ g S
• 4-1 *rl
cu o> co x -H
0 rl 42 O •
M CO 4-J M ON
3 CU
O TJ cn cu en
eo (2 C cO IH
cfl o cu
Ol -H 0) CU
> ^ eo IH cu
TH o cn co o
4J 3 TH 42
•H IH e cn
60 4-1 CU (2 Ol
3 TH >
M-i cfl CU 60 cfl
> ,c
CO O TH M
cu 4-1 4-1 42 cn
^ C TH --. (3
CO -H 60 0) TH
3 i-H 60
eu -a M-I cB
42 01 42 IH
4-J Cu Ol 42
ri 3 H ^ "aT
O TJ TJ i-H
M-l (2 cfl
CJ • CO 42
01 M 42
•-I CO 60 «* O
42 -rl CN
CO 1-1 42 CU
O 01 42 TJ
•rl O./— » 4J C
iH CU 4-1 Cfl
CU O M-4 4-J
Cu 42 Cfl CM
co in *-H
IH r-H CO
4-1 3 •— ' !H Ol
O 43 01 42
P B Cu 4J
Ol CU
0) 42 vo O Ol
IH 4J • 42 rH
co , 42 3
in i-H H
cu en cu •>
4-1 4J 4-1 CU
CJ C CO • TJ
B 0) B 01 TH
CO 4-1 TH iH 3
^ C X 3
CO O O 4J /—>
CU O V-l Cfl 4J
Cu IH MH
01 42 P. 0)
cn en co cu r»
cu co B •
42 IH cn a) VJD
H 4J -H 4J ^^
(0
171
-------
the emissions to the atmosphere. Studies need to be performed to speciate
the arsenic contained in cotton gin emissions in order to assess the
severity of the inorganic arsenic problem, since inorganic arsenic is the
determined hazardous air pollutant. No speciation for arsenic has been
performed on emissions from gins handling arsenic acid-desiccated cotton.
However, EPA is preparing to reanalyze the hi-vol filter samples taken
during the Durrenburger tests of 1974 (Section 7.1.4) to speciate the
arsenic contents. The results of this analytical work should be available
from the Environmental Monitoring Systems Laboratory (EMSL) of EPA, Research
Triangle Park, North Carolina by mid-1982. To justify the development of
emissions standards for inorganic arsenic from cotton gins, the level of
inorganic arsenic in gin particulate matter emissions must be quantified.
An entire approach to source testing cotton gins for arsenic emissions
and arsenic removal efficiencies for both process and fugitive sources needs
to be developed and verified if a decision is made to study cotton gin
arsenic emissions. The sampling of gins for arsenic may present problems
due to the following:
- difficulty in determining the amount of arsenic coming into the
ginning process,
- difficulty in measuring inlet streams to the cyclones by
conventional methods due to the large particle sizes of the
trash,
- non-existence of test data on fugitive arsenic emissions from
gins,
- no testing protocols for gin fugitive emissions have been
demonstrated.
Another factor to be considered in any decision to further investigate
cotton gins for inorganic arsenic standards development is the view held by
the States involved towards gin sources. Neither Texas nor Oklahoma view
cotton gins as major emission sources of inorganic arsenic. As a result
neither State attempts to directly regulate or monitor potential arsenic
emissions from cotton gins. Although the States do indirectly control
potential arsenic emissions by means of particulate matter regulations, the
level of control achieved is not definable because the particulate
regulations are not rigidly enforced. Source testing for compliance
determination purposes is not performed in either State. Primarily, Texas
and Oklahoma attempt to get gins to install designated control devices, such
as cyclones and filters, to reduce the gross particulate emissions problem
to a tolerable limit. In the absence of public nuisance complaints, the gin
sources are left alone.
Neither Texas nor Oklahoma pursue cotton gins as violators of the NAAQS
for particulate even though several measurements around gins have shown
gross exceedences of the standards. State officials indicate that because
of the many other sources of windblown particulate in these areas, it is not
possible to place a blame or a percentage of blame on gins for causing the
172
-------
NAAQS exceedence. Agricultural fields other than cotton are contributors in
these areas. As with the particulate SIP's, the States do not believe that
the seasonality, agricultural nature, and comparatively small size of the
gin sources warrants regulatory action to reduce the gins' contribution to
the problem. Any gross particulate reductions made by the gins to aid in
reducing the NAAQS exceedence problem would have a positive effect in
reducing arsenic emissions.
In the event inorganic arsenic regulatory development is deemed
necessary, close consultation and coordination between Federal and State
agencies is recommended in order to produce a workable, more effective
standard.
173
-------
7.7 REFERENCES
1. PEDCo Environmental, Inc. Arsenic Control Costs for Industrial Sources.
(Prepared for U. S. Environmental Protection Agency.) Research
Triangle Park, North Carolina. EPA Contract No. 68-02-2842. June 30,
1978. Section 7.2.8.1.
2. Development Planning & Research Associates, Inc. Economic Impact
Screening Study of Arsenic Sources: Primary Copper Smelters and Cotton
Gins. (Prepared for U. S. Environmental Protection Agency.) Research
Triangle Park, North Carolina. EPA Contract No. 68-02-3535. February
1982. p. III-5.
3. Abernathy, J. R. Role of Arsenical Chemicals in Agriculture, (Paper
Presented at Arsenic Symposium Sponsored by Chemical Manufacturer's
Association and the National Bureau of Standards, Gaithersburg,
Maryland. November 4-6, 1981.)
4. The Biologic and Economic Assessment of Pentachlorophenol, Inorganic
Arsenicals, and Creosote - Volume II: Non-Wood-Preservatives. U. S.
Department of Agriculture. Technical Bulletin Number 1658-11.
November 1980. pp. 24-26.
5. Miller, C. S., and E. M. Bailey. Arsenic Acid Use and Hazard
Assessment in the Desiccation of Cotton. The Texas Agricultural
Experiment Station Report MP-1434. October 1979.
6. Telecon. Brooks, G. W., Radian Corporation, with Parnell, C., Texas
A & M Agricultural Experiment Station. January 25, 1982. Conversation
concerning cotton gin arsenic emissions.
7. Reference 2, pp. III-6 to III-7.
8. Reference 4, pp. 38-39, 60.
9. Telecon. Brooks, G. W., Radian Corporation, with Baker, R., U. S.
Department of Agriculture. January 25, 1982. Conversation concerning
cotton gin arsenic emissions.
10. Telecon. Brooks, G. W., Radian Corporation, with Ivey, D., Texas
Department of Agriculture. January 29, 1980. Conversation concerning
cotton gin arsenic emissions.
11. Reference 4, pp. 47-48.
12. Reference 4, p. 46.
13. Miller, C. S., et al. Pesticide Residues in Cotton Gin Wastes. Texas
Agricultural Experiment Station Report MP-1184. April 1975.
17 A
-------
14. Telecon. Brooks, G. W., Radian Corporation, with Bodovsky, P.,
Continental Conveyor. February 22, 1982. Conversation concerning
cotton gin emissions control.
15. Environmental Engineering, Inc. Background Information for
Establishment of National Standards of Performance for New Sources:
Cotton Ginning Industry - Draft Final Report. Prepared for U. S.
Environmental Protection Agency.) Research Triangle Park, North
Carolina. July 15, 1971. p. 2-4.
16. Rawlings, G. D., and R. B. Reznik. (Monsanto Research Corporation.)
Source Assessment: Cotton Gins. (Prepared for U. S.vEnvironmental
Protection Agency.) Cincinnati, Ohio. EPA-600/2-78-004a. January
1978. p. 15.
17. McCaskill, 0. L., and R. A. Wesley. Unifilter Collecting System for
Cotton-Gin Waste Materials. U. S. Department of Agriculture Report
ARS-S-144. September 1976.
18. Parnell, Jr., C., and R. Baker. Particulate Matter Emissions by a
Cotton Gin: A Study. The Cotton Gin and Oil Mill Press. April 17,
1971.
19. Reference 4, p. 54.
20. Reference 4, p. 27.
21. Reference 4, pp. 49-50.
22. Telecon. Brooks, G. W., Radian Corporation, with Brooks, W., Pennwalt,
Inc. January 7, 1982. Conversation concerning arsenic acid
production.
23. Telecon. Brooks, G. W., Radian Corporation, with Smith, D., Voluntary
Purchasing Group. January 22, 1982. Conversation concerning arsenic
acid production.
24. Telecon. Brooks, G. W., Radian Corporation, with Abernathy, J., Texas
Agricultural Experiment Station. January 7, 1982. Conversation
concerning cotton gins.
25. Telecon. Brooks, G. W., Radian Corporation, with Metzer, R., Texas
A & M University. February 4, 1982. Conversation concerning arsenic
acid substitutes.
26. Telecon. Brooks, G. W., Radian Corporation, with Wallin, G., Texas Air
Control Board. January 29, 1982. Conversation concerning cotton gin
emissions.
175
-------
27. Durrenberger, C. Cotton Gin Report. Texas Air Control Board. May 31,
1974.
28. Miller, C. S., and M. M. Aboul-Ela. Studies of Arsenic Acid Residues
in Cotton. Texas Agricultural Experiment Station Report MP-771. June
1965.
29. Miller, C. S. The Efficiency of Arsenic Acid in Cotton Desiccation.
Weed Science. 22(4): 388-393. July 1974.
30. Miller, C. S., and W. H. Aldred. Determination of the Efficiency of
Stalk Application of Desiccants. Texas Agricultural Experiment
Station. (undated)
31. Johnson, D., and R. Braman. Alkylated Arsenic Compounds in Atmospheric
Samples. In: Proceedings of the First International Conference on
Heavy Metals in the Environment. Toronto, Canada, October 27-31, 1975.
pp. D-92 to D-93.
32. Telecon. Brooks, G. W., Radian Corporation, with Irgolic, K. , Texas
A & M University. March 2, 1982. Conversation concerning arsenic
speciation.
33. Irgolic, K. Speciation of Arsenic Compounds in Water Supplies.
(Prepared for U. S. Environmental Protection Agency.) Cincinnati,
Ohio. HERL-Ci-353. March 1982. pp. 1-6.
34. Fish, R. H., et al. Fingerprinting Inorganic Arsenic and Organoarsenic
Compounds in Situ Oil Shale Retort and Process Waters Using a Liquid
Chromatograph Coupled with an Atomic Absorption Spectrometer as a
Detector. Environmental Science and Technology, 16, No. 3, 1982.
pp. 174-179.
35. Reference 4, p. 41.
36. Reference 15, p. 1-3.
37. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for U. S. Environmental Protection Agency.) Research
Triangle Park, North Carolina. EPA Contract No. 68-02-3173. May 1982.
p. 4-16.
38. Reference 16, pp. 34-36.
39. Reference 16, p. 9.
40. Parnell, Jr., C., and R. Baker. Particulate Emissions of a Cotton Gin
in the Texas Stripper Area. U. S. Department of Agriculture Report
No. 149. May 1973.
176
-------
41. Research Triangle Institute and PEDCo Environmental, Inc. Capital and
Operating Cost Study of Model Cotton Gin Plants with Pollution Control
Systems. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, North Carolina. EPA Contract No. 68-02-0607.
May 1974. p. 4-6.
42. Reference 15, p. 3-10.
43. Reference 16, p. 33.
44. Reference 15, p. 2-19.
45. Gillum, M. N., et al. Primary and Secondary Cleaning of Lint Cleaner
Exhaust Air: A Progress Report. American Society of Agricultural
Engineers Paper No. 80-3077. June 1980.
46. Reference 41, p. 4-7.
47. Reference 16, p. 23.
48. Reference 15, p. 3-6.
49. Telecon. Brooks, G. W., Radian Corporation, with Wallin, G., Texas Air
Control Board. March 25, 1982. Conversation concerning cotton gin
emissions.
50. Telecon. Brooks, G. W., Radian Corporation, with Oklahoma Air Quality
Division. April 6, 1982. Conversation concerning cotton gin
emissions.
51. Telecon. Brooks, G. W., Radian Corporation, with Bradford, C., Texas
Air Control Board. January 28, 1982. Conversation concerning cotton
gin emissions.
52. State of Texas. Texas Administrative Code, Title 31, Natural
Resources, Chapter 111 - Particulates, Sections 111.71 - 111.76.
Amended April 6, 1981.
53. Reference 37, pp. 5-17 to 5-18.
54. Reference 4. p. 45.
55. Telecon. Brooks, G. W., Radian Corporation, with Gordon, C., OSHA.
April 12, 1982. Conversation concerning OSHA arsenic standards.
56. Reference 15, p. 4-2.
57. Reference 16, pp. 41-42.
177
-------
58. Cotton Ginners Handbook. U. S. Department of Agriculture. Handbook
No. 503. July 1977. p. 81.
59. Baker, R. and V. Stedronsky. Gin Trash Collection Efficiency of
Small-Diameter Cyclones. U. S. Department of Agriculture ARS 42-133.
July 1967.
60. Wesley, R. A., et al. An Evaluation of the Cyclone Collector.
American Society of Agricultural Engineers Paper No. 70-848. December
1970.
61. Wesley, R. A., et al. A Comparison and Evaluation of Performance of
Two Small-Diameter Cyclones for Collecting Cotton Gin Waste. U. S.
Department of Agriculture Report ARS 42-167. January 1970.
62. Reference 16, p. 45.
63. Reference 58, pp. 82-83.
64. Baker, R. and C. Parnell, Jr. Three Types of Condenser Exhaust Filters
for Fly Lint and Dust Control at Cotton Gins. U. S. Department of
Agriculture Report ARS 42-192. September 1971.
65. Reference 2, pp. 111-21 to 111-27.
66. Reference 41, pp. 2-15 to 2-19.
67. Telecon. Brooks, G. W., Radian Corporation, with Bodovsky, P.,
Continental Conveyor. February 1, 1982. Conversation concerning
cotton gin control.
68. Letter and attachments from Bodovsky, P., Continental Conveyor to
Brooks, G. W., Radian Corporation. February 18, 1982. 40 p. Cotton
dust control.
69. Letter from Furr, R., Anderson-Bigham Sheet Metal Works, Inc. to
Brooks, G. W., Radian Corporation. March 22, 1982. 3 p. Cyclone
control costs.
70. Parnell, Jr., C. and D. Davis - Predicted Effects of the Use of New
Cyclone Designs on Agricultural Processing Particulate Emissions.
(Paper Presented at Southwest Region Meeting of American Society of
Agricultural Engineers. Hot Springs, Arkansas. April 25-27, 1979.)
71. Hughes, S. and M. Gillum. Collecting Particles from Gin Lint Cleaner
Air Exhausts. American Society of Agricultural Engineers Paper No.
81-3565. December 1981.
178
-------
72. Gillum, M. N., et al. Effect of Cyclones in Series on Gin Emission
Reduction. American Society of Agricultural Engineers Paper No.
80-3563. December 1980.
73. Telecon. Brooks, G. W. , Radian Corporation, with Bigham, V.,
Anderson-Bigham Sheet Metal Works, Inc. March 3, 1982. Conversation
concerning cyclone costs.
74. Memo from Paul, J., PEDCo Environmental, Inc. to Mappes, I., PEDCo
Environmental, Inc. May 8, 1981. 3 p. Arsenic Acid Usage and Cotton
Gin Emissions.
75. U. S. Environmental Protection Agency. Identification and Testing of
Hazardous Waste. Section 261.4, Exclusions. 45 Federal Register
33120. Washington, D. C. U. S. Government Printing Office. May 19,
1980.
76. Feairheller, W. and D. Harris. (Monsanto Research Corporation.)
Particulate Emission Measurements from Cotton Gins - J. G. Boswell
Company El Rico #9. (Prepared for the U. S. Environmental Protection
Agency.) Research Triangle Park, North Carolina. EMB Report No.
72-MM-19. (undated).
179
-------
8.0 GLASS MANUFACTURING
This chapter discusses the glass manufacturing industry and its
associated arsenic emissions. Sections 8.1 and 8.2 present a brief source
category description. Section 8.3 discusses the regulations currently
covering the industry and the resulting level of control and associated
arsenic emissions. Section 8.4 presents a discussion of the estimated best
control (EEC) options and Section 8.5 discusses the impacts of applying
these options. Section 8.6 presents the input data to be used in assessing
the population exposure to arsenic emissions from glass manufacturing. The
ease of standards development is discussed in Section 8.7.
8.1 SOURCE DESCRIPTION
This source category only includes glass manufacturing plants that use
inorganic arsenic compounds as a process raw material. The only sector of
the glass industry that presently uses inorganic arsenic raw materials is
classified under the Standard Industrial Classification (SIC) code 3229,
Pressed and Blown Glass Not Elsewhere Classified. There are about
150 plants in SIC code 3229, however, only 15 plants have been found to
currently be emitting inorganic arsenic. These 15 plants have a total of
30 glass melting furnaces emitting arsenic. In 1978 a detailed EPA survey
indicated that 17 glass plants were emitting inorganic arsenic, however, one
of these plants no longer uses arsenic and one has been closed down. ' The
companies that currently operate glass plants emitting inorganic arsenic are
Anchor Hocking, Inc., Corning Glass Works, Owens-Illinois, RCA-Picture Tube
Division, and the Lamp Division of GTE Sylvania. Table 8-1 lists the
locations of glass plants emitting inorganic arsenic.
The primary types of glass where arsenic is used include lead^opal,
lead silicate, borosilicate and aluminosilicate-ceramic (PyroceranP).
Inorganic arsenic compounds are used in glass manufacturing for a
combination of reasons depending on the particular glass being produced. In
the majority of cases arsenic compounds act as fining or clarification
agents. During the melting of the glass batch raw materials, gaseous
reaction products such as oxygen, nitrogen, and carbon dioxide are evolved
and rise through the glass melt in the form of bubbles. These bubbles
greatly reduce the overall quality of the glass. The addition of the
inorganic arsenic material causes the bubbles to rise more rapidly to the
melt surface and dissipate. It also appears that chemical reactions brought
about by the use of the arsenic,.reduces the release of some bubbles caused
by nitrogen and carbon dioxide.
A second function of arsenic in glass is to act as a decolorizing
agent. The effectiveness of arsenic trioxide in this use is based on the
180
-------
TABLE 8-1. GLASS MANUFACTURING PLANTS EMITTING INORGANIC ARSENIC0
PLANT NAME, LOCATION
PRIMARY PRODUCTS
RCA, Circleville, OH
GTE, Versailles, KY
Owens-Illinois, Columbus, OH
Anchor Hocking, Lancaster, OH
Anchor Hocking, Baltimore, MD
A
B
C
D
E
F
G
H
I
TV Picture Tube Components
Glass Tubing
TV Picture Tube Components
Tableware Glass
Container Glass
Heat Resistant Globes,
Electric Light Covers
Lead Glasses
Tableware Glass
TV Picture Tube Components
Optical Glasses
TV Picture Tube Components
Tableware Glass
Glass Tubing
Aluminosilicate (Pyroceram"
and Lead Glasses
Lead Crystal Glass
Plants A through J coded at the request of manufacturer, locatirns
available from the Strategies and Air Standards Division of EPA.
181
-------
ease of interconversion of the various oxidation stages. This
interconversion helps establish a concentration of arsenic pentoxide under
the equilibrium conditions of the melt. Arsenic pentoxide oxidizes divalent
iron impurities in the melt (which impart a greenish color to the glass) to
a trivalent iron, which results in a yellowish color glass. The
simultaneous addition of other elements such as nickel oxide, cobalt oxide,
and rare earth oxides provides balancing colors which produce a colorless
finished glass.
Inorganic arsenic is also used in some instances in special glass types
to impart particular properties that are needed for the end use of the
glass. For example, arsenic can provide stable fixation of certain colors
for optical glass by stabilizing selenium, provide high glass permeability
to infrared light for camera lenses, and provide a high degree of energy
transmission for solar collector glass.
With the advent of environmental and occupational health laws for
inorganic arsenic in the late sixties and early seventies, glass companies
began reducing arsenic usage and initiating research into arsenic
substitutes. In 1968 approximately 3900 Mg (4300 tons) of arsenic
(elemental),were used, while in 1981 total usage was estimated at 730 Mg
(800 tons). ' Five percent of the total arsenic consumed in the United
States goes into glass production. Several manufacturers of television
picture tube funnel glass have totally eliminated inorganic arsenic
compounds from the feed batch raw materials. ' Another manufacturer
reports that since 1978 arsenic usage was reduced 50 percent in a fluoride
opal glass and 30 percent in a borosilicate glass.
8.2 PROCESS DESCRIPTION
The production of an arsenic glass involves melting a uniform mixture
of raw materials in a furnace to obtain a homogeneous mass. Typical
materials include sand, limestone, soda ash, feldspar, sodium sulfate and
nitrate, anhydrous borax, potassium carbonate, and arsenic trioxide or
arsenic acid. Arsenic compounds may be introduced into the batch as either
arsenic trioxide powder or liquid arsenic acid with no effect on the overall
glass making process. Inorganic arsenic is also introduced into the batch
feed as a constituent of the return cullet or scrap glass. The level of
inorganic arsenic in the cullet is the same as the percent retained in the
glass of the total batch fill.
The batch raw materials are mixed and entered in the melt furnace where
they float on a bed of molten glass until they dissolve. The temperature of
the melt furnace is approximately 1500 to 1700°C (2700 to 3100°F). The
majority of the input inorganic arsenic is permanently fixed in the molten
glass and is not emitted from the furnace. Data submitted by the glass
industry indicates that from 70 to 99..percent of the total arsenic in the
batch material remains in the glass. ' ' The average percent of arsenic
retained as reported by three specialty glass producers was about
182
-------
94 percent. The remaining arsenic material in the form of arsenic trioxide
is evaporated from the furnace melt and is either eventually emitted as
arsenic vapor or it,.condenses as submicron arsenic particles or on other
particulate matter. ' ' Arsenic emissions from uncontrolled and
controlled plants are affected by the size of the particular glass furnace,
the type of glass produced, the type of controls used, and the size of the
controls used. The majority of the arsenic particulate matter is entrained
in the exhaust flue gas, however, a small portion can be deposited-as
particle fallout in the refractory checkerwork of the furnace. ' The
amount of arsenic being emitted to the atmosphere as a function of the
amount in the batch feed has not been quantified.
Inorganic arsenic emissions from the glass manufacturing process are
influenced by the type of furnace used to melt the batch feed raw materials.
The majority of existing plants producing arsenic-containing glass use
either regenerative- or recuperative-type furnace systems. The two types of
furnaces differ primarily in the method by which they recover heat from the
hot furnace exhaust gases. Regenerative furnaces utilize two chambers of
refractory material called checkerwork. At any one time, while combustion
flue gases heat the refractory in one checkerwork chamber, the other
checkerwork preheats furnace combustion air. After intervals ranging from
10 to 30 minutes, this gas flow is diverted so that combustion air is drawn
through the chamber previously heated by flue gases, and flue gases heat the
refractory in the other chamber previously used to preheat combustion air.
Recuperative furnaces use one continuously operating shell and tube type
heat exchanger to preheat combustion air insteacLof7the checkerwork heat
exhangers used in the regenerative furnaces. ' ' In the United States
regenerative-type furnaces predominate. In regenerative furnaces, arsenic
that has condensed on larger particles may be deposited in the checkerwork
from particle fallout. The extent to which arsenic may be removed in the
checkerwork has not been quantified. Recuperative furnaces would not have a
similar arsenic removal capability.
Regenerative and recuperative furnaces, which are fossil-fuel fired,
are being modified and in some cases replaced by systems using electrical
current to melt glass. Electric boosting is the term applied to the furnace
modification in which an electric current is used to augment furnace firing
of gas or oil. Electrical energy is converted to heat because of the high
electrical resistence of the molten glass. The effect of electric boosting
is to decrease the required furnace bridgewall temperature, which in turn
decreases the fyel consumption rate, thereby decreasing the pollutant
emission rate.
In some limited applications more traditional regenerative furnaces
have been totally replaced by systems known as all-electric melters.
All-electric melters produce less than 10 percent of the glass in the United
States and none of the units produce more than 136 Mg (150 tons)/day. The
surface of the melter in a cold top all-electric furnace is maintained at
ambient temperature and fresh batch raw materials are continuously fed over
183
-------
the entire surface. Because energy is supplied internally to the glass, a
higher percentage of the total energy expended can be converted into usable
heat for melting than with fossil fuel fired,melters.„„ Pollutant emissions
from all-electric melters are virtually nil. ' ' ' The main limitation
with these systems is that they cannot be used to produce all varieties of
glass. Not all glasses possess the electrical properties required for these
melters and some glass formulations actually corrode the electrodes
presently used in the all-electric melters.
A second source of inorganic arsenic emissions is the raw materials
handling part of the glass plant where powdered arsenic trioxide is
received. Fugitive arsenic dust emissions from raw materials handling are
highly controlled due to OSHA regulations for airborne inorganic arsenic in
the workplace. The situation here is analogous to the arsenic emissions
from arsenic trioxide handling in the arsenic chemical industry (see
Chapter 9). The glass industry uniformly controls these fugitive arsenic
dust emissions by enclosing the unloading, conveying, and storage areas and
venting them to fabric filters. ' As demonstrated in the arsenic
chemical industry (Chapter 9), this type of control method is effective in
lowering arsenic emissions to negligible levels.
The OSHA regulations are the primary reason that some glass companies
have switched from using powdered arsenic trioxide to liquid arsenic acid.
By using the liquid arsenic acid as a batch raw material, no arsenic is
released into the workplace. One manufacturer has indicated that they now
use only liquid arsenic acid for their glass manufacturing in order to
protect their workers and comply with the OSHA standard. Because this
source of potential arsenic emissions is already controlled with the
estimated best technology and the arsenic emissions are orders of magnitude
less than controlled process arsenic emissions, it will not be discussed
further in this chapter.
8.3 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING REGULATIONS
8.3.1 Regulatory Impacts
Atmospheric inorganic arsenic emissions from glass plants are presently
being controlled indirectly as a result of State and Federal particulate
matter regulations. ' ' The particulate matter regulations are met by
installing an add-on control device, installing an electric boosting process
modification, or collecting particulate matter as fallout in the furnace
checkerwork. Of these particulate control methods, the use of add-on
control devices does the best job of reducing arsenic emissions, while the
checkerwork fallout technique does the worst. Discussions with the
companies involved in this source category and State air pollution control
representatives indicate that the plants in Table 8-1 are in compliance with
applicable particulate matter regulations. However, arsenic emissions from
the facilities that do not have add-on controls are significant when
compared with the arsenic emission rates of plants with control devices.
184
-------
None of the plants in Table 8-1 are specifically regulated for atmospheric
emissions of inorganic arsenic.
8.3.2 Baseline Inorganic Arsenic Emissions and Controls
The estimated hourly and annual inorganic arsenic emissions from
existing glass plants producing arsenic-containing glass are given in
Table 8-2. The majority of the emission estimates were provided by the
companies involved. Total existing arsenic emissions from the source
category are about 51 Mg (56 tons)/yr. As discussed in Section 8.2, on the
average only about 6 to 10 percent of the arsenic in the glass furnace melt
is emitted with the furnace exhaust gases. The remaining 90 to 94 percent
is fixed in the glass material. The high temperatures of the melt
volatilize arsenic into the furnace offgas stream. Before the offgas stream
is emitted to the atmosphere it will be cooled down by heat losses in the
ductwork, by its passage through the heat recovery checkerwork, and possibly
by applying a gas cooling chamber, Upon cooling some of the arsenic in the
gas stream will condense and form submicron arsenic particles or will
condense on other particles in the stream. Test data on the uncontrolled
exhaust gas streams of two typical regenerative furnaces indicated.that from
80 to 92 percent of the arsenic found was in particulate form. ' The
remainder is emitted into the atmosphere as vaporized arsenic.
Because inorganic arsenic emissions from glass plants are not
specifically regulated, baseline arsenic emission levels are equal to the
levels generated as a result of." existing particulate matter regulations.
The worst situation for arsenic emissions occurs when existing plants are
able to rely on particle fallout in the heat recovery checkerwork as the
means of complying with State particulate matter regulations. In this case
the flue gas is at a high temperature as it passes through the checkerwork
such that little or no arsenic would be condensing and forming particulate
matter. No arsenic particle fallout would occur bee?'se the submicron
particles would not likely be caught by the checkerwork. This method of
particulate control is used in at least two existing arsenic glass plants.
The second method that is used to reduce particulate and arsenic
emissions is the electric boosting process modification. The use of
electric boosting has been demonstrated to reduce overall particulate and
arsenic emissions. One company's experience with boosting indicated that
reductions of from 30 to 60 percent were achievable when compared to
uncontrolled emissions. The use of electric boosting reduce- arsenic
emissions because it lowers the required furnace bridgewall temperature such
that less material evaporation occurs. Also, the offgas temperature would
be reduced, thereby promoting a greater amount of condensation and less
vapor loss.
The arsenic glass plants that use add-on fabric filter or electrostatic
precipitator (ESP) control devices do the best job of controlling inorganic
arsenic emissions. As shown in Table 8-2, ESP devices predominate in this
185
-------
TABLE 8-2. ESTIMATED INORGANIC ARSENIC EMISSIONS AND PARTICULATE
CONTROLS FOR EXISTING ARSENIC GLASS PLANTS6
Plant
RCA
CircleviUe, OH
GTE,
Versailles, KY
Owens-Illinois,
Columbus, OH
Anchor Hocking,
Lancaster, OH
Anchor Hocking,
Baltimore, HD
A
B
C
0
E
F
G
H
I
J
Existing PM Controls that
Reduce Arsenic Emissions
ESP preceded by evaporative cooler,
99% efficient on arsenic
ESP, 95% overall efficiency
ESP, 96% overall efficiency
None
None
FT?, 93% efficient on arsenic
ESP, 95% overall efficiency
EB, 60% reduction on arsenic
EB, 50% reduction on arsenic
None
ESP preceded by an evaporative
cooler, 99% efficient on arsenic
ESP preceded by a spray dryer
and cyclones
ESP's preceded by evaporative
cooler
FF's and EB
Done
kg/hr
0.0036
0.0082
0.011
2.59
0.54
0.023
0.006
1.81
0.45
0.0045
0.0027
0.699
0.027
0.096
0.014
Baseline Arsenic Emissions3
(Ib/hr) Mg/yr
(0.008)
(0.018)
(0.025)
(5.7)
(1.2)
(0.051)
(0.013)
(4.0)
(1.0)
(0.01)
(0.006)
(1.52)
(0.06)
(0.21)
(0.03)
0.032
0.069
0.089
21.8
4.5
0.19
0.026
15.2
3.78
0.038
0.023
3.81
0.237
0.80
0.118
(ton/yr)
(0.035)
(0.076)
(0.101)
(24)b
(5)b
(0.21)b
(0.028)
(16.8)b
(4.2)b
(0.042)b
(0.025)b
(4.2)
(0.26)
(0.88)
(0.126)b
Emissions expressed as total elemental arsenic.
For plants where no information was available on the hours of
CF7 - fabric filter
ESP « electrostatic precipitator
EB " electric boosting
PM • particulars matter
No controls on arsenic glass furnaces.
"References 10. 27, 28, 29, 30, 31, 32. 33
operation 8,400 hr/yr was assumed.
186
-------
sector of the glass industry. Testing performed by industry and EPA has
shown that ESP devices can reduce arsenic emissions by 95 to
99 percent. ' ' One company noted that its "needle design" ESP is
particularly efficient at removing arsenic and lead from their plant's
furnace exhaust gases. Based on the company's domestic experience with the
ESP and other similar experience in Japan, they believe the needle design to
be more effective in this application than a "plate design" ESP. The only
other particulate control device used in this source category is a fabric
filter. Fabric filter devices have been shown to be at least 93 percent __
efficient at specifically removing arsenic from glass furnace exhausts. '
One manufacturer noted that cost and plant space limitations were two
important factors in choosing an add-on control device, particularly since
these devices are not currently installed specifically for arsenic
control.
As shown in Table 8-2, some facilities use gas cooling devices prior to
their add-on control device for the purposes of lowering the stream
temperature to prevent damage to the device and to reduce the amount of
material that escapes the control device as a vapor. In the plants that use
gas coolers, water evaporative-type coolers are commonly found. In one
source test of an arsenic glass plant with an evaporative gas cooler in
place prior to its ESP, the form of the arsenic_in the control device inlet
stream was essentially 100 percent particulate.
8.4 DEFINITION OF EBC OPTIONS
Of the three techniques described in Section 8.3.2, which are
specifically installed for gross particulate control, only one, the use of
add-on control devices, does an effective job of reducing inorganic arsenic
emissions from glass furnaces. Electrostatic precipitator devices are by
far the most predominant add-on control device used in the arsenic glass
plant source category. However, fabric filter devices are also in place,
but they have not demonstrated an arsenic removal performance equal to that
of ESP's. The EPA tests on two glass plants, one with a fabric filter and
the other with an ESP/gas cooler, demonstrated that the ESP system was
99 percent efficient for arsenic removal, while the fabric filter was only
about 93 percent efficient. ' Despite operating at a lower temperature
(138°C compared to 204°C), the fabric filter system yielded an outlet
arsenic concentration that was six times higher than that measured from the
ESP/gas cooler system. ' However, any comparisons between the use of
ESP's and fabric filters to control arsenic emissions should include an
examination of the type of glass produced, as this will influence the type
of arsenic-containing particles to be collected. ESP's have been proven to
be effective at controlling lead glass particulates, while fabric filters
are more effective on borosilicate particulates due to the higher electrical
resistivity of the borosilicate particles. There are, however, plants with
furnaces producing multiple types of glass including lead, borosilicate, and
aluminosilicate that are controlled by ESP's.
187
-------
Despite the fact that both ESP's and fabric filters are effective at
controlling glass furnace arsenic emissions, ESP's were selected as the EEC
for the purposes of the analyses in this report. In the arsenic glass
plants examined ESP's predominated and they demonstrated higher arsenic
removal efficiencies. The selection of ESP's as EEC does not preclude the
use of fabric filters as EBC in situations where the glass particle
resistivity hinders the effectiveness of ESP use.
In addition to applying an add-on control device after the melt
furnace, it is recommended that the EBC option include the use of a gas
cooling device to reduce the amount of arsenic emitted in vapor form. Tests
at a typical arsenic glass plant equipped with an evaporative cooler
demonstrated the effectiveness of cooling the off gas stream in that only,...
0.1 percent of the arsenic entering the plant's ESP was in a vapor form.
The temperature of the furnace offgas stream was not measured in this test,
but the temperature at the inlet to the plant's ESP was about 209°C (408°F).
As discussed in Chapter 2, at this temperature significant amounts of
arsenic trioxide would be expected to be in a vapor form. However, because
of the ready conversion between oxidation stages of arsenic trioxide (As_0.)
and arsenic pentoxide (As^Cv), the less volatile, pentavalent arsenic oxide
may predominate in the offgas stream. Arsenic trioxide is oxidized to
arsenic pentoxide initially in the glass melt by other melt components
including nitrates. At the elevated temperatures of the glass melt the
arsenic pentoxide releases oxygen to perform its clarification function and
arsenic trioxide results. The extent to which this series of reactions is
carried to completion could determine the levels of As?0 and As_0,. in the
furnace offgases. One source, in fact, reports that a borosilicate arsenic
glass furnace emits primarily arsenic pentoxide. The possible
predominance of the pentavalent arsenic oxide may explain why essentially
all of the arsenic was in the particulate form even at 204°C (400°F).
The recommended EBC option to reduce inorganic arsenic emissions from
glass plants involves cooling down the melt furnace offgases by standard gas
cooling devices to minimize the amount of vaporized arsenic, and then
directing the arsenic particulate matter to an ESP control device. Because
of a lack of test data on glass plants with gas coolers, it is difficult to
estimate the optimum cooling temperature that needs to be achieved. The one
plant that has been tested uses a cooler to lower the gas stream temperature
to about 204°C (400°F). An analysis of the sampling train filter and
impinger sections indicated that over 99 percent of the arsenic was found in
a particulate form. Based on the information received during this study,
three of the arsenic glass plants given in Table 8-2 are using EBC including
RCA-Circleville, OH and plants F and H. Five other plants including
GTE-Versailles, KY, Owens-Illinois-Columbus, OH, plant A, plant B, and plant
E have controls in place that reduce arsenic emissions to levels comparable
to plants equipped with EBC. Because EBC is not in place at the majority of
arsenic glass plants, model plants need to be developed in order to estimate
the typical cost incurred by various size glass plants to install EBC.
188
-------
These model plants and the level of arsenic emissions under EBC will be
developed in the next section.
8.5 DETERMINATION OF MODEL PLANTS, EBC EMISSIONS, AND THE INCREMENTAL
IMPACTS OF EBC
8.5.1 Model Plants
Model plants for the glass manufacturing source category were developed
based on data from the background information document (BID) for the glass
manufacturing NSPS and actual plant data submitted by the involved companies
given in Table 8-1. Stack parameters and plant production sizes for the
small and medium arsenic glass model plants were taken from data given on
the pressed and blown - other than soda lime glass plants in the glass NSPS
document. The stack parameters for the large model plant were scaled up
appropriately from the small and medium plants. The 181 Mg (200 ton)/day
production rate of the large facility was included because actual arsenic
glass plants exist that are this size and larger.
The model plant for this source category is defined to be a single
regenerative-type, fossil fuel fired glass melting furnace. The furnace is
operated on basically a continuous basis. Due to the proprietary nature of
information concerning furnace—operations for particular glass types (e.g.,
lead, borosilicate, Pyroceram, etc.), no specific glass is assumed to be
manufactured by the model plant. Instead, it is assumed that either
powdered arsenic trioxide or liquid arsenic acid is added to the batch raw
materials that enter the furnace and an arsenic-containing glass is
produced.
The arsenic emission rates for each model plant are based on actual
plant arsenic emission rates for glass plants that are approximately the
same size as the model plants. This simple procedure for estimating model
plant arsenic emissions had to be followed because no data were available on
the amounts of arsenic input to the glass making process and on the amounts
retained in the glass for the various plants in this source category. The
glass companies hold these data to be business proprietary. Without these
data a material balance approach to estimating model plant arsenic emissions
was not possible. The arsenic emission rates and emission stack parameters
of the model plants are given in Table 8-3. The existing arsenic emission
rates in Table 8-3 are assumed to be occurring as a result of no gas cooling
equipment and no add-on control devices being applied. The glass model
plant is, however, assumed to be in compliance with applicable State
particulate regulations. These assumed conditions represent the conditions
of existing glass plants that have the highest levels of arsenic emissions.
In the following sections the environmental, energy, and economic
impacts of the recommended EBC option on the model glass plants will be
assessed. For the glass plants in Table 8-2 that have poor arsenic control,
189
-------
CO
pi
Ed
H
W
T
PS
^
P-l
H co
53 53
55
53 Ed
M CO
Pi Pi
S3 <
H
U U
<; M
pl^ *^
bZD ^i
53 p
< pi
s o
CO M
CO
*ti O
hJ 53
0 r^
T3
, -l "^ —
C -~ < ^3
O /^- rH
•H C Q) >-^
4-1 O C
3 N-' rH J=
•T3 CD ~-~
O M CO 00
^ S co ,^-
p , pQ
^-^
0
CN
•
O
**— ^
i— t
ON
O
O
y-"^.
v3-
O
•
O
*^s
OO
r-H
O
O
/— v
CN
O
•
O
x^/
CN
ON
O
O
•
O
M
Cfl
C
o
•H
ca
CO
•H
Ed 'M'
fCt
CJ **^^.
•H ^D
(2 *"*"t
d) s^x
co
M V-i
^ >-Tn
•*^^..
a oo
PQ ^M
Ed
O
O
O
oo m r- o -a-
* 00 * £^ ^Q
*^ ^^x v^ ^^ ^^^
*«-^ >s-^ i^s
a^ cs
csi * — i r-* •
• lO • P"* O
CN cs CM -w^
CO ON ^
-a- • t-H
X^ X-/ Sw/
CN ON
CN • O P"» \O
• in • r^* •
CO CN *™ ^ "vf P^
CD
rH ^^
,£) 4-1
CO x-» «4-(
0 4J 4J v-' /-x
•H 3 <4-l tl<
c & v-' e o o
C1J 0 CO
^ *£2 4-) 4-J B
1— 1 *^^ ^^ ^ »
co c3 o oo B * **
4.J O rH »H CO CX CU
C •H M v^ CU *rH B 4^
UJ 4J O SC & Q) CO ^^
B O PQ M H Pi B
CD 3 td nG »i
M CO CO Ed ft.
^-^
0
Vw'
t-H
,
O**
^-s.
O
x^
»— 1
*
ON
/•— .
O
*W-^
i»H
•
&\
O
'-
CO
^ Ou
O M-l
rH x^
Eb
190
-------
estimates will be made of the potential reduction in arsenic*emissions
achievable from the application of the EEC option.
8.5.2 Incremental Impacts of the EBC Option
8.5.2.1 Environmental Impacts. The atmospheric arsenic emissions
resulting from the application of the EBC option to the model glass plants
are shown in Table 8-3, along with the incremental arsenic reduction
attributable to EBC. The potential atmospheric arsenic emissions resulting
from the application of EBC to the actual arsenic glass plants are estimated
in Table 8-4. Currently, the degree to which plants control arsenic is
heavily dependent on the stringency of that State's particulate matter
regulations. As shown in Table 8-4, several plants have control equipment
in place for particulate that results in arsenic emissions comparable to or
less than those from plants designated to be using EBC. In these cases it
is assumed that arsenic emissions under baseline control and EBC are equal.
Arsenic emissions reductions resulting from the use of EBC are estimated to
be significant in plants that are uncontrolled, including Anchor Hocking -
Lancaster, OH and Baltimore, MD, plant C, plant D, and plant J and in plants
that are controlled but have high temperature exhausts, including plants G
and I. In plants G and I only gas cooling devices would be required to
reach EBC.
No adverse water pollution impacts are assumed to occur from the use of
the BAT control option. Neither the ESP nor the gas cooling device will
produce wastewater. Water used in the gas cooling device would be
completely evaporated during normal operation of the device.
The arsenic-containing particulate collected by the ESP could present a
solid waste impact from EBC, however, it is assumed that this solid waste
material is recycled to be used as a batch feed raw material. This
technique of handling the ESP catch is currently practiced by plants in this
source category. ' Therefore, it is assumed that the FBC option would
not create a solid waste problem.
8.5.2.2 Energy Impacts. The largest energy impact resulting from the
EBC option involves the electrical energy consumption of the ESP control
device. The annual energy requirements of the model plants for ES?
operation are as follows:
- large plant: 1940 Gj (5.4 x 10^ kWh)
- medium plant: 970 Gj (2.7 x 10 kWh)
- small plant: 480 Gj(1.3 x 10 kWh)
The derivation of these quantities is summarised in Appendix A-8. The
energy requirement numbers presented here represent the incremental energy
impact of the EBC option.
191
-------
^J
pa
OS
Ed
Q
2!
3
CO
25
O
CO
1 — 1
^*
Ed
M
2
S
Pu
CO
CO
>J
CJ
CJ
h~^
2
W
CO
PS
<
•
1
00
W
hJ
pa
H
c
o u *-.
•H pa n
u W A
o --.
3 O J2
O >»•
eti cu
^ ja
CO CO
C 3
CJ .0 *•*
S i-l £
V4 4J 60
O W ^^
B
0 3
pa
w
CO
a
o
1-1
CO x-v
-rt £
[d J3
i-l
O ^"^
<
Mi co
•H as
0) > O O
i— 1 CO C
i-l U i-l •
•H i-l i-H U)
> iH fH 3
cu 1-1 M .a
|-i ffl 1 E
U CO CO 3
•» W4 • la C. I— I
S-H td a) mo
CJ H > SO
a: u O
^v
(**
*^
U*^
\^
ON
^f
•
{S|
^s
CO
o"
O
0
r**
•
m
m
01
M
OOP:
C O
•H
^ M
O M
O a)
S 4J
CO
i-i co
O U
.c e
u co
e -J
^^
m
—4
•
f~4
"**^
O
CM
in
d
>•""*• ^**
oo —
•3- m
o o
o o
CM CO
CM CM
O O
d d
f^
/-^ m
CM O
-4* d
CO
-a- «M
m o
O 0
A
•H
*s «
U 01
O t-i
* i
lJ -H
O *J
*£• i~4 ^
U CO
e pa
,*^.
<"*^ x"^ /-i> /—^ C^
^ *J} «^ CT^ CM
00 CT> *^ -*4 O
• • • • •
CO O —i O O
O 0 O O
00 CO
«C CO vO 00 «— 4
p^ *ir ^o c^ co
• • - • • •
— o o o o
,— ,
^•N /— * /— x CM
CO/-s/»-x-.\O/— >/--00 — <
SCO KO —4 O
o o o o
ooooooooo
u"i r- — 4 \o
vO CM CO *y CM vO f**« OO CO
§t~--HOOCOCMOO
OOOOOOOO
dddoddood
f—\ S~\
OOOOomOcMO
• ••*•• •*•
OST-4OO--OOO
in r>-
VO *3" CM P^ \D
-------
8.5.2.3 Control Cost Impacts. This section presents the capital and
annualized costs associated with the application of the ESP/gas cooler EEC
option to the glass manufacturing model plants. The actual costs for EBC
calculated for the model plants represent the incremental costs of control
because the EBC option is being added to and not substituted for any other
means of control. The capital and annualized costs of EBC are presented in
Table 8-5. ' The bases for the capital and annualized costs are given in
Appendix B-8.
8.5.2.4 Economic Impact Resulting From EBC. An economic impact
analysis was conducted for the glass manufacturing model plants to estimate
the potential impact of requiring EBC. Using the cost data in Table 8-5 and
the methodology given in Appendix C, revenue and product price increases,
required by glass plants to maintain the same net present values before and
after the installation of EBC, were estimated.
For the purposes of determining a cost for product glass the following
classifications between glass model plants were assigned:
- small model plant produces only glassware
- medium and large model plants produce only television tube and
bulb blank glass.
This type of classification system was required due to the inability to
estimate a single glass product price, because glass plants using 'arsenic
produce a multitude of final products.
The annual revenue increases required by the glass model plants range
from $350,000 in the glassware model plant up to $620,000 in the large
television tube and bulb blank model plants. The greatest unit price
increases were $22 per Mg ($20 per ton) in the glassware model plant and
$15 per Mg ($14 per ton) in the small television cxib^ ind bulb blank model
plants. The glassware model plant and the medium bulb blank plant
demonstrated the greatest product price increases, 1.7 and 1.6 percent,
respectively. The lowest product price increase was 0.7 percent in the
large television tube model plant. Table 8-6 summarizes the revenue and
price impacts of EBC on the glass manufacturing model plants.
To test the sensitivity of the economic impact results to the various
input data, a sensitivity analysis was conducted for the glass model plants,
assuming simultaneous 15 percent decreases in baseline price ard annual
output and a 15 percent increase in the weighted cost of capital.
In the sensitivity analysis the glassware model plant still has the
greatest unit product price increase at $27 per Mg ($25 per ton). The large
television tube model plant had the lowest product price increase at $12 per
Mg ($11 per ton). The percent price increase ranged from 1.0 percent, in
the large television model plant, to 2.4 percent in the glassware model
193
-------
CO
H
r4
p-l
r4
O
s
O
2
M
rv]
|
<:
£
&
^^
^ri
to
CO
3
o
Pi
o
rS
rj
pq
W
O
CO
H
CO
O
O
o
N
M
<£
C3
^
C
0)
u e
W -H
3
tr
w
o
•H
4-1
Cfl
01 H
a. co
O 0)
^-*
14H
O rJ
01
CO P-l
3
O
TJ
01
0) rH
N rH
•r-4 3
CO D.
4-1 >1
C cfl
CO T3
rH •>--.
P-( ^*^
CO
rH C
01 O
T3 W
O ^-^
00
a
y^\
00
"^
O
X«^
CM
CM
•
o
0
0
m
*
r- (
00
ro
•co-
C
0
3
rH
•H
S
m
•
_i
•CO-
^^
!-i J3
01
^H T3
O 0)
o 3 p-
CJ O CO
rH W
CO rH
CO O C
CJ tu CO
O
O
^f
00
s"**
O
m
\«/
•
^^
V-t 43
01
rH T3
O 01
o 3 cu
CJ O CO
rH W
CO rH
Cfl O C
O fl* CO
O
o
*^^
00
^— S.
o
o
I— 1
s^/
r-.
o
CJN
^-^
CN
.
»— I
s^
VO
oo
•
O
0
0
m
A
CN
O
r^
^^
•
1— (
00
•
CJ
pa
W
M-J
O
4-1
CO
O
G
rH
CO
4-1
0
01
g
• Q)
CO )-<
(-( 0
" S
r^ *rt
rH
O 01
T) 43
4J
OO CO
-------
w
C/)
f—a <^
S3 w
w pi
o u
W M
Cd
O U
W H
Pi Pi
M Pn
O-H
W M
Pi S3
I—
O
W 60
H W SS
1-4
O.
en
£8
O
o ss
W M
M W
O O i-H
O* M r-H
W pi O
Pi P-l T3
O
O
CN
CN
W
td (0
> W H
w cn to
Pi
-------
plant. The revenue and product price impact of all the glass manufacturing
model plants are summarized in Table 8-7.
8.6 POPULATION EXPOSURE DATA
The data in Table 8-8 were developed during this study for use in an
air pollution dispersion model to estimate the population exposure to
arsenic from glass manufacturing. The Strategies and Air Standards Division
(SASD) of EPA is using these data in an ongoing effort to estimate exposure
to arsenic from this source category.
8.7 EASE OF STANDARDS DEVELOPMENT
In the event that arsenic standards development is initiated, the task
of arsenic emissions measurement would not present a problem for the glass
source category. The EPA, glass companies, and State and local agencies
have conducted source emission tests at glass plants to specifically measure
arsenic. The methodology for measuring these emissions as they are emitted
to the atmosphere is established, and the methodology has been shown to be
effective and accurate. The EPA reference method generally used to measure
glass plant arsenic emissions is Method 108 or a slight variation of it.
Method 108 is designed for use at non-ferrous smelters where high levels of
sulfur dioxide (SO.) are expected. In these smelter applications solutions
of hydrogen peroxide (H^O,,) are used in the sampling train impingers.
However, glass furnace exhausts have much lower levels of SO , therefore,
water (H_0) is generally substituted in the impingers for the H_0?." '
The arsenic emissions caught by the sampling train are quantified through
Atomic Absorption Spectrophotometry. Speciation of the arsenic materials is
not necessary because no organic arsenic species are evolved from a glass
melt furnace.
196
-------
H
CJ
VD
CO
CM
CO
CO
W
H co
M <;
Z w
00
a z
w M
cd
M w
at
Cu
CO
)-,
cd
00
CSl
CM
CM
CM
CM
O* I—1 iH
W Pi O
pi p., -O
> W
W co
Q 03 O
W CJ T3
M M O
O O
o- o
vO
CO
CO
00*
CO
00
in
r-
oo
6
3
•H
13
0)
41
00
CO
e
•H
T3
CD
s
a»
00
o <
O hJ
y p_j
a)
ca
ca
cd
ca
-------
Cd
OS
H3 U
co a
S S
X D
CdH
a
•H tO
M d
O i-l
T3
2 V
H 0
O 0
CJ
•M
P3
to
-J
CM
•£
JS
"*-•*
,n
^H
N«— '
^
,d
**«^
00
S-*i
j^
42
,0
i— 1
J-t
^3
4_t
d
•r^
o
i— i
CO
4-1
d
o
N
•H
^4
O
33
i— 1
tO
CJ
•H
4-1
^1
CO
<1)
d
SI
0)
S
to
a
00
o
o
0,
v£>
CO
o
o
o
^-^
00
o
o
o
vO
ro
o
o
o
co
co
CU
o
o
P-I
S
1— 1
•
CN
CO
CO
e
CN
•
1— 1
00
CO
-*"
r^
33
o
»
d)
-H
f— (
•H
cu
rH
. U
< n
CJ t-t
CXi O
00
o
•
o
**— '
CM
oo
o
o
•
o
/— s
00
1— 1
o
•
o
CM
oo
o
o
o
CO
CO
cu
a
o
a!
M
~^
•
t--
Q>
vO
g
w^
,»|
•
CO
--4
CM
-a-
\o
^-1
Ky*r
A
CO
CU
-H
r-l
•H
to
•> CO
Cd li
H CU
o >
m
CM
o
•
o
N—^*
t— 1
,-H
o
o
X-N
m
(N
0
o
»-H
t— (
o
o
CO
CO
0)
CJ
o
eu
I
0
f— 4
CO
CO
Q\
•
i—4
CN
-^-
•*
r^
^
CO
•H
O 33
d o
•H
1— 1 «
i-l CO
H 3
co B
d 3
0) ^
S 0
O CJ
^_^
CO
CM
•
o
>*-^
o
I— 1
•
o
t~\
p^
•
m
o«*
m
CM
co
co
CU
o
o
£
J
f»N»
•
*sT
^o
CO
g
w*
m
•
vO
CO
•*
r~-
^
60
d 33
•H O
CJ «
o n
33 0)
4-1
t-l CO
O tO
jC O
u d
d to
o cu
33 V4
o
o -5
JS -I-1
O -4
d to
< (O
•—4
m
0
o
^•^
CO
CM
o
d
^— N
-H
m
0
o
CO
CM
0
o
co
co
cu
CJ
o
PH
B
oo
•
r-t
o
CO
S
oo
•
o
•*3°
vO
•*
<^
^
^ ^ ^ ^
^H \O "^ >— 1
o ^ o o
o o o o
in
^«O CN 00 *^"
o r- -H o
o o o o
* • • *
o o o o
s-*^
CO x-s
t-H /^*\ /•— \ t— <
o o o o
O -* — i O
m
o -H m o
O 00 <3- O
o ^ o o
co co co co
CO CO CO CO
cu cu cu cu
a o o o
o o o o
P-i Cu P-J Cu
G B S
CN O CM
• * *
i-l CO CO
cr\ -* m
vO CM \O
B B B
r-~ co co
• » •
O O\ OO
oo vo o
— ' co m
"* "* "*
^O 00 vD
CO U Q Cd
vO
O
O
O
^
CM
O
o
•
o
s~^
*iO
o
o
•
o
CM
O
0
O
CO
co
cu
0
o
PH
J
^.
•
-^
vO
CN
B
o
•
-^
CM
in
-------
oo
I
oo
cd *-•
CO 43
C ->»
O JQ
to v-'
CO
e
Cd
/-v m m CM
m m co •— i o
o o o o o
Q\
co co \o r*** CD
CM CM —1 O O
O O O O O
co
co
CU
o
o
£
Js
^
CT\
m
e
m
rn
-a-
J
^,
-^
^
o
CO -H
4»i 13
43 0 CU
en 4J cd co
•H 4-> U
co 3 co cd
43 CO HI 00
CU -H C
cj 4-> o a. -H
(3 cd -H 4J
Cd ti 4J 4-1
_l J_l ^ .H
ft a 3 s
^ iw e cu
<~v x-\ 00
vO •-"* O
O O O
• » •
o o o
^-^ ^^ S— ^
t/t ^
r*-fc »«d* ro
CM O O
o o o
0 0 CD
^O i-H O
O O CN
O O O
m
p~ ~* i— i
CM O O>
o o o
o o o
co co
co co
cu cu
o o
0 O
CU Oi
00
c
•H
4-1
CM -rl
43 CO /-s
cu u
33 H-l 4-« O T-l
cd e cu
— 1 1-" 03
(X 3 M
>-^ u-i cd
O
o
O
O
o
o
•
o
O
•
O
CO
CO
a)
o
o
o
•H
C
CU
co
e
cu
cu
.—i
cu
,—i
-------
o
z
3
rS 0
HH CJ
EH ^
•» CX
4-1 u-t
o
o
CU
a
3 cu
O CO
rH -»»
fa S
"i"
<4-i
• CJ
01 CO
4J ^^^
OS
3 CJ
o cu
rH CO
CO
e
• fa1
Cue
pg i ^
CU
H
4-1
•H
^
W t^
o
/— s
" 4J
Li 14-4
cu ^
cu
e
CO
•H
Oe
ts
•> 4-1
4J IM
pi ^^^
00
•rl
CU
s: e
4-1 CU
a e
CO cO
P-i
s~\ ***\
v»O vO
H
*.
cu
rH CO
rH CU
•H rH
> rH
V -H
rH CO
• CJ -CO
CJ -H HO)
Pi CJ O >
CO
•H
0
C
•H
rH
rH
M
1
ca
C
cu
3
O
^
•
^H
-*
^
•
CN
i-H
O
o
CN
M
CN
CM
m
,
o
rH
0*
in
C
O
in
<^
^.^
CN
in
o
-H
O
N^^
CO
•H
CN
S!
O
•»
CO
3
rO
S
rH
O
u
00
C
•H
^^
a
0
53
^|
2
0
C
/-">.
0
CO
«— 1
•
ON
o"
O
O
^>
vO
CN
o"
CO
o
o
f^
f*x
^^
^
r^.
•
NO
rH
CN
m
00
CT^
•
m
CM
32
O
M
Li
CU
4-1
CO
3
C
CD
/-«X4
0
co
rH
•
ON
o"
o
o
CM
CO
rH
«
in
rH
O
O
*^s
p^.
[•«*
^^
m
f>^
•
f«^.
CM
ON
•
CM
rH
CO
CO
CO
O
^j*
sr
/_^
ON
•
m
CO
rH
m
CO
rH
x**'
rH
•
rH
"*
CQ
s^* ^^\ *~*. /~*\
— i m o r-
—1 CM CO «*
-------
M
Crf
82
•< fa
»j ^
o a
CO
W CO
£3
o
1
fa
W
OJ M
3 (3d
e o
•H M
4J O
W
OO
00
W
CO
J-l
0)
^ t
0)
s
CO
1-1
tO
y
^j
to
f». Du
4-1 U-4
y
o
cu
y
3
-------
8.8 REFERENCES
1. Telecon. Brooks, G. W., Radian Corporation, with Drake, R., Glass
Packaging Institute. December 29, 1981. Conversation concerning
arsenic glass plants.
2. Glass Manufacturing Plants Background Information: Proposed Standards
of Performance. EPA-450/3-79-005a. U. S. Environmental Protection
Agency. Research Triangle Park, North Carolina, p. 8-47.
3. Telecon. Brooks, G. W., Radian Corporation, with Simmons, C.,
Mississippi Air Quality Division. January 5, 1982. Conversation on
status of General Electric glass plant.
4. Telecon. Brooks, G. W., Radian Corporation, with West Virginia Air
Control Board. December 31, 1981. Conversation concerning status of
Libbey-Owens-Ford glass plant.
5. Peters, A. The Use of Arsenic with Particular Reference to the Glass
Industry. Glastechn. Ber., Volume 50, No. 12, 1977. pp. 328-335.
6. Crose, P., Acurex Corporation. Arsenic: An Environmental Materials
Balance - Draft Final Report. EPA Contract No. 68-01-6017. March
1981. pp. 3-3, 3-14.
7. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for Emission Standards and Engineering Division, Office of
Air Quality Planning and Standards, U.S. Environmental Protection
Agency.) EPA Contract No. 68-02-3173. May 1982. pp. 3-31 to 3-33.
8. Letter from Cherill, J., Corning Glass Works to Hellwing, G. V., PEDCo
Environmental. June 1, 1981.
9. Telecon. Brooks, G. W., Radian Corporation, with Armstrong, F., RCA.
February 19, 1982. Conversation concerning arsenic usage by RCA.
10. Telecon. Brooks G. W., Radian Corporation, with Riley, L., RCA.
March 5, 1982. Conversation concerning estimates of RCA emissions.
11. Letter from Mosely, G., Corning Glass Works to O'Connor, J., U. S. EPA.
August 28, 1978.
12. Reference 2, p. 3-6.
13. Letter and attachments from Swander, T., RCA to Goodwin, D., U. S. EPA.
September 27, 1978.
14. Letter and attachments from Weikel, P., GTE-Sylvania to Goodwin, D.,
U. S. EPA. August 30, 1978.
202
-------
15. Reference 2, p. 3-16.
16. Suta, B. E., SRI International. Human Exposures to Atmospheric
Arsenic. EPA Contract Nos. 68-01-4314 and 68-02-2835. May 1980.
p. 97.
17. Reference 2, pp. 3-7 to 3-8.
18. Reference 2, p. 4-4.
19. Schorr, J. R., et al., Battelle-Columbus Laboratories. Source
Assessment: Pressed and Blown Glass Manufacturing Plants.
EPA-600/2-77-005. January 1977. p. 24.
20. Memorandum from Cuffe, S., U. S. EPA to O'Connor, J., U. S. EPA.
July 17, 1978. Emissions of Arsenic from Glass Melting Furnaces.
21. Memorandum from Herring, W., U. S. EPA to Cuffe, S., U. S. EPA.
December 27, 1978. Arsenic Emissions from Glass Manufacturing Plants.
22. Telecon. Brooks, G. W., Radian Corporation with Cherill, J., Corning
Glass Works. January 29, 1982. Conversation concerning Coming's use
of arsenic.
23. Reference 2, pp. 4-6 to 4-7.
24. Reference 2, p. 3-10.
25. Reference 19, p. 34.
26. Telecon. Brooks, G. W., Radian Corporation, with Cherill, J., Corning
Glass Works. April 20, 1982. Conversation concerning emission control
devices.
27. Letter from Goebel, G., Kentucky Bureau of Environmental Protection to
Brooks, G. W., Radian Corporation. February 11, 1982.
28. Letter and attachments from McVay, D., Rhode Island Department of
Environmental Management to Brooks, G. W., Radian Corporation.
January 7, 1982.
29. Letter and attachments from Cherill, J., Corning Glass Works to
Brooks, G. W., Radian Corporation. April 7, 1982. (Project
Confidential File).
30. Thalman, M. T., et al., Monsanto Research Corporation. Arsenic Glass
Manufacturing Emission Test Report - Corning Glass Works, Central
Falls, Rhode Island. EMB Report 78-GLS-3. February 1979. pp. 3-9.
203
-------
31. Thalman, M. T., et al.» Monsanto Research Corporation. Arsenic Glass
Manufacturing Emission Test Report - Corning Glass Works, State
College, Pennsylvania. EMB Report 78-GLS-4. February 1979. pp. 3-9.
32. Telecon. Brooks, G. W., Radian Corporation, with Iden, C.,
Owens-Illinois. March 3, 1982. Conversation concerning Owens-Illinois
arsenic emissions.
33. Letter from Armstrong, F., RCA, to Brooks, G. W., Radian Corporation.
March 8, 1982.
34. Reference 2, p. 4-32.
35. Telecon. Brooks, G. W., Radian Corporation, with De Leon, F., Anchor
Hocking. March 23, 1982. Conversation concerning Anchor Hocking
arsenic emissions.
36. Reference 2, p. 6-12.
37. Reference 2, p. 8-58.
38. Neveril, R. B. and M. L. Kinkley, CARD, Inc. Capital and Operating
Costs of Selected Air Pollution Control Systems. EPA-450/3-76-014.
May 1976. pp. 4-85 to 4-86.
39. Reference 30, p. 15.
40. Reference 31, p. 17.
204
-------
9.0 ARSENIC CHEMICAL PRODUCTION
This chapter discusses the arsenic chemical manufacturing industry and
its associated arsenic emissions. Section 9.1 presents a brief source
category description. Section 9.2 discusses the regulations currently
covering the industry and the resulting level of control and associated
arsenic emissions. Section 9.3 presents a discussion of the estimated best
control (EEC) options and Section 9.4 gives the input data to be used in
assessing the population exposure to arsenic emissions from arsenic chemical
manufacturing.
9.1 SOURCE DESCRIPTION
The manufacture of chemicals containing arsenic is an obvious potential
source of atmospheric inorganic arsenic emissions. As many as 25 arsenic-
containing chemicals are produced, of which 18 are inorganic and seven
organic in nature. These chemicals are identified in Table 9-1. The
organic chemicals are included in this discussion because inorganic arsenic
compounds are used in the production of organic arsenicals and may be
emitted during raw materials handling and process reaction steps. Arsenic
chemical production consumes approximately 90 percent of the total arsenic
used in the United States. Of this 90 percent, about 70 percent goes into
pesticides manufacture and 20 percent to wood preservative manufacture.
Nine of the compounds given in Table 9-1 are considered to be of major
importance because they constitute 90 to 95 percent of the production in the
arsenic chemicals industry. ' ' The nine chemicals are as follows:
- orthoarsenic acid
- monosodium methylarsonate (MSMA)
- disodium methylarsonate (DSMA)
- dimethylarsinic acid (cacodylic acid)
- calcium arsenate
- lead arsenate
sodium arsenite
- chrome copper arsenate (CCA)
- ammoniacal copper arsenite (ACA)
The production of these nine chemicals will be discussed in depth in the
following section.
205
-------
TABLE 9-1. ARSENIC COMPOUNDS AND THEIR USE
ARSENIC COMPOUNDS MAJOR USE
INORGANIC
Orthoarsenic Acid Desiccant
Arsenic Disulfide (tri and penta) Textile painting
Arsenic Iodide (arsenous iodide) Antiseptic
Arsenic Pentafluoride Laboratory Research
Arsenic Thioarsenate Scavenger
Arsenic Tribromide (arsenous bromide) Medicinals
Arsenic Trichloride (arsenous chloride) Herbicide
Arsenic Trifluoride (arsenous fluoride) Laboratory Research
Arsenic Trioxide (arsenous oxide, Arsenical Precursor
arsenous acid)
Arsine (arsenous hydride) Semiconductor Industry
Calcium Arsenate Insecticide
Calcium Arsenite Insecticide
Copper Arsenate Wood Preservative
Copper Arsenite Insecticide
Lead Arsenate Insecticide
Sodium Arsenate Herbicide
Sodium Arsenite Herbicide
Zinc Arsenate Insecticide
ORGANIC
Arsanilic Acid (sodium arsanilate) Feed Additive
Cacodylic Acid (dimethylarsinic acid) Defoliant
Fluor Chrome Arsenate Phenol Wood Preservative
Methanearsonic Acid (mono- and disodium salts, Herbicide
MSMA/DSMA)
Methanearsonic Acid (calcium salts) Herbicide
Methanearsonic Acid (ammonium salts) Herbicide
Copper Acetoarsenite (Paris green) Insecticide
206
-------
9.1.1 Orthoarsenic Acid
Orthoarsenic acid (H-AsO.) functions as both an end product and as a
chemical intermediate to the production of other inorganic and organic
arsenicals. As an end product arsenic acid is used as an agricultural
desiccant/defoliant. It is primarily used to desiccate short-season cotton
that will be harvested by a mechanical stripping technique. Desiccation of
the cotton plant is required to prevent plant moisture from staining the
cotton fibers prior^to ginning, and thereby, lowering the quality grade of
the product cotton. ' ' It also has other uses as a glass manufacturing
raw material.
Arsenic acid is manufactured by reacting powdered arsenic trioxide with
nitric acid. The reaction is generally carried out as a batch process in a
variety of reactor vessels. An example process flow sheet is given in
Figure 9-1. Reactor types may vary from pressurized, temperature controlled
vessels to simple stirred tank reactors operating at ambient conditions.
Arsenic trioxide (As^CL) generally arrives at the facility in drums. The
handling and unloading (dumping) of the arsenic trioxide powder constitutes
the primary source of inorganic arsenic emissions from arsenic acid
manufacture. The arsenic trioxide drums are dumped into storage bins or
directly into the materials coveying system. The action of dumping the fine
powder creates fugitive dust emissions. These inorganic arsenic emissions
are well-controlled by existing plants primarily to comply with QSHA 1.
regulations and to recover the arsenic trioxide for process use. ' ' The
majority of plants are also affected to some degree by State particulate
matter regulations. The methods of trioxide emissions control are discussed
in detail in Section 9.2.2.
Because the demand for arsenic trioxide exceeds the supply of the sole
domestic producer (ASARCO), quantities must be imported from Mexico, Sweden,
and England to make up the difference. This situation of tight supply
motivates companies towards arsenic trioxide raw material recovery and
reuse. One major arsenical manufacturer indicated that they would like to
produce more arsenic acid but could not due to the shortages of arsenic
trioxide.
Besides the raw materials handling step, other process steps in the
production of arsenic acid such as the arsenic trioxide-nitric acid
reaction, product purification, and product packaging, may be sources of
atmospheric inorganic arsenic emissions. However, compared to the raw
materials handling step, inorganic arsenic emissions from the process
reaction, product purification, and product packaging steps are negligible
or non-existent. This is true because these process steps involve
handling wet, low temperature arsenic materials, therefore, no dust problem
exists and no arsenic volatilization occurs.
Two companies in the United States manufacture liquid arsenic acid as
their chemical end product. These companies are the Agchem Division of
207
-------
o.
3 W
I .
C
U U O CQ
IV
•o
O »H
(0
-H 00
V4 C
(U -H
*J -T3
eg co
2 O
j
•H O
C -O w
0) -H O
01 U (0
co
03
0)
U
o
VJ
Cu
C
o
o
3
O
a.
t3
•H
O
O
•H
CD
0)
t-i
M
•H
U
-------
Pennwalt Corporation in Bryan, Texas and the Voluntary Purchasing Group
(VPG) in Bonham, Texas. ' In 1980, VPG produced about 2.27 million
liters (600,000 gal) of arsenic acid, while in 1981 Pennwalt manufactured
approximately 2.65 million liters (700,000 gal) of the same compound.
Three other companies, Koppers Company, Mineral Research and Development
Corporafion, and Osmose Wood Preserving Company all manufacture arsenic acid
as a chemical intermediate to the production of arsenical wood preservatives
(chrome copper arsenate and ammoniacal copper arsenite). The arsenic acid
produced by these companies is used.for in-house captive purposes only and
is not sold on the open market. ' '
9.1.2 MSMA, DSMA, and Cacodylic Acid
The production of MSMA, DSMA, and cacodylic acid arsenicals is
important in assessing inorganic arsenic emissions because arsenic trioxide
is in most cases the starting material for these end product chemicals.
MSMA and DSMA are selective herbicides for post-emergent weed control of
grassy weeds in cotton, citrus, turf, and non-crop areas. Cacodylic acid
is a non-selective, post-emergent herbicide that is chiefly used in lawn
renovation and weed control in non-crop areas. Cacodylic acid also has
significant uses as a defoliant particularly for machine-picked cotton in
California and Arizona.
These three chemicals are frequently grouped together because they are
manufactured by a relatively similar process. The major difference in
determining the end product is,the number of reaction steps and the
utilization of side-reactions. The process starts by manufacturing DSMA by
reacting arsenic trioxide with sodium hydroxide and water to form sodium
arsenite, which is in turn treated under pressure with methyl chloride to
form DSMA. The DSMA may then be removed and sold as a herbicide or further
reacted to form MSMA. To produce MSMA sulfuric acid is added to DSMA to
adjust the pH. This material is then centrifuged to remove sodium chloride
and sodium sulfate waste salts and the resulting MSMA solution is
concentrated by evaporating the water.
To produce cacodylic acid, sulfur dioxide is first added to the
methanearsonic acid (acidified DSMA) to reduce this to the trivalent form.
A second methylation with methyl chloride is then performed yielding
cacodylic acid. As with MSMA, waste salts are removed, the filtrate
evaporated to the desired-concentration, and unreacted trivalent arsenic
species are oxidized. ' A schematic of the overall production process
for DSMA, MSMA, and cacodylic acid is given in Figure 9-2.
The primary points of potential inorganic arsenic emissions during the
manufacture of the organic arsenicals are the arsenic trioxide handling area
and the sodium arsenite process unit. Only two companies manufacture these
organic arsenicals such that inorganic arsenic is emitted. These companies
are Diamond Shamrock Corporation (Agricultural Chemicals Division) in Greens
Bayou, Texas and Vineland Chemical Company in Vineland, New Jersey. Diamond
209
-------
o a
K4J J->
IT, f<
n n
^ w
TJ
•H
a
•a
o
o
Cfl
W
Q
Vj
O
u
•H
U
0)
jr
a
«
c
o
•H
U
O
3
t3
O
VJ
Pu
(N
I
CT\
4)
^
3
«l
•H
fe
210
-------
Shamrock's primary product is MSMA with very little DSMA and no cacodylic
acid being manufactured. Vineland produces all three chemicals and is the
only domestic manufacturer of cacodylic acid. Inorganic arsenic emission
controls are in place at both facilities and these will be discussed in
detail in Section 9.2.2. Another company, W. A. Cleary, Incorporated in
Somerset, New Jersey produces DSMA, but not by the method given in
Figure 9-2. Instead Cleary buys MSMA and chemically reacts it to transform
the compound back to DSMA. The company has been manufacturing DSMA this way
for about ten years. Representatives of Cleary indicated that the company
had realized early the health hazards and environmental implications of
handling inorganic arsenic substances and had made the decision to not
manufacture DSMA in the classical manner.,. No inorganic arsenic emissions
are associated with the Cleary facility.
9-1.3 Calcium Arsenate, Lead Arsenate, and Sodium Arsenite
Calcium arsenate, lead arsenate, and sodium arsenite are inorganic
arsenicals that function as insecticides, herbicides, and pesticides for a
variety ofApplications. The primary uses of each compound are given in
Table 9-2. Following a thorough survey of the arsenical pesticide
industry it was determined that there are no domestic producers of lead
arsenate, one producer of calcium arsenate, and only one producer of sodium
arsenite.
Calcium arsenate was manufactured by Woolfolk Chemical Works in Fort
Valley, Georgia up until 1974. ' They produced calcium arsenate by
starting with dry arsenic trioxide, which they reacted to form arsenic acid,
and then used the acid to form the arsenate. In 1981 the company decided to
start up production of calcium arsenate again, however, this time as a
liquid product process and not a dry one. Delays were incurred in getting
the facility on-line in 1981, and the most recent Woolfolk estimates
indicate that startup will not occur until the fall of 1982. The process
to manufacture calcium arsenate as a flowable product will proceed by
reacting purchased liquid arsenic acid with lime and running the mixture
through a wet grinding mill. As a result of the wet processing involved,
no inorganic arsenic emissions are expected from the facility.
The one current ..manufacturer, of sodium arsenite is Sharpe Chemicals in
Burbank, California. » » » » Sodium arsenite is produced by reacting
arsenic trioxide, sodium hydroxide, and water. Sharpe only manufactures
sodium arsenite on a bench-scale basis for university teaching and research
purposes. Production is intermittent and only 0.45 to 2.3 kg (1 to 5 Ibs)
are made during anyfisingle batch. Total annual production is less than
45.4 kg (100 Ibs). Because of the size of Sharpe's production and the
essentially zero arsenic emissions from the company's facility, it was
excluded from further examination in this study.
The most recent commercial producer of sodium arsenite was found to be
Colorado Organic Chemical Company in Commerce City, Colorado. They
211
-------
TABLE 9-2. USES OF ARSENICAL PESTICIDES, INSECTICIDES, AND HERBICIDES
COMPOUND
CROP/SITE
OBJECT OF USE
Calcium Arsenate
44 different crops,
flowers
Grapefruit
Poultry houses
Turf
Snail bait
Control acidity
Fly control
Poa annua, soil
insects
Lead Arsenate
Grapefruit
Control acidity
Sodium Arsenite
Grapes
Livestock quarantine
Measles, dead arm
Texas cattle fever
tick
212
-------
manufactured small batch quantities of sodium arsenite for other companies
to use as an input raw material to the production of organic arsenicals.
However, the company has ceased production of the arsenite and a company
official has indicated that they do not plan to resume production.
Therefore, Colorado Organic was also excluded from further consideration.
9.1.4 Chrome Copper Arsenate and Ammoniacal Copper Arsenite
Until recently three arsenical wood preservatives, chrome copper
arsenate (CCA), ammoniacal copper arsenite (ACA), and f luor chrome arsenate
phenol (FCAP), were produced for wood fungicide purposes. According to
several sources FCAP is no longer being-produced, and therefore, will not be
considered further in this document. ' '
Of the two remaining arsenical preservatives, CCA is by far the most
important. The CCA preservative is manufactured by reacting arsenic acid,
chromic acid, and copper oxide. Five companies were identified that are
producing CCA including: Koppers Company, Conley, Georgia and Valparaiso,
Indiana; Mineral Research and Development, Concord, North Carolina;
C. P. Chemicals, Sewaren, New Jersey; Osmose Wood Preserving,-Memphis,
Tennessee; and Chemical Specialties, Valdosta, Georgia. ' ' ' ' As
discussed in Section 9.1.1, three of these companies, Koppers, Mineral
Research and Development, and Osmose Wood Preserving, produce their own
arsenic acid to use as a reactant for their CCA production. For this reason
these three companies present the greatest potential for inorganic arsenic
emissions because they handle the arsenic trioxide raw material. The
control mechanisms used by these companies to reduce arsenic trioxide
particulate emissions are discussed in Section 9.2.2. C. P. Chemicals and
Chemical Specialties do not produce their own arsenic acid. They purchase
liquid arsenic acid, and therefore, do not have the problem of arsenic
trioxide dust emissions.
Ammoniacal copper arsenite is onlynproduced by one company,
J. H. Baxter of San Mateo, California. ' The use of ACA is almost
entirely limited to the West Coast (California, Washington, and Oregon) as a
fir wood preservative. The ACA preservative is manufactured by reacting
arsenic acid and copper oxide. Baxter also buys liquid arsenic acid to
perform all of their production. By analogy to the C. P. Chemicals and
Chemical Specialties processes above, inorganic arsenic emissions would be
negligible with this process.
All of the six companies in the arsenical wood preservative manufac-
turing sector have the potential to emit inorganic arsenic in the form of
their final product (CCA or ACA) during process reaction, product storage,
or product packaging steps. Again, however, arsenic emissions from these
process steps would be negligible or nonexistent because only wet, low
temperature materials are involved.
213
-------
9.1.5 Summary of Producers
Although there are 14 companies involved in the production of the nine
important arsenical compounds (given in Section 9.1), only seven of these
companies handle dry powdered arsenic trioxide and have the potential to be
significant inorganic arsenic emission sources. These seven are listed in
Table 9-3.
9.2 INORGANIC ARSENIC EMISSIONS OCCURRING UNDER EXISTING REGULATIONS
9.2.1 Regulatory Impacts
The primary environmental regulations affecting atmospheric arsenic
emissions from arsenic che'mical plants are the particulate matter
limitations discussed in Section 2.2. These types of regulations are
applied because the inorganic arsenic emissions from plants in the source
category are arsenic trioxide particulate dust. The particulate matter
regulations generally take the form of an hourly emission limit based on
process weight throughput.
For each of the eight plants given in Table 9-3, existing emissions of
inorganic arsenic particulate matter are dramatically less than the
allowable emissions dictated by current State particulate regulations. To
illustrate this point, Table 9-4 presents the estimated existing and
allowable inorganic arsenic emissions from several plants in this source
category. All eight plants are in compliance with their applicable
particulate regulations.
Though important, the particulate emission limitations are not: the main
factor motivating inorganic arsenic emissions control at arsenic chemical
plants. Instead, the OSHA regulations on airborne workplace inorganic
arsenic play., the. primary role in reducing atmospheric arsenic
emissions. ' ' ' ' The specific limitations of the OSHA inorganic
arsenic standard are detailed in Section 2.3. The OSHA rules have motivated
plants handling arsenic trioxide raw materials to install systems capable of
capturing, collecting, and conveying trioxide dust to an appropriate control
device in order to comply with the workplace limit of 10 ug/ra of inorganic
arsenic. Some plants in this source category have been closed down because
of the inability to keep the workplace inorganic arsenic levels below the
OSHA limit. ' One facility, Pennwalt in Bryan, Texas, performs voluntary
monthly monitoring of arsenic levels to insure OSHA compliance.
9.2.2 Baseline Inorganic Arsenic Controls and Emissions
The arsenic chemical plants examined in this report are in compliance
with applicable EPA, SIP, and OSHA regulations, therefore, baseline arsenic
controls and emissions are equivalent to currently existing controls and
emissions. The control systems used in existing arsenic chemical plants
consist of an emission capture system and control device. ' ' '" The
214
-------
CO
o
M
CO
CO
1— 1
JVfl
CJ
M
W
CO
<
CJ
2j
o
Pi
O
M
H
tjP.
^E
CJ
M
M
CJ
M
CO
13
M
&
CO
H
PM
CJ
M
SC
CJ
CJ
M
W
CO
Pi
^^
•
7
w
M
<3
H
CJ
M
O
Pi CO
0 &
S3 O /-•>
M M l-i
CO >,
rJ CO ~-
Q S »H
23 ,S **^
55
,
W Ed 00
H CO ^
^>
c
CO
CU
a
o
CJ
en
Ju
cu
CU
Cu
0
/**\
m
en
o
o
^w^
**o
r— 1
O
O
CL>
•H
JJ
CO
g
cu
en
0)
^
" P-<
*o
o
o
•*^
o
CJ
J2
rQ1
^j
O
CJ
o
u
cu
M
o
.e cj
o
M u'
co C
cu cu
en a
0) Cu
erf o
rH
rH CU
CO >
\-t CU
cj o
•rH o£J
2d
^^
o^
•
vO
CO
NwX
m
•
a\
ro
0)
•H
*4_j
CO
S
cu
CO
cu
M
PM
T3
O
O
<^
O
CJ
12
£_|
Cfl
•H
f^
Cu
a
cu
S
•
o
c
h-1
A
•a oo
o c
O -H
^ ^
}_t
cu cu
CO CO
O CU
a >-<
CO CM
O
/•"•v
CN
i— H
0
•
o
••— '
vO
m
o
0
o
T3
•H
CJ
^J
o
•H
C
CU
CO
}M
•
rH
CO
a
•H
a
CU
CJ
-a
c
cfl
iH
cu
C
•H
t>
x1^
O
<]•
O
•
o
**••••'
oo
1 — 1
o
•
o
*73
•H
0
^
CJ
•H
C
cu
CO
^J
<^
X
H
,
e
cfl
f!
C
O
«
Cu
O
a
00
c
>, -H
)-i CO
cfl cfl
•U X
C a
3 VJ
rH 3
O CM
215
-------
TABLE 9-4. COMPARISON OF EXISTING AND ALLOWABLE
INORGANIC ARSENIC EMISSIONS14'31'32'33
PLANT ALLOWABLE EMISSIONS EXISTING EMISSIONS,
UNDER SIP PARTICIPATE REGULATIONS,
Mg(Tons)/yr Mg(Tons)/yr
Osmose Wood Preserving
1.13 (1.25)
0.039 (0.043)
Vineland Chemical
1.97 (2.19)
0.0000025 (0.0000028)
Mineral Research &
Development Corp.
4.14 (4.6)
0.000016 (0.000018)
216
-------
purpose of the capture system is to collect particulate matter that can be
emitted during the handling and processing of powdered arsenic trioxide raw
material. The control device removes the collected arsenic particulate from
the exhaust gas stream, and the material is eventually returned to the
arsenic chemical process.
The eight plants of interest in this source category use a similar
system to capture arsenic trioxide particulate emissions, though the exact
configuration of the system varies slightly from plant-to-plant. The basic
capture system involves the use of hooding or total enclosures to collect
the emissions, which are in turn pneumatically conveyed to a control device.
The arsenic trioxide raw material generally enters the plant in large drums
which are opened and dumped by a mechanical arm in either an enclosed room
or under a hooding system. Arsenic trioxide dust collected from these
operations is vented to either a fabric filter or wet scrubber control
device. Empty raw material drums are cleaned by vacuum pick-ups and then
washed out under a collection hood to capture any remaining arsenic trioxide
dust that could be displaced. Drum wash waters are returned to the process
to avoid discharging any arsenic trioxide to the environment.
Of the eight plants with the potential to emit arsenic trioxide dust,
three use fabric filters to control emissions, four use wet scrubbers, and
one uses a fabric filter followed by a wet scrubber. Arsenic trioxide
particulate emissions have been reduced by an average_of,.99.5 percent in the
three plants using fabric filter control devices. ' ' The collected
arsenic trioxide is returned to the process. In addition to its primary
fabric filter control device, one plant also uses a high efficiency
particulate air (HEPA) filter on the exhaust gas stream from its fabric
filter. The efficiency of the HEPA filter, as estimated by the plant, is
99.9+ percent. When required, the HEPA filter is washed out and the wash
waters returned to the arsenic chemical process.
Although no test data are available on the arsenic removal efficiency
for the four plants using wet scrubbers, an examination of these plants'
inorganic arsenic emision rates in Table 9-5 verifies that the scrubbing
method of control is comparable to fabric filters in this
application. ' ' The spent scrubbing solution resulting from this
control method, which contains arsenic, is returned to the arsenical process
thus eliminating potential arsenic water pollution.
The remaining arsenical facility operates a fabric filter, estimated to
be 99 percent efficient, to control arsenic trioxide emissions from the
arsenic trioxide handling operations. The exhaust stream from the fabric
filter is sent to a wet scrubber before final exhaust to the atmosphere.
While an across-the-device efficiency has not been determined for the
scrubber by testing, the efficiency was estimated to be at least 95 percent
at its installation.
217
-------
CO
z
o
M
CO
CO
l-l
g
CJ
1— 1
z
CO
el
^C
u
p*1
<
Oi!
0
z
M
Q
^C
CO
t-a
o
el
H
Z
O
CJ
U
z
H
CO
LJ
w
fe
o
^x
<
CO
m
1
CT>
Ed
nJ
CO
^C
*•
1
^
*
CO
1
O
o
Ed
C*4
CJ
i— i
CJ
el
o
CO
i-J
CJ
1— 1
EC
CJ
CJ
Ed
co
(X2
<
Ed
I—I
I-**
H
Z
M
O /-s
H-t «O
g C
t/J 60
•k
O >•
Irt ^
Z —
u a
c z
z <
IK
O
en
O
en .a
en iH
i— i ••*'
3 60
^
u
z -
Irt >•
H J
en &
Z §
w S
Ucl S
Ǥ
U > W
Z U U
Irt O Irt
H to
C/l J tK
5Sf
8^4
|L4
&£
gg
3- H en
z O z
H " 8
M Z
H
2
Cu
o
vO
O
O
CN
O
o
^
o
o
0
o
o^
M^
O
O
§
o
OS
OS
u
OJ
3
CJ
tn
4J
eu
3
eg
eg
O
eg
c
u
•a
B
O
O
jj |*
CJ
o •
I* 3
S X
•c to
en 09
TJ CO
E E
§jj
OJ
eg i-i
•rt U
a
m O
00
SS
CM O
O O
o' o
s~\ S~*
CM •»
O O
O O
o o
o o
o o
•rt CN
So
0
§o
0
0 0
U l-l
OJ O
.a .0
•§•§
t-i i-i
CJ U
en tn
4J W
01 OJ
3 3
CO eg
eg eg
a eg
*•» 4-1
B C
cy 01
•O T3
•rt -rt
B B
o o
u u
>> z
C >~
eg
B-« O
O O to
O i-l
• eg
b CJ CO
o i-i o.
a E --i
c. o re
o o >
in
en
O
es
o
o
00
o
§
o
o
as
0
§
o
o
o'
l-l
eu
•rt
h
U M
•fi Os
eg os
Ik
ta
irt
to
•rt . B
i-i n eu
OJ T3 eg JJ
u B eu u to
eg 08 w to x
60 T2
•O B 60 Q.
C) -rt B E 3
CO irt -rt 3 1
O T! 0. 3 J!
irt B B U CJ
u eg 3 eg -rt
S X -0 > e.
_£.
O •
eg B
OJ 0) CJ
n B Z
0) O.
as O -
•rt -0
•rt eu M
eg > o
l-l CJ CJ
01 Q C
B o
•rt « CJ
°I
SO
00
in
Os
tn
CN
^
O
o'
1— 1
o
o
ffi in
OS Os
«•» to -^
l-l 1-
0X0
•rt -U 3
(K eu l-i
3 CJ
u o tn
•rt »rt
l-l -rt 4J
JO O 01
eg u 3
[K
CO
•rt
•rt . B
u co eu
oj "O eg u
4-* B OJ AJ CO
eg eg u. co X
E ec 3 co
60 T3
•0 E 60 Q.
01 -rt C B 3
CO rt -rt 3 i
O -O C. 3 ^
rt C S O O
E x •a > a.
Id
»
CJ
E
Irt
1 ff*
O iH -
3 e^s
CJ OJ £
to eo ex
S i-i 0)
to e. £
O
es
O
O
sC
m
°.
0
in
^
1
O
r^.
O
|
O
eg
eg
a
.rt
eg
•rt
B
01
•o
•rt
irt
B
O
U
eg
eg
O
eg
•rt
iJ
E
OJ
•a
•rt
1
U
o
B
Irt
X
- H
u
iH -
eg B
3 eg
C >,
E u
ci as
A-
*-s
in
O
in
CM
o
o
o
Cjs
o
o
o
o
o
o
o
sr
o
c
o
o
o
o
o'
3^
OS
OS
Os
IM
01
.c
CJ
en
1 1
OJ
3
*
CO
4J B
rt s eu
rt -rt 4J
eg o 4j co
CX CD X
B to 3 CO
O rt 60t3
eg E a.
60 -rt -rt H 3
E l-l -4 3 1
•rt CI 73 3 Ji
•a u B u cj
O eg eg eg -rt
0 i JB > o.
•>
ec
CJ
•rt
B "J
CJ Z
U •
•O
•C E
B eg
eg rt
rt OJ
a E
C -rt
•rt >
>
o
o
o
00
o
o'
m
o
o
o
o
o
r*»
O
O
o
o
o
o
cX
•f
CJ
•rt
[K
u
•rt
U
eg
[U
CO
rt
eg »
•rt tn B
W eg CJ
CJ CJ 4-J
4J ^4 4-1 CO
eg eg en X
B 3 CO
60 13
•o E a
01 i-l S 3
to rt 3 1
O T3 3 ^
rt E CJ O
o ta eg 1-1
e .= > p.
«
&
CJ
OCX
•5 H
U CO -
eg eg B
*J JZ eg
E U £
3 1- E
rt 3 o
O Bu 03
>
tn
E
o
-rt
•rt
ig
O
y
6D
B
10
i-l
X
01
o
E
eu
eg
•rt
0)
OJ
eg
CO
B
O
•rt
CO
CO
•rt
01
•e
c
eg
CO
o
l-l
B
o
CJ
01
E
•rt
111
CO
CO
f
^
O
CJ
eg
CJ
01
CJ
3
O
CO
eg
u
"B
eu
CJ
u
•rt
B
OJ
CO
to
CJ
^
p
0
to-
co
•
sO
en
in
„
m
*•
CM
en
Os
«
2
—
CM
*^
K
w^
CO
V
o
E
eu
CJ
irt
CI
a:
J3
213
-------
A summary of inorganic arsenic emissions and emission control devices
for each cf the eight arsenic chemical plants is presented in Table 9-5.
The majority of the particulate emission rates given in the table were
determined by source testing according to EPA reference methods. Annual and
hourly inorganic arsenic emissions from these facilities are small. The
total estimated inorganic arsenic emissions from the eight facilities
examined are 39.6 kg/yr (87.1 Ib/yr). One facility in Table 9-5, Osmose
Wood Preserving, has higher emissions than the other plants that were
studied. Controlled arsenic emissions are greater due to the comparatively
larger quantity of arsenic trioxide raw material that the plant handles. As
a consequence of handling more material, uncontrolled arsenic trioxide
emissions from Osmose would be greater than those from the other plants.
The measured arsenic emissions from the scrubber stack agree well with the
emissions estimate for the plant which can be calculated from the estimated
uncontrolled emission rate and the estimated control device efficiencies.
These comparative calculations are detailed in Appendix A-9.
It is also important to note that inorganic arsenic emissions from this
source category are not continuous. Emissions would be occurring only
during periods of arsenic trioxide handling, dumping, and processing. For
the arsenic chemical plants examined, the periods of arsenic trioxide
handling and dumping ranged from 800 to 6240 hours per year. The length of
the arsenic trioxide handling and dumping cycle is a function of the amount
of arsenic trioxide used and the capacity of plant equipment.
9.3 DEFINITION OF EEC OPTIONS
An examination of Table 9-5 indicates that emission capture and control
systems are extensively used in existing arsenic chemical plants, and they
are effective in reducing atmospheric inorganic arsenic emissions. The
magnitude of the arsenic emission rates in Table 9-5 signifies that EEC is
in place for this source category and is represented by the control systems
currently in use. The EEC system for this source category consists of a
particulate emission capture system in the form of hooding or enclosures,
with the collected material being pneumatically conveyed to a fabric filter
or wet scrubber control device. As emphasized in Section 9.2, the need for
EEC-type equipment has been dictated by OSHA workplace regulations for
airborne inorganic arsenic.
Because EEC is in place in the eight arsenic chemical plants, no model
plant development for the purposes of an economic analysis is required.
There would be no impacts of EEC because emissions under baseline control
are equivalent to emissions under EEC control.
9.4 POPULATION EXPOSURE DATA
The data in Table 9-6 were developed during this study for use in an
air pollution dispersion model to estimate the population exposure to
219
-------
"O
to
w:
_]
Z
0
M
M
*™
O
S
cn
O
X
w
z
o
M
H
55
0-
0
eu
Ed
X
H
pi
0
cn
H
Cd
5
OS
Pu
H
1-2
CU
•z.
M
1
Cd
SO
H
CO x->
C h
o j:
CO jO
CO — 1
IT
eo J=
41 C ~—
C O J3
•H -H iH
^H CO w
91 CO
CO «rt IJ
pa u ^
OO
CO
e
o
&.
c
o
-H
CO
i
a
e
o
H
-H
hi
CO
41
•a u <-i
•H to eg
V- C 0
•o u
HO 41
9 0 >
U
cu
o
N
e
o
T4
U
CO
U
o
c
to
r-l
0*
^^
•— CM
i§
§§
o o
y"»
•— •»
— • CN
§§
88
o o
^
h »
91 3
J3 re
u
ta
J*
^
o
eo
J
r*
^
CM
cn
in
"H
U "
O 3
Ui o
e x
CO eo
.c pa
U)
CO
•o e
C 91
§91
M
CO U
•H
O
^^ X^
CM -3-
— CN CN *»
§§§§
o o o o
0 0 0 O
o o o o
x-^ ^-s
CM sr
—t CM CM ^
o p o o
§000
o o o
o o o o
o o o o
%^ v*»
hi CO W CO
01 3 41 3
iO eo ^ Q
&£.£>*:
u o
jl J
O C7>
oo -a-
r- •»
ji -i
CO -9-
in —
CM O^
!n 5
i£ VO
Z
C9 •
a.< o
g CJ CO
O -H
CJ • eo
X 1*
0) 41 CO
p •-! cu
41 C >-4
a o a
o
^^
CO CO
§s
§8
o o
o o
^
co oo
o —
o o
o o
o o
^-^
l-i m
4) v_*
^H CO
tb CO
3
u eo
•H f
h X
U.
e
^
CO
cn
in
a
CM
CO
cn
p^
„
.
e-
t5
CO 4J CJ
41 B Z
CO 41
eu B -
K 0.-0
C hi
rt -H O
CO 91 U
u * §
X •*
^^
r*» in
Os fsl O O
— » -«• o o
o o o o
0 S ° S
^— .
r*» in
- §i
O CM O O
So o o
o o o o
»w s^
91
U
•H
91
hi CO CO
41 3 «-l 3
J3 CO o eo
u o
w u
1 J
ON 0
s0 CM
\o m
r* r*»
J J .
«* en
>O vO
00 C^
co cn
cn cn
in sr
„
o z o
CJ P C
•0 MX
O 00 - H
o e co •
3 .H -< u '
> j: ^ B
41 w & eg eo
co 41 B 3 X
O to 41 C U
g4) 2C B Pa
1- 41
Oh ft.
O" O%
000000
000000
0 S
<• ej>
000000
000000
0 S
U CO
41 3
js eo
14 M
CJ
C/1
e
I/I
MM
es
in
1
CM
0
CM
5
CO
f-l
eO
U
•H "1
e z
CJ -0
c
t> eo
§ «
— i e
41 i-l
C >
•H
>
^—^
r^ in
§0
O O
S-*.
r- m
§0
§§
o o
o o
*•**
41
|H
b. co
3
u eo
•H Si
A U
CO
u.
J
®1
«•*
gj
*"*
J
r*
00
i—
«^
f— <
00
c
"to
CO
f.
u
3 X
^
eo co
c a. c
3 3 O
•H o ta
0 U
> u
220
-------
•o
0)
s
•H
[ t
O
O
vO
CTN
W
iJ
P3
H
en
at
til
u
1
<
0.
g
(H
C/l
x
U x-^
U C.
O u-
41
> u
41
0 -2
M S
Ifc
U U
co eg
Oi "^
go
41
E^
8
41
U
3
u
4) h
!• ^
H ^
0
T-i
X
W
h
u IT
1 ii
•* s
r- '"'
00 m
•H *^
41
ss a
c
o
CO
U
o
a
P»
^
^-\ in ^ *~*. x^
CA CM -if M O
ON o m ao CTV
t^ ^- s^ CM CM
VO
o% o so in co
•"« O M GO 00
O O O O O
O in r* co o
r* —t
CM «tf fl *5f
in o co «r M
<*1 O O O O
c c e c
U Q) W fl)
1 111
Si —.
• XN •
— * CM CM
a ja
00 ^^ ^
o o o
m o
-^ o* o" M \o
•^r o <*^ ^^ **i
^^ ~«^ **^ ^*
C7*
in m ~^ c*^ oo
CM o ON o O
M n -—
C
0
•H
C CO
61 0
^ Z J= 0.
U > >s M U M
O 3 C h O
u o a * o tj
ax ou< o 41
co q S u to co w
am o 1-1 41 e
en o * co BS 4i
<0 X kl S
•a c co 41 a MO.
e v u M a a o
O 41 41 B M MM
§ M a o co cuo
co y a* cj > fl >
•HO -rf 4)
a MI so
•j
^
^C
•*
r*.
O
r*
O
U
en
^
^-^
ON
<*N
o
o
z
•a
M
0
a
a
CM
p.
^
m.
~&
§
m
M4
B
01
|
^^
CM
C
m
o
in^
^^
'""'
in
^
>,
C
<0
a.
B
O
U
60
e
>
41
<0
V
a. z
^H
o -
O CO
3 M
a) a
co e
0 41
3 x
tn
o
u
oj in
in si
os o
^i >rf
^.
CM
". "*.
ca o
p m
O oo
O
•3-
M O
c e
4) 4)
1 1
*T in
r- CM
C C
4) CU
c*l oo
m
t" U -0
" C
u - TJ CO
M C CM
» X M C
C U 41 -H
B si e >
41 M
a> >
CU
in
ON
CJ
°1
oo
S
a
c
1
^
•
C
«J
o
CM
fO
^^
41
r~-
a\
a.
3
O
O
00
e
•H
CO
CO
,£
U
M
3 X
a. t-
X •
u B
CO co
e c
MS
5
^
CN
u
y-i
m
m
o
CVJ
£
^^
in
o
o
CO
CJ
CO
u
CO
o
CO
2
CO
aj
-a
•rt
tn
c
M
en
^^
0 CN
e>J »
e -
en
00
O c^
• CM
O
M
CO <•
U CM
CO M
JJ
CO •
0
CO
41 CO
M
-------
arsenic from arsenic chemical manufacturing. The Strategies and Air
Standards Division (SASD) of EPA is using these data in an ongoing effort to
estimate exposure to arsenic from this source category.
222
-------
9.5 REFERENCES
1. Suta, B. E. (SRI International.) Human Exposure to Atmospheric
Arsenic. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, North Carolina. EPA Contract Nos. 68-01-4314
and 68-02-2835. May 1980. p. 78.
2. Cruse, P. (Acurex Corporation.) Arsenic: An Environmental Materials
Balance - Draft Final Report. (Prepared for U. S. Environmental
Protection Agency.) Washington, D. C. EPA Contract 68-01-6017.
March 1981. pp. 3-2 to 3-4.
3. PEDCo Environmental, Inc. Extended Source Survey Report for Arsenic.
(Prepared for U. S. Environmental Protection Agency.) Research Triangle
Park, North Carolina. EPA Contract No. 68-02-3173. May 1982.
pp. 3-26 to 3-27.
4. PEDCo Environmental, Inc. Arsenic Control Costs for Industrial
Sources. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, North Carolina. EPA Contract No. 68-02-2842.
June 30, 1978. Section 7.2.7.
5. Burrus, Jr., R., and D. Sargent. (Versar, Inc.) Technical and
Microeconomic Analysis of Arsenic and Its Compounds. (Prepared for
U. S. Environmental Protection Agency.) Washington, D. C. EPA
560/6-76-016. April 12, 1976. pp. 57-68.
6. Miller, C. S., and E. M. Bailey. Arsenic Acid Use and Hazard
Assessment in the Desiccation of Cotton. The Texas Agricultural
Experiment Station Report MP-134. October 1979.
7. Abernathy, J. R., Role of Arsenical Chemicals in Agriculture. (Paper
Presented at Arsenic Symposium Sponsored by Chemical Manufacturer's
Association and the National Bureau of Standards. Gaithersburg,
Maryland. November 4-6, 1981.)
8. The Biologic and Economic Assessment of Pentachlorophenol, Inorganic
Arsenicals, and Creosote - Volume II: Non-Wood-Preservatives. U. S.
Department of Agriculture. Technical Bulletin Number 1658-11.
November 1980. pp. 24-25.
9. Telecon. Brooks, G. W., Radian Corporation, with Raven, A., Colorado
Organic Chemical. January 18, 1982. Conversation concerning emission
of arsenic dust.
10. Telecon. Brooks, G. W. Radian Corporation, with Pazianos, G.,
Pazianos Associates. January 6, 1982. Conversation concerning arsenic
chemical manufacturers.
223
-------
11. Telecon. Brooks, G. W., Radian Corporation, with Brooks, W., Pennwalt,
Inc. January 7, 1982. Conversation concerning arsenic acid
production.
12. Letter from Price, R., Diamond Shamrock to Brooks, G. W., Radian
Corporation. February 25, 1982. 2 p. Arsenic acid manufacture.
13. Telecon. Brooks, G. W., Radian Corporation, with Smith, D., Voluntary
Purchasing Group. January 22, 1982. Conversation concerning arsenic
acid production.
14. Trip Report. Mappes, T., PEDCo Environmental, Inc. to file. March 30,
1981. Report of trip to Osmose Wood Preserving, Memphis, Tennessee.
March 30, 1981.
15. Telecon. Brooks, G. W., Radian Corporation, with Scharff, R., Mineral
Research and Development Corporation. January 12, 1982. Conversation
concerning wood preservative production.
16. Reference 2, pp. 3-17, 3-19, and B-2 to B-4.
17. Ansul Company. Report on Air Emissions from Organic Arsenical
Herbicide Manufacturing. June 1975.
18. Telecon. Brooks, G. W., Radian Corporation, with Nelson, R.,
W. A. Cleary, Inc. January 22, 1982. Conversation concerning
DSMA/MSMA production.
19. Letter and attachments from Alden, J., Woolfolk Chemical Works to
Brooks, G. W., Radian Corporation. January 5, 1982. 9 p. Arsenic
chemical manufacture.
20. Telecon. Brooks, G. W., Radian Corporation, with Alden, J., Woolfolk
Chemical Works. February 18, 1982. Conversation concerning arsenic
chemical production.
21. Telecon. Brooks, G. W., Radian Corporation, with Alden, J., Woolfolk
Chemical Works. January 5, 1982. Conversation concerning calcium
arsenate production.
22. Telecon. Brooks, G. W., Radian Corporation, with Alden, J., Woolfolk
Chemical Works. February 2, 1981. Conversation concerning calcium
arsenate production.
23. Telecon. Brooks, G. W., Radian Corporation, with Raven. A., Colorado
Organic Chemical. February 11, 1982. Conversation concerning sodium
arsenite production.
224
-------
24. Telecon. Brooks, G. W., Radian Corporation, with Livingston, M., Blue
Spruce Company. January 7, 1982. Conversation concerning sodium
arsenite production.
25. Telecon. Brooks, G. W., Radian Corporation, with Crenzie, W., United
Mineral and Chemical. January 11, 1982. Conversation concerning
sodium arsenite production.
26. Telecon. Brooks, G. W., Radian Corporation, with Black, J., Sharpe
Chemicals. January 18, 1982. Conversation concerning sodium arsenite
production.
27. Telecon. Brooks, G. W., Radian Corporation, with Duffy, D.,
C. P. Chemicals. January 12, 1982. Conversation concerning wood
preservative production.
28. Telecon. Brooks, G. W., Radian Corporation, with Renfroe, R., Chemical
Specialties, Inc. January 15, 1982. Conversation concerning wood
preservative production.
29. Letter from Baldwin, W., Koppers Company to Brooks, G. W., Radian
Corporation. April 22, 1982. 1 p. Arsenic chemical production.
30. Telecon. Brooks, G. W., Radian Corporation, with Morgan, J.,
J. H. Baxter. January 18, 1982. Conversation concerning wood
preservative production.
31. Letter and attachments from Allen, T., North Carolina Department of
Natural Resources & Community Development to Brooks, G. W., Radian
Corporation. January 20, 1980. 17 p. Arsenic chemical production.
32. Letter and attachments from Tivald, J., New Jersey Department of
Environmental Protection to Brooks, G. W., Radian Corporation.
January 25, 1982. 22 p. Arsenic chemical production.
33. Permit Applications and attachments from Clark, R. W., Osmose Wood
Preserving, Inc. to Memphis and Shelby County Health Department,
Tennessee. December 21, 1978.
34. Telecon. Brooks, G. W., Radian Corporation, with Beck, C., Los Angeles
Chemical Company. January 6, 1982. Conversation concerning arsenic
chemical production.
35. U. S. Environmental Protection Agency. Arsenic Source Survey Category.
No. 81/01. Confidential File of Emission Standards and Engineering
Division, Office of Air Quality Planning and Standards.
36. Memo from Brooks, G. W., Radian Corporation to file. March 15, 1982.
1 p. Arsenic Emissions from Voluntary Purchasing Group Plant.
225
-------
APPENDIX A - BASES FOR THE CALCULATION
OF ARSENIC EMISSIONS ESTIMATES
The purpose of this appendix is to present the methodologies,
assumptions, and supporting data used to calculate arsenic emissions
estimates for the source categories. The appendix has been organized to
correspond with the arrangement and order of the individual source category
chapters in the report. For example, information on emissions estimates
from copper smelting, which is Chapter 3 in the report, would be found in
Appendix A-3.
A-l
-------
APPENDIX A-3
1. Estimation of Primary Copper Smelting Process Arsenic Emissions
Under Existing, Baseline, and EEC Levels of Control
A. Existing Emissions
Estimates of process arsenic emissions from the primary
copper smelters in their existing condition were taken
directly from the arsenic material balances presented in
reference 1-.—
B. Baseline Emissions
In those cases for which the baseline configuration of
the smelter varies from the existing configuration, original
estimates of the baseline arsenic emissions from the smelters
were made. The estimates were based on the arsenic
distributions presented in reference 1. For example,
consider the Kennecott/Hurley smelter. At present the plant
has the configuration:
Reverberatory Furnace/ESP •*• Converter/Ac id Plant.
As a result of the regulatory baseline Kennecott/Hurley will
be replacing its reverberatory furnace with a flash furnace,
which will allow the plant to install additional acid plant
capacity to control SCL emissions from the smelting furnace
step. The baseline configuration of the Kennecott/Hurley
facility will then be identical to the existing configuration
of the Phelps Dodge/Hidalgo facility (Flash Furnace ->•
Converter/Acid Plant), for which arsenic emissions estimates
are available. The assumption is made that for identical
process and pollution control configurations, process arsenic
emissions would be proportional to the arsenic content of the
smelter feed. Therefore,
-------
estimated process arsenic
emissions from Kennecott/
Hurley under the regula-
tory baseline
x
estimated process arsenic
emissions from Phelps Dodge/
Hidalgo under existing
conditions
arsenic content of feed
materials to Kennecott/Hurley
per unit time
arsenic content of feed
materials to Phelps Dodge/
Hidalgo per unit time
The estimated process arsenic emissions from the Phelps
Dodge/Hidalgo smelter are taken from the arsenic
distributions presented in reference 1, and the arsenic
content of the feed materials to the Phelps Dodge/Hidalgo and
Kennecott/Hurley smelters are taken from Table 3-2 of
Chapter 3:
estimated process arsenic
emissions from Kennecott/
Hurley under the regula-
tory baseline
= 0.3 kg_
hr
0.2 kg/hr
20 kg/hr
0.003 kg/hr = 0.007 Ib/hr
C. EEC Emissions
In all but three cases, one of the approaches in
(l.A) or (l.B) was used to generate EEC arsenic emissions
estimates. A different approach was needed for estimating
EEC emissions from Kennecott/Hayden, Magma/San Manuel, and
White Pines/Michigan. For these smelters, neither the
existing nor baseline process emission control systems are
equivalent to EEC for arsenic. New arsenic emissions
estimates were made to assess the improved arsenic removal
capabilities of the existing particulate removal equipment
after the addition of spray chamber gas cooling systems.
These estimates were made by modifying the arsenic
distributions in reference 1 using the following assumptions:
A-3
-------
(1) Uncontrolled arsenic emissions from a given process
source follow the same pattern before and after
implementation of the new control equipment. For
example, the fractional volatilization and slagging in
the converter expressed as a fraction of the arsenic
input to the converter is assumed to remain fixed.
(2) Captured flue dusts that are presently recycled to the
smelting process can continue to be recycled to the
process regardless of the additional magnitude.
For example, consider the Kennecott/Hayden facility. The
arsenic distribution estimates presented in reference 1 and
shown in Figure 1 indicate that:
(1) 89 percent of the arsenic in the feed to the roaster is
volatilized, and 11 percent remains in the calcined
product. Dust captured by the cyclone in the roaster/AP
train is recycled to the reverberatory furnace feed.
The cyclone removes 83 percent of the arsenic in its
feed. Sludge produced by the scrubber in the roaster/AP
train is recycled to the roaster feed. The scrubber
sludge contains 98 percent of the arsenic in the
scrubber feed.
(2) 37 percent of the arsenic in the feed to the
reverberatory furnace is volatilized, 15 percent is
slagged, and 49 percent remains in the matte. Dust
captured by the hot ESP is recycled to the reverberatory
furnace feed. The ESP has an estimated arsenic
collection efficiency of n = 0.4 at its current
temperature of operation (650°F).
(3) 70 percent o£ the arsenic in the feed to the converter
is volatilized, 29 percent is slagged, and 1 percent
remains in the blister copper product. Dust captured by
the ESP in the converter/AP train is recycled to the
reverberatory furnace feed. The ESP removes 40 percent
of the arsenic in its feed. Sludge produced by the
scrubber in the converter/AP train is recycled to the
roaster feed. The scrubber sludge contains 90 percent
of the arsenic in the scrubber feed.
The basis of the estimate approach is to develop a material
balance for the smelter using the above information.
A-4
-------
OT
d
i
3
*-*
•r
a
i
••
**»
*"
(S.
3-
an c
— *
IM «4
3 •• j*i in-
12 -*
3*
te
^^
0»
9
0
•^
3
I
O
9
I
•rt
1 - n) F2
Scrubber Sludg«
r*
<*t
.. «
U M
ia ^"
o
OT
0
•*
O*
1
(•
bt
r»
O
&
|
Ml
3
i
•4
a
j
X
h'
—„ 1_— ___ *^
i.a -
j, "^ C
euperAtur*
epmulant)
Duat
i*
•
•8
» 1? ^
c
- .r -
a «
* : *
n
d
i i
Cft
>•
91
a
c
^*
en
*
u I
) « , =
) Hi
4)
•2 ~
1 S
M
8 'C-
:«
a ujs
* •.""
"*• o
ft0 -.
41
ac
;
«N
h.
C*lcln« 1
w
«
M
j
;
«r
k°
I
»*r
41
So
3d
Mfc.
•M 9%
^. S°' 5-
II ^
Is §
S-
O)
0.
a.
o
o
c
OJ
T3
^
X
3 TJ
•s
— j o
QJ
0)
? ^
:<-"• ^
^\
_ «2
St «
-° 3
Si c
S o 0
4->
rs
ja
•^
-M
crt
•r~
T3
o
'£
" 0)
2
— <
i— i
ai
s_
01
u.
A-5
-------
Let: F^ = amount of arsenic in the smelter feed (kg/hr)
FI = amount of arsenic in the roaster feed (kg/hr)
F_ = amount of arsenic in the reverberatory furnace
feed (kg/hr)
F- = amount of arsenic in the converter feed (kg/hr)
n = arsenic removal efficiency of the reverberatory
furnace ESP
E = arsenic emissions from the reverberatory furnace
ESP (kg/hr)
All of the offgas, dust, and scrubber sludge streams can be
expressed as shown in Figure 1 in terms of feed rates F-, FI,
F_, F_, and efficiency n. There are a total of four unknowns
(F, , F_, F_, and E) and two known parameters (F,., n). Four
equations are needed to express the unknowns in terms of the
parameters F« and n. These are provided by writing a
material balance equation for the feed to each of the
smelting processes and by writing a relation between E and
V
Roaster feed:
FL = FQ + .89(1-83)(.98)F1 + .7(1-.40)(.9)F3
or
0.85FJ-0.38F2 = FQ
Reverberatory Furnace Feed:
F2 = .11F1 + .89(.83)F1 + .37nF2 + .7(.4)F3
or
0.85F1 - (l-.37n)F + 0.28F3 = 0
Converter Feed:
F3 = .49F2
or
0.49F2 - F3 = 0
Reverberatory Furnace Emissions:
E = 0.37(l-n)F0
or
0.37(l-n)F2 - E = 0
A-6
-------
The system of equations to be solved is:
(i) 0.85F - 0.38F - F
(ii) 0.85F; - (l-.37n)F7U+ 0.28F
(iii) 0.49F* - F - 0 Z
(iv) 0.37(f-n)Fo - E = 0
3
Simultaneous solution of the four equations for the
reverberatory furnace arsenic emissions E yields:
(1 ' "> V(1'86 ~ *
\ ESP /
Similarly, it can be shown:
(1 - .426n) FQ/(.672 - .364n)
FQ/(.679 - .383n)
FQ/(1.39 - .752n)
The feed to the acid plant circuit can also be expressed in
F- .
.89 (1-.83) (1-.98) F + .7 (1-.4) (1-.9) F
terms of F. and F- .
\ plant/
or
/F . . \- 0.003 F. + 0.042 F0
/ acid I 1 J
\ planty
Using these equations, the arsenic distribution at
Kennecott/Hayden can be computed provided Fn and n are known.
From Table 3-2 of Chapter 3, FQ = 8,5 kg/hr = 18.7 Ib/hr.
The reverb ESP collection efficiency n depends on the
temperature at which the device is operated. The available
data suggest that a spray chamber/ESP system should be
capable of achieving 96 percent arsenic collection
efficiency. Using n « .96 and 7 = 8.5 kg/hr =-18.7 Ib/hr,
it follows that
^reverb) = °'38 kg/hr * °-84 lb/hr
ESP J EEC
A-7
-------
Under existing and baseline control n = 0.40, so that
:'reverb!
ESP / baseline
=3.5 kg/hr = 7.7 Ib/hr
Thus, the algebraic model predicts an achievable emissions
reduction of 3.5 - .38 = 3.1 kg/hr =6.8 Ib/hr from the
reverberatory furnace ESP. The amount of arsenic eliminated
as slag increases, as does the amount of arsenic entering the
acid plant unit.
2. Estimation of Primary Copper Smelting Fugitive Arsenic Emissions
Under Baseline and Existing Levels of Control
A. Baseline Emissions
Estimates of baseline fugitive arsenic emissions from
the primary copper smelters were taken directly from
reference 1, except for the Cities Services plant in
Copperhill, Tennessee for which fugitive arsenic emissions
estimates were not given. For this plant, the approximate
fugitive arsenic emissions distribution shown in Table 3-3 of
Chapter 3 is used to generate baseline (uncontrolled) arsenic
emissions estimates. From Table 3-2 of Chapter 3, the
arsenic content of the feed material at the Copperhill plant
is found to be 0.05 kg/hr (0.1 Ib). The fugitive arsenic
hr
emissions estimate from an arbitrary source S within the
plant is generated as follows:
fugitive arsenic
emissions from
source S
amount of arsenic
in the feed material x
to the plant per unit
as shown in Table 3-2
fraction of input arsenic
released as fugitive emissions
from source S as shown in
Table 3-3 of Chapter 3
For example, Table 3-3 of Chapter 3 indicates that
1.8 percent of the arsenic entering a typical copper smelter
is released as fugitive emissions from the converter. Thus,
the uncontrolled converter fugitive arsenic emissions at the
Copperhill Plant are estimated as follows:
A-8
-------
["uncontrolled fugitive arsenic
emissions from the converter
[at Cities Services/Copperhill.
0.05 kg/hr x 0.018
9 x 10~4 kg/hr - 2 x 10"3 Ib/hr
Similarly, it follows:
*
Uncontrolled fugitive
arsenic emissions from
the matte and slag
tapping operations at
Cities Services/Copperhill
B. EEC Emissions
0.05 kg/hr x (.0034 + .0001)
- 1.8 x
kg/hr
3.9 x 10 Ib/hr
Estimates of EEC fugitive arsenic emissions from the
primary copper smelters were obtained by applying the
following assumed EEC capture and collection efficiencies to
the baseline (uncontrolled) fugitive emissions estimates
obtained above:
- for furnace tapping operations, 90 percent capture
efficiency is assumed for the EBC ventilation system and
96 percent arsenic collection efficiency is assumed for
the EBC fabric filter collection system. The overall
control efficiency is:
n
overall
0.90 x 0.96 = 0.864
- for converter operations, 95 percent capture efficiency
is assumed for the EBC air curtain/secondary hood
ventilation system and 96 percent collection efficiency
is assumed for the EBC fabric filter collection
efficiency. The overall control efficiency is:
overall
0.95 x 0.96 = .912
The following equation was used to estimate EBC fugitive
arsenic emissions for the two largest sources of fugitive
arsenic emissions in the primary copper smelting industry:
EBC fugitive
arsenic emissions
.estimate for source S
"uncontrolled fugitive
arsenic emissions estimate
for source S
1 - n
overall
A-9
-------
For example, EEC fugitive arsenic emissions from the
converter operation at Cities Services/Copperhill are
estimated as follows:
EEC Emissions =• 9.0 x 10~A kg/hr x (1 - .912)
= 8 x 10 kg/hr - 1.7 x 10 Ib/hr
A-10
-------
REFERENCES FOR APPENDIX A-3
1. U. S. Environmental Protection Agency. Preliminary Draft Report -
National Emission Standard for Arsenic Emissions from Primary Copper
Smelters. Research Triangle Park, N. C. EPA Contract No. 68-02-3060,
1979.
A-ll
-------
APPENDIX A-4
1. Primary Lead Smelting Process Arsenic Emissions under Existing Level of
Control
The methods used to obtain estimates of the process arsenic emissions
for each of the five primary lead smelters are outlined below.
A. ASARCO/East Helena
Process arsenic emissions from this smelter were taken from
the test results shown in Table 4-3 of Chapter 4. It was assumed
that the offgas streams from Baghouses 1, 2, and 3 are vented to
the blast furnace/reverberatory furnace stack.
B. ASARCO/E1 Paso
(i) Blast furnace/Reverberatory furnace
Arsenic emission data from the blast
ace/reverberatory furnac
(0.16 Ib/hr, or 0.07 kg/hr).
(ii) Zinc fuming furnace
Arsenic emissions from the zinc fuming furnace were
estimated by analogy to ASARCO/East Helena. The assumption
is made that for comparable emission control systems, arsenic
emissions from a particular plant component are proportional
to the arsenic content of the feed material to the plant.
E =» E
zinc fuming zinc fuming
furnace ,„, N furnace /„..„, \
(El Paso) (East Helena)
weight percent smelter
arsenic in feed ,_, . capacity . .
(El Paso) * J (El Paso)
2
furnace/reverberatory furnace stack were provided by ASARCO
weight percent smelter
arsenic in feed ,„ . „ - . capacity /•„..„, x
(East Helena) v J (East Helena)
A-12
-------
Since the two plants process western lead ore and since
the capacities of the plants are identical (82,000 Mg/yr) it
follows that the estimated arsenic emissions at ASARCO/El
Paso are identical to the measured arsenic emissions at
ASARCO/East Helena. Therefore
E , - . E . , . 0.2 kg/hr = 0.4 Ib
zinc fuming = zinc fuming = r—•
furnace furnace
(ASARCO/El Paso) (ASARCO/East Helena)
(iii) Sinter/acid plant stack
ASARCO/El Paso sends its entire sinter machine offgas
stream to an acid plant unit (100 percent gas recirculation).
Arsenic emissions from the sinter/acid plant stack are
assumed to be negligible in comparison to emissions from the
other plant components.
C. AMAX/Boss
Emissions test data are not available for process arsenic
emissions from this smelter... However, material balance
information provided by AMAX indicates that 12.9 percent of the
arsenic entering the plant cannot be accounted for in the
materials leaving the smelter. This amounts to 0.019 tons of
arsenic per blast furnace day, or about 0.72 kg/hr (1.6 Ib/hr).
Thus, the total arsenic emissions from the plant (process +
fugitive) are bounded from above by 0.72 kg/hr (1.6 Ib/hr).
The "best estimate" of process arsenic emissions from the
plant was obtained by analogy to ASARCO/East Helena. Emissions
from the sinter machine, blast furnace, and reverberatory furnace
were assumed proportional to the arsenic content of the feed at
the two plants. The weight percent of the ore concentrate at AMAX
is 0.02 percent and at ASARCO/East Helena, about 0.4 percent. The
production capacities of the plants are 127,000 and 82,000 Mg/yr
lead, respectively. Thus,
F » / K + F + F" 1
"main I sinter blast reverb/East Helena
x 127,000
82,000
= 0.2 kg/hr - 0.4 Ib
hr
A-13
-------
St. Joe/Herculaneum
No process arsenic emissions data are available for
St. Joe/Herculaneum. Estimates for total maximum arsenic
emissions (process + fugitive) were obtained by analogy to
AMAX/Boss. Estimates for process arsenic emissions were made by
analogy to ASARCO/East Helena. The arsenic content of the ore
concentrate smelted at St. Joe/Herculaneum is 0.02 weight percent.
The production capacity of the St. Joe smelter is 204,000 Mg/yr
lead.
E = E
process process
+ fugitive (St. Joe) + fugitive (AMAX)
"main
stack
x .02 x 204,000
.02 127,000
1.12 kg/hr = 2.5 Ib
hr
(St. Joe)
E . „ + E. . + E J
sinter olast reverb)
East
Helena
x .02 x 204,000
.4 82,000
- 0.3 kg/hr - 0.7 Ib
hr
ASARCO/Glover
No process arsenic emissions data are available for
ASARCO/Glover except for the venturi stack (0.0005 kg/hr, or
0.001 Ibs/hr). Estimates for total maximum arsenic emissions
(process + fugitive) were obtained by analogy to AMAX/Boss.
Estimates for process arsenic emissions were made by analogy to
ASARCO/East Helena. The arsenic content of the ore smelted at the
ASARCO/Glover smelter is 0.025 weight percent and the production
capacity of the plant is 100,000 Mg/yr lead.
"process ASARCO/
- . . Glover
+ fugitive
E /AVTAVN x -025 x 100,000
process (AMAX) -^- 127>000
+ fugitive
0.55 kg/hr
1.2 Ib
hr
A-14
-------
main
stack ASARCO/
Glover
'sinter
^1,
ast
""reverb)
East
Helena
x 0.025 x 100,000
0.4 82,000
- 0.2 kg/hr - 0.4 Ib
hr
2. Primary Lead Smelting Fugitive Arsenic Emissions Under Existing Level
of Control
Fugitive arsenic emissions data are available for ASARCO/East Helena
and ASARCO/Glover (Tables 4-4 and 4-5 of Chapter 4). Estimates of fugitive
arsenic emissions at the other primary lead smelters were obtained by
analogy to these two smelters. As in the case of process arsenic emissions,
the fugitive arsenic emissions at a smelter were assumed to be proportional
to the arsenic content of the feed material.
For example:
JfUgitive /AMAX/\
I Bossy
I Glover I
wt % of arsenic in
feed to AMAX/Boss
wt % of arsenic
in feed to ASARCO/Glover
production capacity
at AMAX/Boss
production capacity
at ASARCO/Glover
0.005 kg_ x 0.02 x 127,000
hr 0.025 100,000
0.005
hr
0.011 Ib
hr
A-15
-------
REFERENCES FOR APPENDIX A-4
1. U. S. Environmental Protection Agency. Emissions Measurement at the
ASARCO Lead Smelter in East Helena, Montana. Research Triangle Park,
N. C. EPA Contract No. 68-01-4140. May 1980.
2. Letter from Richardson, J. B., ASARCO, to Fensterheim, R. J., Chemical
Manufacturers Association. May 17, 1982. ASARCO response to Radian
Emission Estimates.
3. Letter from Kearney, W. M., AMAX Lead Company of Missouri, to Lee Beck,
EPA:ISB. August 6, 1981. Arsenic material balance at the AMAX
smelter.
4. U. S. Environmental Protection Agency. Sample Fugitive Lead Emissions
from Two Primary Lead Smelters. Research Triangle Park, N. C.
Publication No. EPA-450/3-77-031. October 1977.
A-16
-------
APPENDIX A-5
1. Estimation of Primary Zinc Smelting Arsenic Emissions Under Baseline
and EBC Levels of Control
A. Electrolytic Zinc Plants
The only source of significant arsenic emissions at an
electrolytic zinc plant is the roaster acid plant tail gas.
Fugitive arsenic emissions are assumed to be negligible from the
fluid bed roasting systems. Two methods were used to estimate the
acid plant tail gas arsenic emissions.
(i) The first method assumes that the tail gas leaves the acid
plant stack with an arsenic content corresponding to
saturated As.O, vapor at 60°C (140°F). This is the typical
temperature at which the offgas leaves the preliminary wet
scrubber in the acid plant circuit. The assumption is made
that the condensed arsenic is removed by the ESP mist
precipitator that generally follows the scrubber. Regression
of the A.s.0, vapor pressure data presented in reference 1
over the temperature range gO < T < 150°C (194 < T < 302°F)
yields the relation log%(, P = 12.77 - 6067 (1/T), where T =
temperature in °K and P = vapor pressure of As,0- in mm Hg.
Extrapolating the vapor pressure relation back to 333°K
(60°C) yields P » 3.57 x 10 mm Hg @ 60°C. Assuming ideal
gas behavior, this-corresponds to an arsenic concentration of
5.7 x 10 g As/Nm . The arsenic emissions estimates for the
roaster acid plant tail gas streams are obtained as follows:
Acid Plant
[Emissions J
Volumetric Flow
Rate of the Tail
Gas Stream at
Standard Conditions
x 5.7 x 10~5 g As/Nm3
For example, the volumetric flowrate for the acid plant
offgas stream at ASARCO/Corpus Christi is 6.9 Nm /sec. Thus,
the arsenic emissions estimate is:
6.9 Nm3/sec x 5.7 x 10~5 g As/Nm3 x 3600 sec/hr
- 1.4 x 10~3 kg/hr - 3.1 x 10"3 Ib/hr
A-17
-------
(ii) The second method assumes that the tail gas leaves the acid
plant stack with the maximum allowable arsenic content for
the inlet gas to the acid plant catalyst (1.2 mg/Nm ). Thus,
the emissions estimate is:
6.9 Nm3/sec x 1.2 x 10~6 kg/Nm3 x 3600 sec/hr
= 3.0 x 10~2 kg/hr - 6.6 x 10"2 Ib/hr
Since baseline and EEC control are equivalent for the
electrolytic zinc plants, the EEC arsenic emissions estimates
are identical to the baseline arsenic emissions estimates.
In the text, the estimates made by method (i) are called the
"best estimates" and the estimates made by method (ii) are
called the "high estimates."
B. St. Joe/Monaca Electrothermal Smelter
The arsenic emissions-estimates for this plant were obtained
by direct correspondence ' with the plant management.
C. New Jersey Zinc/Palmerton Zinc Oxide Plant
There is no available information concerning the distribution
of arsenic at the New Jersey Zinc zinc oxide plant. Production
information obtained from the plant indicates that about 1.2 kg
As/hr is contained in the smelter feed. There are five baghouses
at the plant; two are "hot" baghouses (#1 @ 193°C and #2 @ 140°C)
and three are "cool" baghouses (#3 @ 91°C, #4 @ 82°C, and #5 @
82°C). Two slags are produced in the operation, one from the
waelz kiln and one from the horizontal moving grate furnace. The
following somewhat arbitrary assumptions are made to estimate the
arsenic distribution at the plant (see Figure 1):
(i) Assume that 20 percent of the arsenic entering the waelz kiln
is slagged and 80 percent is volatilized with the waelz
oxide. Assume that arsenic emissions associated with the
fugitive dust baghouse are negligible compared to those from
the other sources. Assume that the 140°C baghouse serving
the waelz kiln process offgas has an arsenic capture
efficiency of 40 percent (by analogy to medium-temperature
copper smelter particulate controls).
(ii) Assume that 25 percent of the arsenic in the sinter machine
feed is volatilized (by analogy to a "typical" copper smelter
roaster) and that the cool sinter machine baghouse has an
arsenic capture efficiency of 96 percent.
A-18
-------
M • mm
S° 0 -
c -0 K y> > •
~ s. K is
!*• •'o «S
.1*. 4« ^2°
II; k -II
-I « J_ « * a
> « o
ii: ' •-
S" e
0 « K
M O
sS >
0
«
M
S ~ ~ *
or- k. e
!£» 0 £
• S * **
S'"' m M
M «•« e
•&J3 * "
• o * • » *
" " -—
3~!i e"1 I
o e IM • 3
w ^ •» • «
£ 2 3 "
0
• c* -S
1 c 5 £
J ". J! 5
?a ^^^ ™ • J!
1 c"1 3*
•4
M ^
H
1.
IK
5*2
M «
•4 k*
1°
fa. h
5 1
M M
M ft
r* •
C
i
*
s
C
H
'1
" t
M
1
M •
• »
S 01
s « -^
21 1 2
5 *" *ij " Ti S
• *• i c * ~ °°
J! 15 if .= ,
5*S oo- JT e i. «"
AMQ IK «««£<
* S w «o o o» *-
t " u *. i °: £ o i o «
1 _. X O *«* O •-». •*». *^
5 tfk *•» «* M 1 «C ¥
£ ^ (C K M K J* .M
. U W mf ^1
,A ' eo x e • • • X^C5** "*T
— _ -*~.KKJ{ UJ^KJ'
H Jt
££ M o ^ «• (13
«r -• « c M - K ° *
*— — 5? ^* xmiD m« ~
'M r*4«0* CDINCD* O
c-<— IB • • o •••e ._.
5oC o'tt 00°K IS
S««« « " " « """^ 3
SSt" f «»»o »»»o -r-
"• e « c" o o o K e' o o ic S-
c • • ' a • • • s U
« 33 = o 33Co ^
e co i i i i iiit
!? «*" UMMW UWUU1 CJ
i 3s sicsic e c 2 £ S
**" o SSSESS assrss to
K R9BMm ffiflCMB ^_
irt ^
p*
O)
•a , iT3
M - o * u E
3 S "S ° -g -^
Z O « « •«> 4->
Bfc. s°:S .. „ E oo
. =!: 3 ° • ; c • uj
' ^S a • £ 3 • i
m "»~--«*r<^.
5 S^'^t15'^^ a;
uocr— us" i.
iii "S i i 3
JrfT- =c- .?
2 I
A-19
-------
(iii) Assume that 20 percent of the arsenic entering the
horizontal grate furnace is slagged and 80 percent
volatilizes with the product ZnO vapor. Assume that the
cool horizontal grate furnace baghouse has an arsenic
collection efficiency of 96 percent.
(iv) Assume 25 percent of the arsenic entering the fume kiln is
volatilized. Assume that the cool fume kiln baghouse has an
arsenic collection efficiency of 96 percent.
Using this set of assumptions, a material balance was
prepared to assess the "best estimate" of existing arsenic
emissions. They are shown on Figure 1. Then, assuming that added
spray chamber capacity is added to waelz kiln baghouse so that the
temperature is lowered to about 100°C and r\ » 0.96, a set of(EBC
emissions are estimated.
Existing As emissions: 0.6 kg/hr =» 1.3 Ib
hr
EEC As emissions: 0.07 kg/hr = 0.15 Ib
hr
A-20
-------
REFERENCES FOR APPENDIX A-5
1. U. S. Environmental Protection Agency. Arsenic Emissions from Primary
Copper Smelters - Background Information for Proposed Standards.
Preliminary Draft. Research Triangle Park, N. C. February 1981.
2. Telecon. Janeck, T., St. Joe Minerals Corporation, with Keller, L.,
Radian Corporation. February 11, 1982. Information concerning the
St. Joe Minerals zinc smelter in Monaca, Pennsylvania.
3. Letter from Janeck, T., St. Joe Minerals Corporation, to Keller, L.,
Radian Corporation. March 10, 1982. Arsenic Emissions from the
St. Joe/Monaca zinc smelter.
4. Telecon. Silvestris, M., New Jersey Zinc Co., with Keller, L., Radian
Corporation.
A-21
-------
APPENDIX A-6
1. Calculation of Arsenic Emissions from Secondary Lead Smelters
Approach
Particulate emissions and typical lead concentrations of particulates
are well documented. Arsenic and lead have been shown to behave similarly
in studies of trace element emissions from coal-fired power plants. In
addition, lead/arsenic ratios of dustfall around secondary lead smelters
have been measured. By using these ratios in conjunction with particulate
and lead emission rates one can estimate an arsenic emission rate.
1. Particulate Emissions from Secondary Lead Smelters.
All furnaces are equipped with fabric filters. Fabric filters are
capable of lowering emissions to below 50 mg/m (0.022 gr/dscf) NSPS limit.
Most test results are below this (see Table 1). For estimation assume
particulate emissions are .022 gr/dscf.
2. Lead Content of Particulate.
Measurements of the lead contents of particulate emissions from
secondary lead smelters show wide variations. In general, the lead contents
increase with decreasing particle size. Table 1 shows average lead contents
to vary from 1.5 to 47 percent. General Battery test data show a range of
20 - 52 percent with a weighted average of 42 percent. This lead content
will be used in this study.
3. Lead/Arsenic Ratios of Two Secondary Smelters.
From Reference 2, page 319, Table 5, we have the following average
arsenic and lead contents in soil and foliage (ppm by weight).
A-22
-------
TABLE 1. EMISSIONS TEST RESULTS FOR SECONDARY LEAD SMELTERS
PLANT
Sanders
Lead
Tonolli7
SOURCE
Blast
Furnace
Rotary
Furnace
(FF & Scrubber in
Series)
Q
General
Battery
A9
B
C
D
E
F
G
H
I
AVERAGE
AVERAGE
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Blast
Furnace
Reverb
Furnace
Reverb
Furnace
Reverb
Furnace
FABRIC FILTER
FF/ SCRUBBER
CONTROL
DEVICE
AB/Fabric
Filter
AB/Fabric
Filter
Scrubber
AB/FF/
Scrubber
AB/FF/
Scrubber
AB/Fabric
Filter
AB/FF/
Scrubber
Scrubber
AB/Fabric
Filter
AB/Fabric
Filter
AB/Fabric
Filter
Fabric
Filter
Fabric
Filter
Fabric
Filter
TEMPERATURE
°F
-
87°
-
-
96°
178°
121°
95°
152°
175°
110°
168°
124°
327°
176°
101°
PARTI CUL ATE
gr/dscf
.0002
.00616
.01264
.024
.020
.01
.0026
.0079
.0143
.0013
.0059
.0142
.0035
.0033
.0022
.0043
.010
LEAD Lead
gr/dscf cone, in
particulate
-
.00013 2%
.00024 1.9
.00018 0
.00007 0
.00014 - 0
.00025
.00034 13
.00012 1.5
.00105 7
.00061 47
-
-
.00061 17
.00038 12
734,100
.000414
.00017
A-23
-------
Fb As_ Ratio Pb/As
Smelter A
Foliage 173 7.4 23:1
Soil 3,106 107.2 28.9:1
Smelter B
Foliage 160 2.4 66:1
Soil 3,292 34.8 95:1
Urban Control Area
Foliage 91 0.9 101:1
Soil 542 9.8 55:1
A-Urban
Foliage 173 - 91 7.4 - .9 12.6:1
Soil 3,106 - 542 107.2 - 9.8 26.3:1
B-Urban
Foliage 160-91 2.4 - .9 99.3:1
Soil 3,292 - 542 36.8 - 9.8 110:1
The range of ratios seen is 12.6:1 to 110:1.
For the purpose of this study we will use 13:1 to maximize arsenic
concentrations. The single flue dust analysis identified during this study
shows an arsenic content of 0.3 percent arsenic in the material captured by
a fabric filter. Using a lead content of 42 percent and a 13:1 Pb/As ratio,
the arsenic in the material escaping is approximately 3.2 percent As. (This
would be .38 percent @ 110:1.) This represents an enrichment ratio of = 11
from the larger to the smaller particulate. A similar enrichment can be
seen between the .1 - 3 um particulate and the 27 - 76 pm particulate in
Table 2.
4. Emissions Estimates
a. Process Emissions _
Using the model plant flow rates proposed by MRI , arsenic
emissions from a typical furnace can now be estimated. For the model
plants we have the following:
Plant Size Small Medium
Flow rate dscfm 20,000 36,000
Hours of operation 6,000 6,000
Lead production tpy 10,000 20,000
Hourly 1.67
A-24
-------
Q
^3
^J
CJ
M
z
w
C/3
OS
^3
OS
o
w
o
z
2J
2 co
O E*"*
PD z
OS <£
W H-*
PJ PH
W O3
CJ W
M 3
> 0
W p i
Q
O
o os
OS M
H Pa
O ,-J
CJ <
O
Q CJ
< S
O
CO OS
o w
M
H CO
OS O
M
H CO
Z co
W M
«S IS
S w
CJ
M Q
Z W
W M
•
CM
W
M
P3
^j
H
CO
o
•H
4J
CO
os
,c
CO
cfl
£*t
rH
tu
O
.— 1
CO
o
M
H
«
OS
H
Z
pd
n ^
CJ CU
M S
OS CU
w w
o
•^ CM CM
O — ( «-<
Q
^^ in *^r
pO vO
•*•«. i— i «
r-~
1
t— 4
t
o
o
•H
CU t3
CO CO
V-i CU
< hJ
• •
C C v£>
i-l -H •
E \o co co @ CM (n m
••-. o ON CM "^. vo •— i r-«
60 in co 60 sr — i
e e
•T
CO
H
SB
cS
w
»j
w
w
CJ
cS
H
23
o
w
CJ
z
S
o^
p
OS
w
oS
w
n
PQ
D
os
o
Q
52
? M
H H
H O — i OH
CO £ OS C CO
O
^
O
c
cu
^ «_J
4J CU 0
0) t-l •!-(
i— I 4J 14-1
C 3
-------
Use the largest flow per ton of product (i.e., the small plain.
assumptions). Hourly arsenic emissions are:
20,000 dscfm x .022 grdscf x 60 min x .032 Ib As
7,000 gr/lb hr Ib particulate
hourly As = .12 Ib/hr
Emission factor for a small plant Ib/ton lead » .12 Ib/hr ==
.072 Ib As/ton lead 1.67 tons/hr
Using the ratios of flowrates we can estimate emissions from other
sites.
For a medium plant emissions = 36,000 x .12 Ib/hr = .21 Ib/hr
20,000
For a large plant emissions - 72,000 x .12 = .43 Ib/hr
20,000
b. Process Fugitive Emissions
Process fugitives include charge preparation, slag tapping and
mef:al pouring emissions. These are currently controlled at all
secondary lead smelters. There have been two studies of fugitive lead
emissions which measured process., fugitives. These are: The General
Battery Fugitive Emissions Study and the East Penn Emissions and
Emissions Control Study . The General Battery Study showed the
following:
Source Lead Emissions Ratio to Process
Process Controlled by 750 1:1
Fabric Filters
Ventilation (sanitary) 558 0.74:1
Fabric Filters
Area Fugitives 1,270 1.7:1
A-26
-------
The East Perm Study showed the following:
Lead Content
3 3
Source Flow Rate m /h Cone, mg/m wt gms/h Ratio to Process
Process 33,000 6 198 1:1
Controlled by
Fabric Filters
Slag Tap Fabric 8,400 1.4 11.76
Filter
Sanitary Fabric 59,000 1.5 88.5
Filter
Total Process
Fugitive 100.26 0.5:1
Choosing the worst case, assume controlled process fugitive emissions
are 74 percent of controlled process emissions.
The MRI report presents an estimate of uncontrolled process
fugitive emissions on page 14. They give 326 tpy of total particulate
for 10,000 tpy of lead. Using our assumptions for the arsenic content
of the particulate yields an uncontrolled arsenic emissions rate of
326 tpy Particulate x .032 ton As » .00104 t As/t lead
10,000 tpy lead ton particulate
Assuming the fabric filter is 99 percent efficient the equivalent
controlled rate would be .0000104 t/t. The estimate using the .74 of
process is .072 Ib/ton x 1 ton x .74 = .000027 t/t.
2,000 Ibs
These estimates are in fairly close agreement. Use the larger of the
two to be conservative.
c. Area Fugitives
Use the General Battery data. Area fugitives are 1.7 x controlled
process emissions. These emissions correspond to a typical existing
level of control. Emissions would be lowered when the SIPs are revised
to allow compliance with the lead NAAQS. Since ambient lead measure-
ments are not available for all smelter locations, no estimates can be
made of the emissions reduction would occur when compliance is
achieved.
A-27
-------
Area fugitives for a small plant:
Arsenic emissions
Ib/ton
1.7 x .072 = .12 Ib/ton
5. EEC Emissions Estimates for Process Sources
Table 1 presents lead and particulate emissions test data for blast and
reverberatory furnaces controlled by fabric filters and scrubbers. The
table shows that a fabric filter/scrubber combination operates at a lower
exit gas temperature than a fabric filter alone and has lower lead
emissions. The incremental reduction in lead emissions attributable to the
combination is 59 percent. Lead and arsenic particulate are controlled to
approximately the same efficiency by other control devices (Table 2).
Therefore, assume arsenic removal efficiency 3 lead removal efficiency.
Arsenic emission factors for EBC » .59 x existing.
A-28
-------
APPENDIX A-6
REFERENCES
1. Schwitzgebel, K., G. S. Gunn, and M. A. Capalongan. (Radian
Corporation.) Fugitive Emissions at the Seconadry Lead Smelter
Operated by General Battery Corporation in Reading, Pennsylvania.
(Prepared for U. S. Environmental Protection Agency, Region III.)
Philadelphia, Pennsylvania. EPA Contract No. 68-02-3513.
December 1981. p. 3.
2. Temple, P. G., S. N. Linzon, and B. L. Chai. Contamination of
Vegetation and Soil by Arsenic Emissions from Secondary Lead Smelters.
Environmental Pollution. 12(4):311-320. April 1, 1977.
3. Midwest Research Institute. Development of New Source Performance
Standards, Project Conclusions Report for Secondary Lead Smelting
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-3059.
January 27, 1981. pp. 8-11.
4. Schwitzgebel, K., and R. Vandervort. (Radian Corporation.) Emissions
and Emission Controls at a Secondary Lead Smelter. (Prepared for U. S.
Environmental Protection Agency and U. S. Department of Health and
Human Services.) Cincinnati, Ohio. EPA Contract No. 68-03-2807.
January 9, 1981. p. 5.
5. Midwest Research Institute. Development of New Source Performance
Standards, Project Recommendation Report for Secondary Lead Smelting
Industry. (Prepared for U. S. Environmental Protection Agency.)
Research Triangle Park, N. C. EPA Contract No. 68-02-3059.
June 12, 1980.
6. Trip report. Maxwell, C. M., Midwest Research Institute, to
Telander, J., EPArlSB. May 19, 1980. 8 p. Report of April 19, 1980
visit to Sanders Lead Company in Troy, Alabama.
7. Trip report. Maxwell, C. M., Midwest Research Institute, to
Telander, J., EPArlSB. May 6, 1980. 109 p. Report of April 22, 1980
visit to Tonolli Corporation in Nesquehoning, Pennsylvania.
8. Trip report. Medepalli, K. S., Midwest Research Institute, to
Telander, J., EPA:ISB. May 5, 1980. 13 p. Report of April 23, 1980
visit to General Battery Corporation in Reading, Pennsylvania.
9. Trip report. Thomas E. Mappes. PEDCo Environmental, to Beck, L.E.,
EPA:ISB. March 30, 1981. Report of March 18, 1981 visit to Refined
Metals Corporation in Memphis, Tennessee.
A-29
-------
10. U. S. Environmental Protection Agency. Background Information for
Proposed New Source Performance Standards: Asphalt Concrete Plants,
Petroleum Refineries, Storage Vessels, Secondary Lead Smelters and
Refineries, Brass or Bronze Ingot Production Plants, Iron and Steel
Plants, Sewage Treatment Plants; Volume 2: Appendix: Summaries of
Test Data. Research Triangle Park, N. C. Publication No.
EPA-APTC-1352b. June 1973. pp. 37-43.
11. Smith, R. D. The Trace Element Chemistry of Coal During Combustion and
the Emissions from Coal-Fired Plants. Progress in Energy and
Combustion Science. 6/1):53-119. 1980.
12. Kaakinen, J. W., et al. Trace Element Behavior in Coal-Fired Power
Plant. Environmental Science and Technology. £(9):862-869.
September 1975.
13. Klein, D. H., et al. Pathways of Thirty-seven Trace Elements Through
Coal-Fired Power Plant. Environmental Science and Technology.
9(10):973-979. October 1975.
A-30
-------
APPENDIX A-7
1. Calculation of Total Arsenic Applied Versus the Amount of Total Arsenic
Found in Gin Trash
The purpose of calculation No. 1 is to compare or check the arsenic
emission estimates that can be made by using either arsenic acid application
data or gin trash arsenic content data. Approximately equivalent emissions
estimates by using both methods indicates that the data associated with both
techniques is fairly reliable.
I. Determination of Total Arsenic Applied
A. Assumptions
1. Application efficiency «• 5% (Section 7.1.3.5)
2. Total hectares affected = 850,000 (Section 7.1.1)
3. Arsenic acid application rate = 4.9 kg/ha
(Section 7.1.2)
4. The ratio of elemental arsenic in arsenic acid (H-AsO.)
is 0.52. J 4
B. Calculations
1. (850,000 ha) x (4,9 kg/ha) - 4.17 x 106 kg
2. (4.17 x 10 ) x 0.05 - 2.09 x 105 kg of total
,. arsenic acid-on the plants
3. (2.09 x 10 kg) x 0.52 = 1.09 x 10 kg of total
elemental arsenic on the plants
II. Determination of Total Arsenic in Gin Trash
A. Assumptions
1. 334 kg lint cotton per hectare of seed cotton harvested
(Chapter 7, ref. 8)
2. 385.6 kg trash per bale of lint cotton (Section 7.1.6.1)
3. Total arsenic in trash = 225 ppm (Section 7.1.4)
B. Calculations
1.
2.
334 kg lint/ha * 226.8 kg lint/bale -1.47 bales/ha
1.47 bales/ha x (850,000 ha) - 1.25 x 10 bales
A-31
-------
3. (1.25 x lo!j bales) x 385.6 kg/bale = 4.82 x 108 kg trash
4. (4.82 x 10 kg trash) x 0.000225 = 1.08 x 10 kg of
total elemental
arsenic in the gin
trash
III. Conclusion
Considering the inaccuracies that are part of the estimating
process, the two methods of determining arsenic emissions are-
fair ly equivalent. The two answers, 1.08 x 10 and 1.09 x 10 ,
represent a difference in the estimating techniques of less than
1 percent.
A-32
-------
2. Example Calculation on Cotton Gin Model Plant Process Emission-s
A. Assumptions
1. Total arsenic content of gin trash = 225 ppm
(Section 7.1.4)
2. 99 percent of trash >_ 74 microns (Section 7.2.2.1)
3. Trash greater than 30 microns removed at 99 percent
efficiency (Section 7.2.2.1)
4. Trash smaller than 30 microns removed at 40 percent
efficiency (Section 7.2.2.2)
5. High pressure section emissions constitute 46.4 percent
of total gin emissions (Chapter 7, ref. 40)
6. 4 bale/hr model will be used as the example, total trash
input to the gin * 385.6 kg/bale or 1542 kg/hr
7. The 7, 12, and 20 bale/hr estimates were calculated
using same assumptions except the trash input rate
changes.
B. Calculations
1. 1542 kg/hr x 0.99 - 1527 kg/hr above 74 microns
1542 kg/hr - 1527 kg/hr = 15 kg/hr below 74 microns
2. 1527 kg/hr x 0.01 = 15.27 kg/hr emitted particulate
15 kg/hr x 0.60 - 9 kg/hr emitted particulate
3. 15.27 kg/hr + 9 kg/hr = 24.27 kg/hr emitted
particulate from total gin
4. 24.27 kg/hr x 0.464 = 11.26 kg/hr emitted particulate
from high pressure section of gin
5. 11.26 kg/hr x 0.000225 = 0.0025 kg/hr of arsenic emitted
from high pressure section of
gin
A-33
-------
3. Example Calculation to Determine Low Pressure Section Arsenic Emissions
A. Assumptions
1. Arsenic content of low pressure emissions is 1.33 ppm
(Section 7.2.2.1)
2. Low pressure section particulate emissions are equal
to the difference of total gin particulate emissions
minus high pressure emissions as presented in
calculation No. 7-2.
B. Calculations
4 bale/hr gin
1. 24.27 kg/hr - 11.26 kg/hr = 13.01 kg/hr of particulate
emissions from low pressure
fi section s
2. 13.01 kg/hr x (1.3 x 10"b) = 1.73 x 10~5 kg/hr of
arsenic emissions from low
pressure section
7 bale/hr gin
1. 42.91 kg/hr - 19.91 kg/hr - 23 kg/hr of particulate
, emissions
2. 23 kg/hr x (1.3 10~°) - 3.06 x 10~3 kg/hr of arsenic
emissions
12 bale/hr gin
1. 73.56 kg/hr - 34.13 kg/hr = 39.43 kg/hr of particulate
, emissions ,-
2. 39.43 kg/hr x (1.3 x 10 ) - 5.24 x 10 kg/hr of
arsenic emissions
20 bale/hr gin
1. 122.6 kg/hr - 56.9 kg/hr ^65.7 kg/hr of particulate
missions -
8.62 x 10"D kg/hr
arsenic emissions
, emissions -
2. 65.7 kg/hr x (1.3 x 10 ) - 8.62 x 10 kg/hr of
A-34
-------
APPENDIX A-8
1. Calculation of ESP Energy Requirements for Glass Plant EEC
A. Assumptions
1. 0.035 kW of electricity is required per cubic meter/min of
gas.
2. There are 8400 hr/yr of operation.
3 Large model plant has 1800 m /min, medium model plant has 906
m /min, and the small plant has 453 m /min.
B. Calculations
1. Large Model Plant Energy Demand
1800 m^ 0.035 _kW 56.9 Btu/min 8400 hr 60 min 2.93 x 10~4 kWh
min cmm kW yr hr Btu
5.29 x 105 kWh
2. Medium Model Plant Energy Demand
906 m^ 0.035 kW 56.9 Btu/min 8400 hr 60 min 2.93 x 10~4 kWh
min cmm kW yr hr Btu
2.69 x 105 kWh
3. Small Model Plant Energy Demand
453 m3 0.035 kW 56.9 Btu/min 8400 hr 60 min 2.93 x 10~4 kWh
^H.H V ."I..- V L. --" _ V U « _ -
min cmm kW yr hr Btu
1.34 x 105 kWh
A-35
-------
APPENDIX A-9
1. Comparison of Calculated and Measured Osmose Wood Preserving Emissions
A. Assumptions
1. From the EPA data on Pennwalt, Inc. 2.6 percent of the
arsenic trioxide that is dumped is emitted as a fugitive
particulate (Chapter 9, reference 35).
2. From the EPA/Pedco trip report and the Tennesee State
permit data on Osmose, plant uses 14,515 kg/day of
arsenic trioxide (Chapter 9, references 14 and 33).
3. Osmose operates one 8 hour shift per day from EPA/Pedco
trip report.
4. Fabric Filter (FF) control device is 99 percent
efficient, scrubber device is 95 percent efficient.
B. Calculation of Controlled Emissions
1. 14.515 kg/day .
8 hr/day' 1'814 kg/hr
2. 1,814 kg/hr x 0.026 (emission factor) = 47.2 kg/hr As^
emitted as
uncontrolled
fugitive
3. 47.2 kg/hr x 0.01 (FF emissions reduction) = 0.472 kg/hr
As.O,
emitted
from the FF
4. 0.472 kg/hr x 0.05 (scrubber emissions reduction) =
0.024 kg/hr
As 0
emitted
from the
scrubber
C. Measured Controlled Emissions
1. 0.019 kg/hr As.O. as measured during compliance test for
TN particulate regulation on 2/81.
A-36
-------
D. Summary
A small increase in the control efficiency, particularly
for the FF would make the two emission rates just about
equal. For example, if the FF was 0.2 percent more
efficient, which is certainly possible with a FF, the two
rates would be the same.
A-37
-------
2. Calculation of Arsenic Emissions from the Voluntary Purchasing Group
(VPG) Arsenic Acid Plant
Information learned from the Texas Air Control Board (TACB) regional
office in Ft. Worth indicated that in 1980 the VPG plant produced
2.27 million liters of arsenic acid. From other TACB state file data it was
learned that the company uses 0.99 kg of arsenic trioxide per liter of
arsenic acid produced. Therefore, total arsenic trioxide raw material use
was put at 2.25 million kg.
Based on confidential data submitted to EPA on two arsenic chemical
plants, an emissions factor of 8.1 x 10 kg of arsenic trioxide was
Mg arsenic trioxide input
developed. The controls of the VPG plant are very similar to those in use
at the plants from which the emissions factor was developed. Inorganic
arsenic emissions can, therefore, be expected to be of a similar magnitude.
The annual VPG inorganic arsenic emissions were calculated as follows:
8.1 x 10~6 As203 228 Mg As^ 0.018 kg As^
X ~
Mg As 0 input yr yr
The hourly emissions of the VPG plant could not directly be determined
because no data were available on the total hours of dumping operations at
the plant. As a result the average annual hours of operation were estimated
by averaging the operation periods for the other plants in this source
category. This average equals 2723 hours per year. Hourly emissions of the
VPG plant were then calculated by the following:
0.018 kg As20 1 6.6 x 10~6 kg As^
x x
yr 2723 hr/yr hr
In summary, annual inorganic arsenic emissions were determined to be
0.018 kg/yr, while the hourly emission rate was put at 6.6 x 10 kg/hr.
These estimates are consistent with other arsenic acid plants employing
similar arsenic trioxide controls.
A-38
-------
APPENDIX B - BASES FOR THE CALCULATION OF ARSENIC
EMISSIONS CAPITAL AND ANNUALIZED CONTROL COSTS
The purpose of this appendix is to present the methodologies, assump-
tions, and supporting data used to estimate capital and annualized costs of
arsenic emissions control in each of the source categories. The appendix
has been organized to correspond with the arrangement and order of the
individual source category chapters in the report. For example, information
on the control costs for secondary lead smelting, which is Chapter 6 in the
report, would be found in Appendix B-6.
B-l
-------
APPENDIX B-3
1. Calculation of Incremental EBC Costs for Primary Copper Smelting
Process Emissions Control Systems
As discussed in Chapter 3, three primary copper smelters that are
expected to continue operations will not be controlled at the EBC level for
arsenic when compliance with the regulatory baseline is achieved. These
smelters are Kennecott/Hayden, Magma/San Manuel, and Copper Range/White
Pines. Each of these smelters would require spray chamber/gas cooling
systems on one or more process offgas streams to reduce the temperature of
existing ESP units that currently operate at elevated temperatures.
As an example of the methodology used to estimate the associated
incremental EBC costs, consider the Kennecott/Hayden smelter. A schematic
diagram of the plant is shown in Figure 1 of Appendix A-3. Currently the
roaster and converter offgases are combined and sent to an acid plant unit.
The reverberatory furnace offgas stream is sent to an ESP operated at 650°F.
Thus, the roaster and converter offgas streams are controlled at the EBC
level as defined in Chapter 3, but the reverberatory offgas stream is not.
The control equipment required to reach EBC for the reverberatory offgas
stream is a spray chamber system capable of cooling the gas to about 110°C
(230°F).
The characteristics of the reverberatory furnace offgas stream are:
3
gas flowrate = 125,000 scfm (59 Nm /sec)
temperature = 650°F (345°C)
S02 content =0.4% „
arsenic content = 27.8 mg/Nm (13.0 Ib/hr)
The purchase cost of a spray chamber system capable of providing the
required cooling is obtained using the spray chamber cost equation given in
reference 4:
Purchase Cost (December 1977) - 0.235 x (gas inlet flowrate, acfm)
dollars + 43,000
For the Kennecott/Hayden reverberatory furnace offgas stream:
Purchase Cost (December 1977) - 0.235 x (650 + 460)°R x 125,000 + 43,000
dollars (70 + 460)°R
= $105,000 (December 1977)
B-2
-------
The purchase cost estimate is updated to final quarter 1981 dollars:
Purchase Cost (final quarter) = $105,000 x 305.3
1977 dollars 204.1
= $157,000 (final quarter 1981)
The required water flowrate for the spray chamber is calculated
assuming that the heat removed from the gas stream is transferred to the
water as latent heat (i.e., ignore sensible heat effects and assume
adiabatic operation). Under this assumption an energy balance yields:
M x
gas
stream
(C ) x
P gas
stream
(AT)
gas
stream
AH . _
vaporization
where:
= required water flowrate
(gmoles /minute)
M = offgas flowrate (gmoles/minute)
gas
stream
C = specific heat of gas stream (cal/gmole - °K)
AT = desired temperature change in gas stream (°C)
AH . , = latent heat of vaporization of water
r
"2
The molar flowrate of the gas stream is calculated using the ideal gas law:
M = PV
gas
stream
gas RT
M =1 atm x 59 m /sec x 60 sec/min
gas -?—r
stream 82.05 x 10 m -atm x 293°K
gmole-°K
« 1.5 x 10 gmole/minute
The specific heat of the gas stream is estimated over the applicable
temperature range of 110°C - 345°C using the weighted average specific heat
corresponding to a composition typical of a reverberatory offgas:
B-3
-------
HO 15%
CO, 8%
0 10%
$2 67%
Using this approach, (C ) =7.5 cal
P g*S l^o~le-°K
stream °
The desired temperature change in the gas stream is:
(AT) - (345 - 110)°C = 235°C
gas
stream
The latent heat of vaporization of water is AH . fc. = 9717 cal/gmole
vaporization
H20
Substituting into the energy balance equation yields:
WL - 1.5 x 103 gmole/min x 7.5 cal/gmole-°K x 235°K
2 9717 cal/gmole
4
= 2.7 x 10 gmole/min water = 490 liters/minute *
130 gallons/minute
So the water flowrate required to provide the desired cooling is 130 gpm.
The power required to pump 130 gpm of water to 100 psig (typical for
spray chamber) is estimated by the equation:
Power (kw) » 0.746 x (GPM) x AP
3960 x 0.6
where: GPM = water flowrate, gallons per minute = 130 gpm
AP = required gauge pressure, ft H_0
Substituting into the equation and assuming that the pump operates
8760 hours per year yields:
4
Energy requirement = 8.3 x 10 kWh/yr
The total installed capital cost and the annualized cost of the spray
chamber are then estimated as shown below.
A. Total Capital Cost
Purchase cost * $157,000
Direct Installation Cost 0 0.5 Purchase Cost - $78,500
B-4
-------
Indirect Installation Cost @ 0.5 Purchase Cost = $78.500
Total Capital Cost - 157,000 + 78,500 + 78,500
= $314,000 for spray chamber system
B. Total Annual Cost
1. Operating Labor
a. operator (% man-hour/shift, $13.80/man-hour)
$13.80 x h man-hour x 3 shift x 365 day = $7.7 x 103/yr
man-hour shift day yr
b. supervision (15% of operator cost)
0.15 x 7.7 x 103 - $1.1 x 103/yr
2. Operating Materials
- negligible
3. Maintenance
a. labor (% man-hour/shift, $15,20/man-hour)
$15.20 x h man-hour x 3 shift x 365 day - $8.3 x 103/yr
man-hour shift day yr
b. materials (100% of maintenance labor)
1.0 x 8.3 x 103 = $8.3 x 103/yr
4. Replacement Parts
- negligible
5. Utilities
a. electricity @ $.0432/kw-hr
/ O
$0.0432 x 8.3 x 10 kw-hr - $3.6 x 10 /yr
kw-hr yr
b. cooling water @ 0.10/10 gallons
0^10 x 130 gal x 60 min x 8760 hr_ = $6.9 x 103/yr
10 gal min hr yr
B-5
-------
6. Overhead (@ 80% of total labor)
0.8 (7.6 + 1.1 + 8.3) x 103 $/yr - $13.6 x 103/yr
7. Property Tax, Insurance, Administration (@ 4% of total
capital cost)
0.04 ($314,000) - $12.6 x 103/yr
8. Capital Recovery (@ 16.7% of total capital cost)
.167 ($314,000) = $52.4 x 103/yr
Total annual cost = (7.6+1.1+8.3+8.3 +33.6 + 6.9 + 13.6 + 12,6
of spray chamber system +52.4) = $114.4 x 10 /yr
2. Calculation of Incremental EEC Costs for Primary Copper Smelting
Fugitive Emissions Control Systems
As discussed in Chapter 3, most of the primary copper smelters will not
be controlled at the EEC level for fugitive arsenic emissions when
compliance with the regulatory baseline is achieved. The additional
equipment that will be required to reach EEC for fugitive arsenic emissions
includes a fabric filter system for furnace tapping ventilation gases, and
an air curtain/secondary hood system for controlling converter fugitives.
A. Capital and Annual Costs for the Furnace Tapping Fabric Filter
System
The capital and annual costs for the furnace tapping fabric
filter system were estimated using references 1 and 2. The
purchase cost of the required fabric filter system was estimated
using the log-linear cost curve for fabric filter systems in the
non-ferrous smelting industry shown in reference 1. A regression
model was used to standardize cost interpolations from the curve.
Corrected to final quarter 1981 dollars, the purchase cost
regression model is represented by:
Purchase Cost - 0.0713 Q°'7854
where: Q = gas flowrate (acfm)
Purchase Cost « thousands of final quarter 1981 dollars
B-6
-------
The purchase cost estimates include the flange to flange fabric
filter module, the bags, fans, motors, ladders, walkway, screw
conveyor, dampers, and expansion duct. They do not include other
ductwork or new stack costs.
The total installed capital cost was estimated as 2.56 times
the purchase cost using the information in reference 2. The
annualized costs for the system were estimated using the algorithm
in reference 2. The following labor requirements, wages, and
utility costs were assumed:
- fabric filter system operating labor is $13.80/man-hour and
3 man-hour/shift are required
- fabric filter system maintenance labor is $15.20/man-hour and
1.5 man/hour/shift are required
- electricity cost is $.04/kWh, and the fabric filter system
operates at a AP of 14 inches of water.
As an example of the cost algorithm, the Kennecott/Hayden
smelter is considered. The ventilation gas flowrate for the
furnace tapping operation at Kennecott/Hayden is unknown.
However., the ventilation flowrate at ASARCO/Tacoma is known to be
1300 Nm /min (46,000 scfm). The assumption is made that the
furnace ventilation gas flowrate at a plant is proportional to the
plant capacity. Thus,
["Estimated ventilation
[flow at Kennecott/Hayden
[Ventilation flow"!
[at ASARCO/TacomaJ x
73,000 Mg/yr copper
.91,000 Mg/yr copper.
1040 Nm /min
37,000 scfm.
Using this gas flowrate and the purchase cost equation shown
above, one finds:
Purchase cost of furnace
ventilation fabric filter
.system at Kennecott/Hayden.
$276 x 10-
Total installed capital cost
of furnace ventilation fabric
.filter system at Kennecott/Hayden.
2.56 x $276 x 10:
$707 x 10-
B-7
-------
The total annualized cost for the fabric filter system at
Kennecott/Hayden is estimated to be $363 x 10 /yr. The breakdown
of the costs is shown in Table 1.
B. Capital and Annual Costs for Converter Fugitive EEC Control
Each of the primary copper smelters considered in this study
except for ASARCO/E1 Paso will not be controlled at the EBC level
for converter fugitive arsenic emissions when compliance with the
regulatory baseline is achieved. The equipment needed to reach
EBC for converter fugitives at these plants consists of air
curtain/secondary hood systems followed by particulate removal in
fabric filter units. The capital and annual costs of the air
curtain/secondary hood systems are estimated using the actual
costs reported by ASARCO for the ASARCO/Tacoma plant. The
capital and annual costs of the fabric filter systems used to
collect particulate matter captured by these devices are estimated
using the method shown in Appendix B-3, Section 2A.
The estimated cost of EBC converter fugitive control depends
on the number of converters at the plant in question. The
ASARCO/Tacoma plant has three converters, and the air
curtain/secondary hooding design is based on having two of the
three converters operating at any one time. The total installed
capital costs for the air curtain/secondary hood system reported
by ASARCO are summarized in Table 2. Updated to final quarter
1981 dollars, the cost data can be summarized by the equation:
Total installed capital cost » $ [3038 + (n-1) 1063] x 103
where n = the number of converters at the plant (n = 3 at
ASARCO/Tacoma)
The annualized costs for the air curtain/secondary hood
systems are estimated using the following assumptions:
- the operating labor requirement per converter
1 man-hour/shift @ $13.80/man-hour
- the maintenance labor requirement per converter is
0.5 man-hour/shift @ $15.20/man-hour
- the energy requirement per converter is estimated by analogy
to the ASARCO/Tacoma plant, where the main fan for a
3-converter system requires a 1250 hp drive. The cost of
energy is $0.04/kWh
- supervisory labor cost is estimated as 15 percent of the
operating labor cost
B-8
-------
TABLE 1. TOTAL ANNUALIZED COSTS FOR FURNACE FABRIC FILTER
SYSTEM AT KENNECOTT/HAYDEN
DIRECT OPERATING COSTS
1. Operating Labor
a. Operator „
$13.80 x 3 man-hour x 3 shift x 365 day = $45.3 x 10J/yr
man-hour shift day yr
b. Supervisor _ ,
0.15 (la.) - 0.15 x $45.3 x 10 /yr = $6.8 x ICT/yr
2. Maintenance
a. Labor
$15.20 x 1.5 man-hour x 3 shift x 365 day = $25.0 x 10^/yr
man-hour shift day yr
b. Materials _ „
1.0 (la.) = 1.0 x $45.3 x l(T/yr = $45.3 x 10^/yr
3. Utilities
$0.04 x 8.9 x 105 kWh = $35.5 x 103/yr
kWh
4. Overhead
0.8 (total labor) = $0.8 (45.3,+ 6.8 + 25.0) x 103/yr
- $61.7 x 10J/yr
5. Property Tax, Insurance, Administration
3
.04 (total installed capital cost) = 0.04 (707 x 10 )
= $28.3 x l
-------
TABLE 2. TOTAL INSTALLED CAPITAL COST OF THE ASARCO/TACOMA
AIR CURTAIN/SECONDARY HOOD SYSTEM
REPORTED COST IN AUGUST 1980 DOLLARS"
ITEM
1. Direct Labor and
Materials
Hood and Door
Air Curtain System
Flue System & Fan
Electrical Distribution
System
2 . Engineering
3. Contractor's Field
Distributables
4. Escalation
5 . Contingencies
6. Fees, etc.
7 . Taxes
1st Converter
249,450
72,600
1,018,150
45,167
266,000
283,000
279,000
230,300
69,000
35,000
Other Converters
264,450/conv
69,800/conv
182,100/conv
45, 167/conv
10,000/conv
86,000/conv
175,000/conv
78,650/conv
23,500/conv
19,000/conv
Total of items 1-7
2,638,000
908,500/conv
B-10
-------
s
o
£
co
a
o
CO
CO
H 1
CK^
s^.
CJ CO
hH fcj
53 cd
W rH
CO r-l
fy^ A
*™*3 »rt
r^i ^.
> oo
1— 1 , >>
B OT •*«.
IM
1-1 o en
« 0
o in ~*
H o
CJ
U (0
o o
H U «•
•a IH
oi a cn
•H 4J O
|H -H —4
a o.
tti CO
CO CJ
H4
U
to
o
1-4 cn
a o
c
e
<
t— (
0 O
H CJ «•
•3 i-i cn
A) 03 O
«H u i— I
a o.
jj (0
CO CJ
B
cn
o u
CJ !N
i-H
Q
3 cn
e o
B —
CO M
4J 03
0 0
H CJ
ID *— 1
oi co cn
•-( 4J 0
n a.
CO CJ
C
u
•o u
CU CO
M •^
•H
O" O
QC Cb
to
ki
4k 01
e
o
CJ
0 0 0 0 f- ff.
\o -^ *o O m o
-« CN «N cn cn •»
*O oo a* «ff -* r^
r^ sA o O oo cn
O oo *o cn cy *o
A • «l • * •
^•4 —4 ^4
cn in 10 oo r*» «— '
co c^ &* oo r** so
vO CO O C*i "^ vO
O CTN r* CJ^ cn xo
cn in cn \o oo ^*
cjs ^ cn cr^ m i^
—i oi cn cn •» in
p*» in cn CN o co
vo -» CN o oo m
cn rg m oo o cn
->—•-< {M CM
vo o> CM in oo *-*
— * CM CN cn cn -o1
§o cp o o o
O O O O O
0 C5 C3 C3 C5 O
en § Co cn § S
-• — < CM en •«• ^i
CN cn *T in vo f^
B-ll
-------
- maintenance material costs are estimated as 100 percent of
maintenance labor costs
- the cost of replacement parts is negligible
- overhead costs are estimated as 80 percent of total labor
costs
- property taxes, insurance, and administration are estimated
as 4 percent of the total installed capital costs
- the capital recovery factor is 0.167.
The fabric filter system costs are estimated as shown in
Appendix B-3, Section 2A. The flowrate Q is estimated by analogy
to the flow requirement at ASARCO/Tacoma. The ASARCO design
flowrate of 200,000 acfm assumes that at any time, 2 out of 3
converters could possibly require the maximum possible individual
converter flowrate of 100,000 acfm. Thus, the design flowrate for
an n-converter smelting plant was estimated using the equation:
Q = n x 2_ x 100,000 acfm
3
The total installed capital costs and total annual costs for
copper smelters having from 2 to 7 converters are summarized in
Table 3. As an example, the annual cost of the air
curtain/secondary hood system with a fabric filter system for the
Kennecott/Hayden plant (three converters) is $2,140,000.
B-12
-------
REFERENCES FOR APPENDIX B-3
1. U. S. Environmental Protection Agency. Arsenic Emissions from Primary
Copper Smelters - Background Information for Proposed Standards.
Preliminary Draft. Research Triangle Park, N. C. February, 1981.
2. Vatavuk, William M., and Robert B. Neveril. Factors for Estimating
Capital and Operating Costs. Chemical Engineering. November 3, 1980.
3. Marsh, A. 0., Jr. ASARCO Incorporated: Converter Secondary
Hooding/Tacoma Plant. Salt Lake City, Utah. January 22, 1981.
4. Vatavuk, William M., and Robert B. Neveril. Estimating the Size and
Cost of Gas Conditioners. Chemical Engineering. January 26, 1981.
B-13
-------
APPENDIX B-5
1. Calculation of Incremental EEC Costs for the New Jersey Zinc Plant
The equipment required for the New Jersey zinc plant to reach EEC for
arsenic control consists of a spray chamber to cool the waelz kiln offgas
stream prior to collection in the existing fabric filter unit. The total
installed capital cost and the annualized operational cost of the spray
chamber system were estimated using the methods shown in Section 1 of
Appendix B-3. The 220,000 scfm waelz kiln offgas stream needs to be cooled
from 140 to 110°C (280 to 230°F) prior to baghouse treatment. Using the
methods shown in Section 1 of Appendix B-3 and assuming the same flue gas
composition, the required water flowrate for the spray chamber system is
32 gpm, the total installed capital cost is $342,000 (final quarter 1981
dollars), and the total annual cost of operation is $112,000.
B-14
-------
APPENDIX B-6
1. Basis of Cost Estimates for Secondary Lead Smelters
MRI provided the following cost estimate for a single scrubber to
control S0_ emissions from a typical small and medium sized lead smelter.
Capital Costs:
Scrubber, with installation
Pump
Stack
TOTAL
$300,000
20,000
5,000
$325,000
Annualized Costs:
Capital charges ($325,000 x 0.13147,
assuming 10 percent for 15 years)
Operating labor
General maintenance
Electricity-fan power
Electricity-pump power
Water
Overhead
Property tax, insurance
TOTAL
$ 42,700
18,100
34,600
57,900
6,100
26,900
28,300
13,000
$227,600
Notice that the annualized cost does not include waste disposal costs.
These costs were derived from data submitted by Tonolli Corporation and
General Battery during the secondary lead NSPS revision project. These
scrubber costs represent a scrubber sized to handle approximately
50,000 acfm with a design AP of 12 inches water and a slurry flow rate of
1,500 gpm. Capital costs for the three plant sizes used to model arsenic
emissions were derived from the above data using scaling factors. Analysis
of the cost data in CARD show that the cost of scrubbers follow an
0.6 exponent.
Cost of a = Cost of b
Capacity of a
Capacity of b
0.6
B-15
-------
Using this rule, capital costs for scrubbers for small, medium, and large
secondary lead smelters were developed. It was assumed that two medium
sized scrubbers would be needed by the large plant. Operating costs were
estimated from the MRI data assuming that maintenance was a fixed proportion
of total installed costs, electricity costs would vary linearly with
scrubber size, and operating labor would remain constant for each scrubber.
The costs given by MRI for water seemed very high and probably included
the lime slurry costs. This cost was adjusted downward to reflect water
usage alone rather than limestone or caustic usage. Using these procedures
the costs in Table 1 were derived:
TABLE 1. COSTS OF SCRUBBERS TO CONTROL SECONDARY LEAD SCRUBBERS
Plant Size
Flow dscfm
acfm
Total installed
cost
Small
20,000
24,000
$194,000
Medium
36,000
43,000
278,000
Large
72,000
86,000
556,000
Split into two streams
2 scrubbers
Annualized
capital cost
@ 13147
Operating labor
General
maintenance
@ 11% of cap
Electricity
costs for
pumps and fan
Water use
Overhead
Property taxes
+ ins. @ 4%
$ 27,600 39,570
$ 34,600
37,800
415
18,300
8,400
34,600
54,200
31,000 56,000
2,150
26,200
12,040
73,097
69,200
108,400
112,000 @ $.04 kwh
4,300
52,400
24,080
-------
POWER REQUIREMENTS FOR PUMPS AND FANS
FANS
Assume scrubber pressure drop is 12-15 inches water.
Assume stack pressure drop is 3 inches water.
Assume ductwork damper elbows etc. is 9 inches.water.
(Approximately 300 feet equivalent length of ductwork)
TOTAL PRESSURE DROP is 27 inches water.
Plant Small Medium
Flow ACFM 24,000 33,000
AP " water 27 27
Fan motor HP 162 288
(from CARD)
PUMP POWER REQUIREMENTS
Assume scrubber water flow is 10 gpm per 1000 ft gas flow
Total head £ 100 feet. Pump efficiency s 50%.
Pump
HP 12 22 44
Total Power requirements - Fan and Pump
HP 174 310 620
kwh/h 130 232 463
WATER USE FOR SCRUBBERS
Incoming exhaust gases from fabric filter has a water content of 3% @
176°F.
At the scrubber exit, the exhaust gases will be saturated with water.
The estimated exit temperature for the scrubber is 100°F.
At 100°F 1 pound of dry air can contain 0.044 Ibs of water.
3% water by volume = .03 moles/mole of mixture
= .03 moles/.97 moles dry air
.03 x 10 lb/mole/.97 x 29 Ibs/mole
.54 Ib H 0/28.13 Ibs dry air
.019 Ibs H20/lb dry air.
Therefore the amount of water absorbed by the exhaust gases from the
scrubber will be (.044 - .019) Ibs per Ib of dry gas.
.025 Ib/lb dry gas s 2 Ibs H20/1000 scf.
B-17
-------
WATER USE BY PLANT SIZE
Plant Size
Flow dscfm
Water use gpm
Small
20,000
4.8
Medium
36,000
8.6
Gallons/year
@ 6,000 hours
operation
1.7 x 10
3.1 x 10
6.2 x 10
6
YEARLY ARSENIC REMOVALS BY SCRUBBERS
The model plant emissions factors are as follows:
Plant
Arsenic from fabric filter
Ib/hr
Arsenic from fabric filter/
scrubber combination
Ib/hr
Arsenic removal
Ib/hr
Yearly removal
Ibs
Particulate removal
.05
.07
420
Medium
.2
.08
.12
720
.16
.24
1,440
Test data indicates that the fabric filter/scrubber combination has
higher particulate emissions than the fabric filter alone. Therefore, no
incremental particulate removal can be estimated. Solid waste disposal
costs could be significant if the scrubber was used for joint SCL and
arsenic control. Large amounts of sulfate sludge would be generated which
would need to be disposed of.
For a water scrubber - assume arsenic content of particulate matter =
3.2 percent. Sludge would be thickened to 30 percent prior to disposal.
Yearly sludge quantities would be as follows:
Size
Dry Particulate
tpy
Sludge
-------
REFERENCES
1. (DRAFT). Development of NSPS Project Conclusions Report for Secondary
Lead Smelting Industry. Prepared by MRI. ESED Project Number 80/33.
January 27, 1981. p. 21.
2. Neveril, R. B. Capital and Operating Costs of Selected Air Pollution
Control Systems. Prepared for the U. S. Environmental Protection
Agency. EPA Publication Number 450/5-8-002.
B-19
-------
APPENDIX B-7
1. Equipment Cost of Long-cone Cyclones for Low Pressure Exhaust Control -
20 bale/hr Gin
!• References for Cost Data
A. Letter from Furr, R., Anderson-Bigham Sheet Metal Works, Inc.
to Brooks, G., Radian Corporation. March 22, 1982.
II. Calculations
A. From Table 7-2, the arsenic emission rate is 0.072 kg
(0.152 lb)/yr. This is to be controlled by 94 percent.
3
B. Total Low Pressure Flow = 2443 m /min, therefore the
following cyclone arrangement was used:
3
Gin Emission Pt. Airflow (m /min) No. of Cyclones
6 599 2 82.6 cm Quads
7 594 2 82.6 cm Quads
8 945 3 86.4 cm Quads
9 306 1 82.6 cm Quad
C. Cyclone costs are as follows:
- 82.6 cm Quad - $2464.80/unit x 5 - $12,324.00
- 86.4 cm Quad = $2606.90/unit x 3 = 7,820.70
$20,144.70
D. Control cost-effectiveness is,
- ($20.144.70) x (Cap. Recovery factor of 0.163) =
(0.072) x (0.94)
$48,516/kg of arsenic removed.
B-20
-------
2. Bases for the Capital and Annualized Costs of Cotton Gin EEC
I. Bases for Capital Costs
A. Total Direct Equipment Costs (TDEC)
1. Primary Equipment Cost = Vendor Cost Data
2. Taxes » 3% of Primary Equipment Cost
3. Freight = 5% of Primary Equipment Cost
B. Direct Installation Costs (DIC)
1. Foundation & Supports = 6% of TDEC
2. Handling & Erection = 40% of TDEC
3. Electrical = 1% of TDEC
4. Piping - 5% of TDEC
C. Indirect Installation Costs (IIC)
1. Engineering & Supervision = 10% of TDEC
2. Construction & Field = 10% of TDEC
3. Construction Fee = 10% of TDEC
4. Startup & Performance Test = 2% of TDEC
5. Contingencies = 3% of TDEC
D. Total Capital Cost = (TDEC + DIC + IIC)
II. Bases for Annualized Costs
A. Direct Operating Costs
1. Utilities = $0.04/kWh of electricity
2. Waste Disposal - No additional costs as a result of EEC
B. Indirect Operating Costs
1. Taxes & Insurance = 2% of Total Capital Cost
2. G & A = 2% of Total Capital Cost
3. Capital Recovery - 16.3% based on 10% interest/10 yr.
equipment life
References for the cost bases were taken from the following:
1. Vatavuk, W. M. and R. B. Neveril. Chemical Engineering.
November 3, 1980. pp. 157-162.
B-21
-------
APPENDIX B-8
1. Bases for the Glass Manufacturing EEC Option Cost Analysis
A. Bases for the Capital Costs
1. The bases for the capital cost of the glass
manufacturing EEC option are identical to those
presented in Appendix B-7 for cotton gins.
2. All costs were updated to last quarter 1981 dollars
using the Chemical Engineering Plant Cost Index
published by McGraw Hill. The subcategory used was
Equipment, Machinery, and Supports which had an index
value of 332.7. The reference for this index is:
Chemical Engineering, Volume 89, No. 7. April 5, 1982,
p. 7.
3. Costs for gas cooling devices were taken from: Chemical
Engineering, Volume 88, No. 2, January 26, 1981, P. 132.
These costs were in 1977 dollars which were updated by
using the Chemical Engineering Plant Cost Index given
in: Chemical Engineering, Volume 85, No. 11, May 8,
1978, p. 190. The index value is 220.9.
4. ESP costs were in 1978 dollars which were updated by
using the same index given in: Chemical Engineering,
Volume 87, No. 1, January 14, 1980, p. 7. The index
value is 247.6.
5. From EPA data on existing glass plants using'ESP control
devices, on average ESP plate area of 1.1 ft /scfm was
selected to cost out ESP devices for the arsenic glass
manufacturing model plants in Chapter 8. The 1.1 value
does not represent a plate area designed to equal EBC.
It was selected as a conservative value to apply to
general ESP cost curves based on plate area. This study
did not specify a design plate area associated with EBC.
The reference for the 1.1 value is: Glass Manufacturing
Plants, Background Information: Proposed Standards of
Performance, EPA-450/3-79-005a, U. S. Environmental
Protection Agency, June 1979, p. 4-23.
B-22
-------
6. Costs for the ESP control devices were taken from the
same reference given in No. 5 above except that the
correct page is p. 8-58.
B. Bases for the Annualized Costs
1. The bases for annualized costs given in Appendix B-7 for
cotton gins are also applicable to glass manufacturing.
2. Labor costs are as follows:
- Operating labor is $13.80/hr
- Maintenance labor is $15.20/hr
These costs were obtained from: Chemical Engineering,
Volume 87, No. 22, November 3, 1980, pp. 157-162. They
were updated by using the Engineering and Supervision
subcategory of the Chemical Engineering Plant Cost
Index.
3. For gas cooling devices operating labor is assumed to be
zero. Maintenance labor is 0.5 man-hours per shift and
maintenance materials is equal to the cost of the labor.
Overhead is equal to 80 percent of total labor. The
additional parameters of the gas cooler annualized cost
are:
- utilities
- taxes and insurance
- G & A
- capital recovery
4. For the ESP devices operating labor is 1 man-hour per
shift. Supervisory labor equals 15 percent of operating
labor. Maintenance labor is 0.5 man-hours per shift
with maintenance materials equal to the labor cost.
Overhead charges equal 80 percent of total labor. The
remaining factors addressed in the annualized ESP costs
were:
- utilities
- taxes and insurance
- G & A
- capital recovery
-------
APPENDIX C - METHODOLOGY FOR THE ECONOMIC IMPACT CALCULATIONS
The purpose of this appendix is to present the methodology used to
calculate the economic impacts of the recommended EEC options for the
primary copper and zinc smelting, secondary lead smelting, cotton ginning,
and glass manufacturing source categories.
C-l
-------
ECONOMIC IMPACT ANALYSIS METHODOLOGY
This arsenic source survey economic impact assessment utilized a
methodology based on a simplified price impact analysis. This methodology
calculates the revenue and price increases required by model plants to
maintain the same net present values (NPV) before and after installation of
arsenic-controlling EEC equipment. This revenue calculation relies on a
single derived equation that requires several types of input data for each
model plant. In addition to the methodology description, this appendix
includes a description of the input data used for the price impact analysis.
A. Revenue and Price Impact Analysis
The purpose of this analysis is to determine the revenue increase that
exactly offsets arsenic emissions control costs so that the NPV of the model
plant remains constant or the NPV of the incremental cash flow is zero at
the stated weighted average cost of capital. In this analysis, capital
costs, operating and maintenance costs, investment life, income taxes and
inflation need to be taken into account. A nominal discount rate is used
for the analysis because it is more intuitive and because most equity
capital cost data are available in nominal terms. The use of a nominal
discount rate requires that revenues and operating costs be properly
inflated. The revenue increase that the analysis calculates is expressed in
base year dollars, in this case the fourth quarter of 1981. The required
revenue increase is converted to a required unit price increase by dividing
the revenue by the annual sales volume. This step requires the assumption
that annual sales volume is constant, indicating perfectly inelastic demand.
The derivation of the basic formula for the price impact analysis
requires the following assumptions:
- the incremental pollution control investment occurs at the
beginning of the base year.
- there is no replacement investment for emission control.
emission control investments have zero salvage value.
- no differential inflation occurs among the cost or revenue items.
- the weighted average cost of capital and the marginal income tax
rate remain constant during the life of the investment.
C-2
-------
- depreciation is based on the 1981 Economic Recovery Act,
Accelerated Cost Recovery System (ACRS) five-year rates.
- a 10 percent investment tax credit is applicable on the total EBC
investment and is realized in the year following the investment.
- NPV of the model plant remains constant at the stated weighted
cost of capital.
The derivation of the basic formula of the methodology is presented in
the context of the standard definition of NPV:
n
CF
y
NPV - -I + Z (1)
o
y-1
where,
NPV * net present value (in base year dollars)
I » investment (period y=0)
y = time period
n = investment life
CF = projected operating cash flows in period y
d y = discount rate or cost of capital
Cash flow, CF, is defined as revenues, R, less operating and mainte-
nance costs, OM, less income taxes, T.
CF - R - OM - T (2)
y y y y
Interest expense is excluded because interest and its tax effect are
accounted for in the discount rate. Revenue and operating costs will
inflate at the same rate, thus, nominal revenue (and O&M) equals the product
of the inflation factor, (l+inf)y, and constant dollar value in y=0 dollars.
That is,
and,
Ry = (RQ) (l+inf)y (3)
OM - (OMQ) (l+inf)y (4)
Income taxes, T , are computed on the basis of nominal current income.
Income taxes are defined to equal the tax rate, t, times taxable net income.
The investment tax credit (ITC ) is subtracted directly from income taxes.
C-3
-------
Taxable income equals revenue less operating and maintenance expense less
depreciation. Thus,
T - t(R - OM - D ) - ITC (5)
y y y y y
Substituting the income tax Equation (5) into the cash flow Equation (2)
yields the following expanded expression for cash flow.
CF - R - OM - T (6a)
y y y y
= R - OM - t(R - OM - D ) + ITC (6b)
y y y y y y
= R - OM - tR + tOM + tD + ITC (6c)
y y y y y y
- (l-t)R - (l-t)OM + tD + ITC (6d)
y y y y
Further, by substitution of Equations (3) and (4) into (6d), operating cash
flow, CF , may be expressed by:
CF = (l-t)R (l+inf)y - (l-t)OM (l+inf)y + tD + ITC (7)
y o o y y
Net present value can now be defined in terms of revenue, operating and
maintenance costs and income tax effects on revenue, operating costs, and
investment including effects of depreciation and the investment tax credit.
By substitution of Equation (7) into Equation (1),
(l-t)R (l+inf)y - (l-t)OM (l+inf)y + tD + ITC
no o y y
NPV = -I + Z (8).
y=1
Equivalently, each term under summation can be summed individually so that
(9)
(l-t)Ro(l+inf)y n (l-t)OMo(l+inf)y n ITC + tD
NPV --I+Z -Z + Z
o
y*1 /i^y yml /i^y y"1
With Equation (9), it is possible to solve for the annual revenue require-
ment, R , (in base year dollars) that is equivalent to incremental emissions
control investment, I , and annual operating and maintenance costs, OM .
First, it is necessary to impose the constraint that incremental NPV=0 and
C-4
-------
helpful to rearrange terms on the right hand side of Equation (9) so that
revenue, operating costs and investment related items are grouped,
(10)
0
n
E
y-i
(l-t)R (l+inf)y
o
n
y-i
(l-t)OMo(l+inf)y
- I +
o
n
I
y-i
ITC + tD
y y
Next, the terms related to emissions control investment are isolated.
Assuming a 10 percent investment tax credit in the year after investment and
five-year ACRS rates the investment, investment tax credit, and depreciation
terms on the right hand side of Equation (10) can be expressed as a product
of a constant, TAXF, and I .
(11)
n
I -
y-1
ITC + tD
y y
(1+d)-
1 -
d+d)
.22t
(1+d)'
.21t
(1+d)-
.21t
d+d)
.21t
(1+d)-
I {TAXF}
o
(12)
where TAXF is the sum of terms in brackets. TAXF is a constant that can be
repeatedly applied to different investments so long as the tax rate,
discount rate and ACRS life remain the same.
Next, I {TAXF} is substituted into Equation (10), the constant terms
are moved outside their summations, and R is isolated on the left hand side
of the equation. R is then expressed as°
C-5
-------
(l-t)OM
n
Z
y-i
(l+inf)y
U+dV
+ I {TAXF}
o
n
(1-t) I
y-i
(l+inf)y
U+d)-
(13)
Simplifying Equation (13) results in
I TAXF
o
R = OM +
o o
n
d-t) I
7-1
1+lnf
1+d
(14)
Equation (14) is the equation used to calculate the annual revenue
increase that exactly offsets EEC capital and operating costs so that the
NPV of the firm remains constant.
The unit price increase that the model plant needs to realize this
annual revenue increase is
R
(15)
PI
where, PI
R
the unit price increase in base year dollars
the required annual revenue increase in base year dollars
annual sales volume in units
C-6
-------
Finally, the percentage increase in unit price is calculated using the
formula
PPI
PI
(16)
x 100
where, PPI = percent unit price increase
PI =» unit price increase in base year dollars
P = pre-EBC unit price in base year dollars
Equations (14), (15), and (16) were used to analyze the effects of arsenic
emission controls.
B. EEC Capital and Operating Costs
The incremental capital and annualized costs for arsenic EEC are
provided in Chapters 3-9. The capital costs consisted of the incremental
capital outlay, stated in 1981 fourth quarter dollars, required for the
technologies needed by the various industries to meet the EEC arsenic
emission levels. The control equipment has a useful life of ten years
except where indicated. The annualized costs also are stated in 1981 fourth
quarter dollars and represent the incremental annual operating and
maintenance costs incurred in association with the EEC equipment.
The annual inflation rate used in this analysis was 6.4 percent for all
industries. This level of price inflation as measured by the broad-based
GNP deflator is the forecast for the next three to five years by the Value
Line Investment Survey.
These costs are summarized in Tables 1 through 5.
C. Product Prices
Information concerning product prices before installation of EEC
equipment was needed to calculate the percent price increase caused by the
controls. All price data were converted to 1981 fourth quarter dollars.
The following sources of price data were used:
- Cotton: "Supplement for 1981 to Statistics on Cotton and Related
Data 1960-1978." U.S. Department of Agriculture,
Economic Statistics Service.
- Copper, Lead and Zinc: "Minerals Yearbook" Volume I, Metals and
Minerals, 1978-79. U.S. Department of the Interior.
C-7
-------
H
Z
fid
ta*3
w
a
0
£
of
b3
r t
J
2
CO
Ed
CL,
Cb
O
u
K_j
es
2j
h*t
^^
&-
0
to
t2
M
cn
M
w
^
u
S§
<
u
,J
M
0
<
1— 1
En
Q
<
U
M
H
1
O
•
vH
w
^^
*™"J
H
H
cj < 'LT
a£ £
u p* o
"*" QJ Q)
u a
t-4 U- N^
u o
3
^^
.J M XJ
< H B
Z < 11
"-I B3 0
O b
OS X 01
£££
(-1
to *->
o o in
03 CJ b
id a
j z 3
< l-l O
3 H -0
.1^0
< te O
fee
M n
CM
— M «M V4 ^4 wrf — M
o c^ \o ^f CM m ^^
CM 00 — ' vO 00 -3 O
CM eo O so x in —4
— — — _ _
•
en m en en en en en
CM CM CM CM CM CM eM
<• •* -4T O f-J CO O
O OO GO se en oo CM
en in in en o CM o>
^4 01
u ca
•O B i-l
^4 oi 60
B ^ >H o> u> a
0> U — 1 -H O O
•o »H ^ iw S a
B *, b o w <: -^
01 a s u a 01 o
•0 S B S O 60 60
>,•••.-•.->•.-•. -w -o
CS u u w 4J O O
s 4-i *j *j 4-1 a o
-~ o o o o
O O O U U CO 11
u oi ot oi oi a. a
0£ B B B C i-4 m
< C E S B 01 01
tn oi oi ai 11 j: j:
^ o j> *^ ^j g^ g^
in in in in in in
r* r^ r» i*. i»
r«. oo — so -a- o-
P^ CM —^ en CM *y
O os CM O in oo
1—4 ff* *^ — ^
O*> so CM en r** ON
ON CM 09 — « in CM
•ff — ON O "tf *n
oo r^ r~ oo CM so
""
en en en f*i t™~' en
CM CM CM CM CM CM
CM SO so 00 09 O
-« — < in -a as so
in O so «a O — «
•H
b
B 01
0 « &
60 60 C.
« S ^ 3
o -c oi-! e •-.
•m -H in D nj a
< S S 3 ft 41
<: -- B z u
01 01 • BJ •*.•»•«
60 60 n Z B >
1) -o 01 C b
O Q C 6 1-1 4)
O O **4 CO *J 5/3
CL C/3 O
CO CO "••* b CO
o. a. oi a -4 o>
^^ ^4 U S Oi ^4
01 01 iH 60 01 JJ
X * t= ™ B it
o» a. 5 z i-i w
•
^^
u
n
CL
(0
u
u-<
o
1
u
N
4J
B
0>
u
b
01
a.
o
oo
m
CJ
E
CO
CO
re
4J
•a
a.
U
<§
a
C-8
-------
en
z
z
o
H
H
O
U
N"3
tc
a
o
fa
0
to
u
M
H
C/3
t— i
OS
te
B
^i
Oi
"
2
o
z
z
f"H
[Xw
g
o
z
1— 1
H
c2
r-1
s.
0
cs
a
CO
H
H
C C- AJ
W < B
H U CJ
O U. M
MOD
b2 CM
•3. H ~
CO
o
U
_] [d JJ
l-l S O
U hi
OS. X U
^s^
H <-«
V] ID
V O 1- '
oa u re
[d i-l
O J Z O
< M -O
= H
Z «C O
z (2 o
< HI 0
0- *—
0
H I?
O <0
u°2
oa j o
W < tJ
So
CM O
< 0
o ^
U
1-1
M U
Z M 0.
•J i-i e
U 0£ M
to c- ra
ca •-<
0
•c
^.
CO
,
t*J
1— 1 M
en 01
a
eo
a
m o o <*>
* • • •
0 0 0 0->
co — — i r-
ui r- 00 CO
-« N «r «o
0 ^ 0> -
to sr o —
fM Cl \O CO
— — — —
""^ ^ ^ **1
•-< V* — < *M
r~ o in o
oo in r~ in
o <-) os O
— i •» O 00
M *
— • CM m *?
C-9
-------
en
f^
^^
^^
^f
*•—
O-i
•J
O
2
s
o
2
^^
£
t~-t
3
Z
^C
*
w
<
o
y^ _
O
h- 1
H
i— i
H
U
0§
S
O
M
CJ
2^
<
h~*
fc.
z
O
Z
M
H
<
OM
Cu
O
•
ro
»o
J
3
H
H
fi &4 W
Id < C
H U Cl
O In U
MOO)
w e.
3 H ^
en
o
u
J Id "u
< H C
£2 S
C ^ U
K. x a
< < D.
Z H N-'
H ^>
n u
o p w
S8J
jg-3
9 H "°
IS8
< W 0
£. -w
O
E- '(0
C/S U
0 eo
8^1
w o
0. O
< 0
'3
u
Z W £.
I™4 U
J M (0
(£ ei u
OT a. is
ca — <
o
*~"
w
41
<0 >>
Id
to (•*
IH 01
CO Ck
cfi
u
H
^
p.
^
U
O
2
ON ON ON O> ON
ON ON ON ON ON
55555
Ct a so ci NO
NO O •* O ^
o o o o o
o o o o o
in PO o en o
— CM m CN en
O co oo flo oo
CM O O NO NO
en -3- •» ON ON
•^ ••« «4
ON h« in r* in
in •*< en »M en
— en NO en ve
s e
** 3 ftl 9 O
•H ^ 60 M 60
§•0 1- -O k-
s
U H e
CO
>s
hi
Ml
a
a
>N
CO
•o
60
C
u
eg
01
O
O
in
en
e
0
•o
Cl
CO
(0
ca
eg
C-10
-------
cn
H
^
,J
PH
rJ
g
ta*
X
OS
W
H
r4
cn
P
w
rJ
{H
PS
<
Q
a
0
o
w
cn
fn
o
CO
O
H^
H
cn
rH
Pi
W
H
CJ
2
<:
oj
^
M
&
H
T-TJ
r*1*
O
a
<:
e>
a
M
H
<
Pi
W
ft.
P^
O
•
< CU
< < a
JE H ""^
H ^
cn co
a o M
ca cj co
Cri rH
0 rH
hJ B5 O
««! H -a
P H
S2 < 0
sg « o
< W 0
PH vx
O
/^•s
rJ CO
5 r-
H CO
H rH
P-j H rH
< cn o
0 O T3
o
a o
pa o
W 0
\_^
/-\
^
>-i
V (1)
a w a
M CJ
J M CO
td Pi M
cn p-i to
00 v£>
1-4 CO
E
rH 3 4)
rH 'H 60
CO -O M
S
-------
w
z
M
O
C/l
C/J
M
Pi
W
z
<
z
M
Pw
<
O
OS
w
w
I
J
<
H
H-f /•"•*
Q eu 4-1
w < c
H CJ 01
pc o
e> ft, M
M O CD
M 0-
f^ r . -^.
»£ tr* ^— '
00
o
c_>
/— \
hJ Cd 4J
<: H c
•?• -i
Pd X OJ
< < p.
a H -^
H ^
W CO
u o n
PQ e> co
W r-H
O rH
h4 Z 0
<; M -o
» H
Z < 0
z oa o
-i
CO Pu CO
<3 rH
W iH
O
-o
^~s
J H
~'
^J O
iJ H
W Z
0 <
i£
tn
•
0
f-H
r~.
-a-
CM
•
m
in
CM
sr
ro
o
•
i—4
o
o
o
o
m
f^4
C-12
-------
- Glass: "Glass Manufacturing Plants Background Information:
Proposed Standards of Performance," June 1979,
EPA-450/3-79-005a.
Product prices for each model plant are summarized in Tables 1
through 5.
D. Discount Rates
For the purpose of this analysis the discount rates used to make the
calculations were the model plants' weighted average cost of capital.
Weighted average cost of capital is calculated using the formula
k = k.(l-t)(W ) + k (W )
ad d e e
where,
k = weighted average cost of capital
k, = before tax cost of debt
t = marginal income tax rate
W, = weight of debt in total capital structure
k = cost of equity
W » weight of equity in total capital structure
Different weights of debt and equity were assigned to model plants,
depending on the industry. A before-tax debt cost of 9 percent was used for
the metals and glass model plants. A higher before-tax debt cost of
13 percent was used for the cotton gin model plants because of their lack of
access to the bond market and their consequent reliance on local commercial
banks. Equity costs varied on an after-tax basis between 10.4 and
14.0 percent among the different industries. Weighted average costs of
capital are summarized in the model plant data in Tables 1 through 5.
C-13
-------
TECHNICAL REPORT DATA
(flease read Instructions on the reverse before completing)
1. REPORT NO.
450/5-82-005
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Preliminary Study of Sources of Inorganic Arsenic
5. REPORT DATE
August 1982
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Miles, A. J., G. W. Brooks, and L. E. Keller
Radian Corporation
8. PERFORMING ORGANIZATION REPORT NO.
DCN 82-240-016-18-12
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
3024 Pickett Road
Durham, N.C. 27705
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3058
Task 18
12. SPONSORING AGENCY NAME AND ADDRESS
Pollutant Assessment Branch
OAQPS
U. S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
FINAL
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The purpose of this study was to analyze the severity of inorganic arsenic
emissions from seven source categories including primary copper, lead, and zinc
smelting, secondary lead smelting, cotton ginning, glass manufacturing, and arsenic
chemical manufacturing. The magnitude of arsenic emissions from each source category
were quantified and control systems were investigated to determine baseline and
estimated best control (EBC) for arsenic. The environmental, energy, and economic
costs of implementing EBC, in source categories where it is not being used, were
estimated. Two source categories, primary lead smelting and arsenic chemical
manufacturing, were determined to have EBC in place as a result of compliance with
existing regulations. The number of people exposed to arsenic from each of the source
categories will be assessed separately by EPA using emissions and stack data generated
in this study. The physical and chemical characteristics of arsenic and their impact
on the control of arsenic emissions from the source categories were also examined.
U.S. environmental Protection Agency
R«Slon V, library f
230 South Dearborn Street
n IHinnio
£04
17.
KEYWORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lOENTIFIGHS/OPEN ENDED TERMS
c. COSATI Field/Group
Arsenic Cotton Gins
Air Pollution Control Glass Manufacturing
Hazardous Air Pollutant Arsenic Chemical
Copper Smelting Manufacturing
Lead Smelting
Zinc Smelting
Secondary Lead Smelting
Hazardous Air Pollution
Arsenic
is. OISTRIBUT.QN STATEMENT
Unlimited
19 SECURITY CLASS /Tins Report)
Unclassified
21.
PAGES
20 SECURITY CLASS /Tins page I
Unclassified
22. PRICE
EPA Form 2220-1 (R«». 4-77) &S5VIOU5 EDITIO>. is OBSOLETE
------- |