CONTROL TECHNIQUES
FOR MERCURY EMISSIONS
FROM EXTRACTION AND
CHLOR-ALKALI PLANTS
U. S. ENVIRONMENTAL PROTECTION AGENCY
-------
CONTROL TECHNIQUES FOR MERCURY EMISSIONS
FROM EXTRACTION AND CHLOR-ALKALI PLANTS
ENVIRONMENTAL PROTECTION AGENCY
Office of -Air and Water Programs
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
February 1973
-------
The AP series of reports is published by the Technical Publications Branch of the Information
Services Division of the Office of Administration for the Office of Air and Water Programs,
Environmental Protection Agency, to report the results of scientific and engineering studies, and
information of general interest in the field of air pollution. Information reported in this series
includes coverage of intramural activities and of cooperative studies conducted in conjunction
with state and local agencies, research institutes, and industrial organizations. Copies of AP
reports are available free of charge to Federal employees, current contractors and grantees, and
nonprofit organizations - as supplies permit - from the Air Pollution Technical Information
Center, Environmental Protection Agency, Research Triangle Park, North Carolina 27711 or
from the Superintendent of Documents.
Publication No. AP-118
For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402
-------
PREFACE
This document contains information about the nature and control of a hazardous air pollutant
— mercury. The primary purpose of this document is to provide information useful to those
involved in the control of emissions of mercury from industrial sources. The language and
approach are largely technical, but the first two sections should be of interest and value to the
general reader.
The requirement to publish this document was established when the Administrator of the
Environmental Protection Agency listed mercury as a hazardous air pollutant by notice in the
Federal Register (Vol. 36, page 5931) on March 21, 1971. The Administrator acted under the
authority granted him by Section 112 of the Clean Air Act, which defines a hazardous air
pollutant as, ".. .an air pollutant to which no ambient air quality standard is applicable and
which in the judgment of the Administrator may cause, or contribute to, an increase in mortality
or an increase in serious irreversible, or incapacitating reversible, illness."
Messrs. S. L. Roy, Jr., G. S. Thompson, Jr., F M. Alpiser, and T. R. Osag of the Office of Air
and Water Programs, Environmental Protection Agency, were primarily responsible for compiling
the information contained in this document. This information represents the efforts of the
Environmental Protection Agency, as well as the advice of the members of the advisory com-
mittees listed on the following pages and the contributions of many individuals associated with
other Federal agencies, State and local governments, and private businesses.
111
-------
NATIONAL AIR POLLUTION CONTROL TECHNIQUES
ADVISORY COMMITTEE
Chairman
Mr. Donald F Walters
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, N.C. 27711
Mr. Raynal W. Andrews
150 Guyasuta Road
Pittsburgh, Pennsylvania 15215
Mr. Robert L. Chass
Air Pollution Control Officer
Los Angeles County Air
Pollution Control District
434 South San Pedro Street
Los Angeles, California 90013
Mr. Charles M. Copley, Jr.
Commissioner, Division of Air
Pollution Control
City of St. Louis
Room 419 City Hall
St. Louis, Missouri 63103
Mr. C. G. Cortelyou
Coordinator of Air and Water
Conservation
Mobil Oil Corporation
150 E. 42nd Street - Room 1650
New York, N.Y. 10017
Mr. Arthur R. Dammkoehler
Air Pollution Control Officer
Puget Sound Air Pollution
Control Agency
410 W.Harrison Street
Seattle, Washington 98119
Dr. Aaron J. Teller
Teller Environmental Systems, Inc.
295 Fifth Avenue
New York, N.Y. 10016
Mr. William W. Moore
President, Belco Pollution Control Corp.
100 Pennsylvania Avenue
Paterson, New Jersey 07509
Mr. William Munroe
Chief, Bureau of Air Pollution Control
State of New Jersey
Dept. of Environmental Protection
P.O.Box 1390
Trenton, New Jersey 08625
Mr. Vincent D. Patton
Executive Director
State of Florida Air and Water
Pollution Control
315 S. Calhoun Street
Tallahassee, Florida 32301
Dr. Robert W. Scott
Coordinator for Conservation Technology
Esso Research and Engineering Co.
P.O. Box 215
Linden, New Jersey 07036
Dr. R. S. Sholtes
University of Florida
Environmental Engineering Department
College of Engineering
Gainesville, Florida 32001
Mr. W. M. Smith
Director, Environmental Control
National Steel Corporation
Box 431, Room 159, General Office
Weirton, West Virginia 26062
-------
Mr. George P. Ferreri
Chief, Division of Compliance
Bureau of Air Quality Control
Maryland State Department of
Health and Mental Hygiene
61 ON. Ho ward Street
Baltimore, Maryland 21201
Mr. Benjamin F. Wake
Director, Division of Air Pollution
Control and Industrial Hygiene
Montana State Department of Health
Helena, Montana 59601
Mr. Charles M. Heinen
Executive Engineer
Materials Engineering
Chrysler Corporation
Box 1118,Dept. 5000
Highland Park, Michigan 48231
Mr. A. J. von Frank
Director, Air and Water
Pollution Control
Allied Chemical Corporation
P.O. Box 70
Morristown, New Jersey 07960
VI
-------
FEDERAL AGENCY LIAISON COMMITTEE
Chairman
Mr. Donald F. Walters
Office of Air and Water Programs
Office of Air Quality Planning and Standards
Environmental Protection Agency
Research Triangle Park, N.C. 27711
FEDERAL POWER COMMISSION
Mr. T. A. Philips
Chief, Bureau of Power
Federal Power Commission, Room 3011
411 G Street, N.W.
Washington, D.C. 20426
GENERAL SERVICES ADMINISTRATION
Mr. Harold J. Pavel
Director, Repair and Improvement Division
Public Building Service
General Services Administration
9th and D Streets, S.W.
Washington, D.C.
NATIONAL AERONAUTICS AND
SPACE ADMINISTRATION
Mr. Ralph E. Cushman
Special Assistant
Office of Administration
National Aeronautics and Space Administration
Washington, D.C. 20546
NATIONAL SCIENCE FOUNDATION
Dr. O. W. Adams
Program Director for Structural Chemistry
Division of Mathematical and Physical Sciences
National Science Foundation
1800 G Street, N.W.
Washington, D.C. 20550
POSTAL SERVICE
Mr. Robert Powell
Assistant Program Manager
U.S. Postal Service
Room 4419
1100 L Street
Washington, D.C. 20260
DEPARTMENT OF TRANSPORTATION
Dr. Richard L. Strombotne
Office of the Assistant Secretary
for Systems Development and Technology
Department of Transportation
400 7th Street, S.W.
Washington, D.C. 20591
DEPARTMENT OF DEFENSE
Harvey A. Falk, Jr., Commander, USN
Office of the Assistant Secretary
of Defense
Washington, D.C. 20301
DEPARTMENT OF HOUSING AND
URBAN DEVELOPMENT
Mr. Samuel C. Jackson
Assistant Secretary for Metropolitan Development
Department of Housing and Urban Development
Room 7100
7th and D Streets, S.W.
Washington, D.C. 20410
vn
-------
DEPARTMENT OF AGRICULTURE
TENNESSEE VALLEY AUTHORITY
Dr. Theodore C. Byerly
Assistant Director of Science and Education
Office of the Secretary
U.S. Department of Agriculture
Washington, D.C. 20250
DEPARTMENT OF COMMERCE
Dr. James R. McNesby
Room A361, Materials Building.
National Bureau of Standards
Washington, D.C. 20234
DEPARTMENT OF THE TREASURY
Mr. Gerard M. Brannon
Director, Office of Tax Analysis
Room 4217 MT
Department of the Treasury
15th and Pennsylvania Avenue, N.W.
Washington, D.C. 20220
DEPARTMENT OF THE INTERIOR
Dr. LeRoy R. Furlong
Research Advisor to the Assistant Secretary
Office of Assistant Secretary — Mineral
Resources
Bureau of Mines
Interior Building
Washington, D.C. 20240
DEPARTMENT OF HEALTH, EDUCATION,
AND WELFARE
Dr. Douglas L. Smith
Department of Health, Education, and Welfare
National Institute of Occupational Health
Rockville, Maryland
Dr. F. E. Gartrell
Director of Environmental Research and Development
Tennessee Valley Authority
715 Edney Building
Chattanooga, Tennessee 37401
ATOMIC ENERGY COMMISSION
Dr. Martin B. Biles
Director, Division of Operational Safety
U.S. Atomic Energy Commission
Washington, D.C. 20545
VETERANS ADMINISTRATION
Mr. Gerald M. Hollander, P.E.
Director of Architecture and Engineering
Office of Construction
Veterans Administration
Room 619 Lafayette Building
811 Vermont Avenue, N.W.
Washington, D.C. 20420
DEPARTMENT OF JUSTICE
Mr. Walter Kiechel, Jr.
Land and Natural Resources Division
Department of Justice
Room 2139
10th and Constitution Avenue, N.W.
Washington, D.C. 20530
DEPARTMENT OF LABOR
Mr. Robert D. Gidel
Deputy Director, Bureau of Labor Standards
Department of Labor
Room 401, Railway Labor Building
400 1st Street, N.W.
Washington, D.C. 20210
Vlll
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES xi
LIST OF TABLES xii
ABSTRACT xiii
SUMMARY xv
1. INTRODUCTION 1-1
2. BACKGROUND INFORMATION 2-1
2.1 DEFINITIONS 2-1
2.1.1 Primary Mercury Extraction 2-1
2.1.2 Chlor-Alkali Plants 2-1
2.2 PROPERTIES OF MERCURY 2-2
2.3 ORIGIN, PRODUCTION, AND USES OF MERCURY 2-2
2.3.1 World 2-2
2.3.2 United States 2-3
2.4 EMISSION MECHANISMS AND GENERALIZED CONTROL PRINCIPLES . . 2-3
2.4.1 Process Gases 2-3
2.4.2 Ventilation Air 2-5
2.4.3 Participate Emissions 2-8
2.5 REFERENCES FOR SECTION 2 2-8
3. MERCURY EMISSIONS, CONTROL TECHNIQUES, AND CONTROL COSTS .... 3-1
3.1 PRIMARY MERCURY PRODUCTION 3-1
3.1.1 Emissions and Process Description 3-1
3.1.2 Emission Reduction Resulting from Process Changes 3-9
3.1.3 Control Techniques 3-11
3.1.4 Control Costs 3-15
3.1.5 Development of New Technology 3-17
IX
-------
3.2 CHLOR-ALKALI PRODUCTION .................... 3-19
3.2.1 History ............................ 3-19
3.2.2 Emissions and Process Description ................ 3-20
3.2.3 Emission Reduction Resulting from Changing to
Diaphragm-Cell Process ..................... 3-24
3.2.4 Control Techniques ....................... 3-24
3.2.5 Control Costs ......................... 3-31
3.2.6 Development of New Technology ................. 3-33
3.3 REFERENCES FOR SECTION 3 .................... 3-33
APPENDIX .................................. A-l
A.1 CALCULATION OF EQUILIBRIUM CONCENTRATION OF MERCURY
VAPOR IN A GAS AND RESULTANT LOSSES IN PROCESS
STREAMS OR IN VENTILATION AIR ................. A-l
A.2 METHODS OF ESTIMATING CONTROL COSTS ............. A-2
A.2.1 Equipment Costs ....................... A-2
A.2. 2 Fixed-Capital Requirement ................... A-3
A.2. 3 Annual Operating Costs ..................... A4
A.3 CONTROL EQUIPMENT COSTS AND EFFICIENCIES
A.4 GEOGRAPHIC LOCATION OF CHLOR-ALKALI PLANTS
THAT USE MERCURY CELLS .................... A4
A.5 REFERENCES FOR APPENDIX .................... A-4
SUBJECT INDEX ............................... M
-------
LIST OF FIGURES
Figure Page
2-1 Mercury Deposits In North America 2-4
2-2 Trends in U.S. Production, Consumption, and Price of Mercury 2-7
3-1 Pyrometallurgical Process for Producing Mercury 3-1
3-2 Multiple-Hearth Furnace 3-5
3-3 Flotation Flow Sheet for Cinnabar Ore 3-6
3-4 Group Horizontal Pipe Retort 3-8
3-5 Two-Pipe, Inclined Pipe Retort 3-9
3-6 D-Retort 3-10
3-7 Mercury Mist Eliminator 3-13
3-8 Manufacturer's Estimate of Purchase Cost of a Fiber Pad Type
of Mist Eliminator . . .... .3-16
3-9 Conceptual Hydrometallurgical Plant Layout and Flow Diagram
Based on Laboratory Data 3-18
3-10 Percentage of Total Installed U.S. Chlorine Capacity for
Diaphragm and Mercury Cells 3-20
3-11 Installed U.S. Chlorine Capacity in Diaphragm and Mercury Cells 3-20
3-12 Basic Flow Diagram for Chlor-Alkali Mercury-Cell Operation 3-21
3-13 Mercury Emissions in Cell Room Ventilation Air 3-23
3-14 Cooling and Condensing of Hydrogen Stream 3-25
3-15 Converging-Diverging Nozzle Mist Eliminator 3-26
3-16 Cooling, Condensing, and Mist Elimination 3-27
3-17 Depleted Brine Scrubbing System 3-27
3-18 Hypochlorite Scrubbing System 3-28
3-19 Activated Carbon Bed System 3-29
3-20 Process Flow Sheet for a Two-Bed Molecular Sieve System 3-30
A-l Equilibrium Concentration of Mercury Vapor in Air as a
Function of Temperature A-l
A-2 Mercury Emission Rate as a Function of End-Box Ventilation
Air Flow Rate at Arbitrary Conditions of Air Temperature
and Percentage Saturation A-3
A-3 Location of Mercury-Cell Chlor-Alkali Plants in the United States A-11
XI
-------
LIST OF TABLES
Table
2-1 Isotopic Abundance of Mercury ...................... 2-2
2-2 Physical Properties of Elemental Mercury .................. 2-2
2-3 Vapor Pressure of Mercury ........................ 2-2
2-4 World Production of Mercury by Country for 1968 through 1970 ........ 2-3
2-5 United States Mercury Statistics ...................... 24
2-6 Mercury Produced in the United States ................... 2-5
2-7 Mercury Consumed in the United States ................... 2-6
3-1 Calculated Volumetric Flow Rates due to Combustion and Excess
Air fora 100-ton/day Ore Rate ...................... 3-3
3-2 Calculated Mercury Emissions from the Condenser Stack of a
Primary Mercury Extraction Facility ................... 3-3
3-3 Calculated Vaporous Mercury Emissions for Selected Condenser
Stack Flow Rates and Temperatures .................... 3-12
3-4 Summary of Primary Mercury Control Technique Costs and Expected
Emissions in Condenser Stack Gas Stream ................. 3-17
3-5 Chlorine Capacity and Production Methods in the United States ......... 3-19
3-6 Summary of Chlor-Alkali Control Costs and Expected Emissions
for Combined Hydrogen and End-Box Ventilation Streams .......... 3-34
A-l Mercury Vapor Losses in a Hydrogen Stream ................. A-2
A-2 Calculated Mercury Vapor Emission Rates in Cell Room Ventilation Air ..... A-2
A-3 Estimated Collection Efficiencies of Control Equipment ............ A-6
A-4 Equipment Costs for Primary Mercury Extraction Facilities ........... A-6
A-5 System Costs for Primary Mercury Extraction Facilities ............ A-6
A-6 Equipment Costs for Mercury-Cell Chlor-Alkali Plants ............. A-6
A-7 Control System Costs for Chlor-Alkali Plants Treating Hydrogen
and End-Box Ventilation Streams ..................... A-7
A-8 Control System Costs and Emission Rates for Combined Hydrogen
and End-Box Ventilation Streams ..................... A-8
A-9 Capital Investment Required to Convert from Mercury to Diaphragm
Cells ................................. A-9
A- 10 Mercury-Cell Chlor-Alkali Plants in the United States ............ A- 10
Xll
-------
ABSTRACT
The toxicity of mercury, combined with its high volatility, creates a potential health hazard.
This publication deals with two sources of mercury emissions, the primary mercury processing
industry and the mercury-cell chlor-alkali industry. An effort is made (1) to identify the process
steps that may produce atmospheric mercury emissions, (2) to summarize the emission control
techniques and low mercury emission processes used or applicable to these industries, and (3) to
evaluate these techniques in terms of cost and effectiveness.
The condenser gas stream is the major source of mercury emissions from a primary mercury
processing plant. The amount of emissions can be reduced by converting to processes that
inherently produce fewer emissions or by treating effluent gases to remove mercury. Process
changes that inherently produce fewer emissions include beneficiation of ore, retort processing,
and hydrometallurgical processing. Appropriate control techniques include cooling and mist
elimination, wet scrubbing, or adsorption beds.
Major emissions of mercury from a chlor-alkali plant using mercury cells are from the
hydrogen gas stream, the end-box ventilation stream, and the cell room ventilation air. The
emissions from all sources can be eliminated by converting to the diaphragm-cell process. The
cost of converting a 100-ton-per-day plant is estimated to range from $3,700,000 to $8,000,000.
Mercury emissions can also be reduced by the installation of control systems and the use of
good housekeeping practices. The hydrogen gas and the end-box ventilation air streams can be
treated by cooling and mist elimination, chemical scrubbing, or adsorption beds. No techniques
are presently available to treat the cell room ventilation air; therefore, the control of mercury
emissions from this source is dependent on good housekeeping practices.
A control system for a primary mercury facility using cooling (down to 45° to 55° F) and
mist elimination would cost between $86,000 and $108,000, depending upon the type of mist
elimination device used. The cost of a similar control system for a chlor-alkali plant is estimated
at $202,000. Chemical scrubbing, which is too expensive for existing primary mercury facilities,
can be applied to the chlor-alkali process at a cost from $160,000 to $350,000 for a 100-ton-
of-chlorine-per-day plant.
The cost of a carbon bed adsorption system for a primary mercury facility is estimated at
$66,000. The capital investment for an adsorption bed system for a chlor-alkali plant of 100-
tons-per-day capacity would range from $279,000 to $349,000.
Key words: mercury emissions, chlor-alkali plants, control techniques, costs.
xni
-------
SUMMARY
BACKGROUND INFORMATION
Mercury, one of the most volatile of all
metals, vaporizes readily at room temper-
ature. Accordingly, the exposure of a mercury
surface to a gas will result in vaporous
mercury emissions. This characteristic, com-
bined with mercury's toxicity, may create a
potential health hazard in some instances. The
saturation concentration of mercury vapor in
a gas can be estimated from vapor pressure
data and the ideal gas law and used in
estimating the vaporous emissions from gas
streams containing mercury.
The most important deposits of mercury
are found in Italy, Spain, the U.S.S.R.,
Yugoslavia, China, the United States, Canada,
Mexico, and the Philippines. The major
deposits of mercury in the United States
occur in the West with California and Nevada
being the major mercury-producing states.
The United States ranked fifth in 1970 and
fourth in 1967 through 1969 in the world
production of mercury.
MERCURY EMISSIONS, CONTROL
TECHNIQUES, AND COSTS
Mercury emissions from process gas
streams are dependent on the volumetric flow
rate and the concentration of mercury in the
gas stream; therefore, appropriate control
techniques include cooling the gas, reducing
the mercury vapor concentration by chemical
scrubbing or adsorption, and reducing the
volumetric flow rate of the gas stream.
Mercury emissions can also occur when
mercury is exposed to circulating ventilation
air. The resultant mercury concentrations are
low, but large emissions of mercury can occur
where large volumes of ventilation air are
used.
Primary Mercury Processing
Mercury can be extracted from ore by
two basic methods, pyrometallurgical pro-
cessing and hydrometallurgical processing.
The former process employs heat to volatilize
mercury from the ore, whereas the latter
method uses an aqueous chemical solution to
extract the mercury from the ore.
In the United States, most of the primary
mercury is produced through a directly
heated pyrometallurgical process. Mercury ore
is directly heated by combustion gases in a
multiple-hearth or rotary furnace to volatilize
mercury. The hot mercury-laden gases leaving
the furnace are cooled in an air-cooled heat
exchanger, where mercury vapor condenses
and collects under an aqueous layer. The
condenser gas stream is then emitted to the
atmosphere through a stack, which is the
major mercury emission point in the process.
A typical emission from condenser stack gases
is approximately 25 pounds of mercury per
day for an existing 100 ton-of-ore-per-day
direct-heat extraction process in which the
exit gas temperature is 110° F.
In another application of a pyrometal-
lurgical process, mercury ore is heated in-
directly by combustion gases to volatilize
mercury in a retort. This batch process is
generally used in small operations to process
high-grade or concentrated ore. The mercury
vapor that is evolved is condensed in a water-
or air-cooled heat exchanger and collected
under an aqueous layer. The major mercury
emission from an indirectly heated retort
probably occurs while the retort is being
charged and discharged. The condenser stack
gas flow rate is low as a result of the indirect
heating method; therefore, the stack emis-
sions from a retort process are low.
xv
-------
Mercury can be leached from mercury ore
with an aqueous solution of sodium sulflde or
sodium sulfite-sodium hydroxide, or by elec-
tro-oxidation in an aqueous sodium chloride
solution. The soluble mercury compounds
that are formed can be precipitated with
metals such as zinc, iron, and aluminum, or
electro-deposited to recover the mercury.
Atmospheric mercury emissions from these
hydrometallurgical processes are negligible. A
potential water pollution problem exists,
however, because soluble mercury compounds
may be emitted with the tailings and leach
solution. Currently, there are no commercial
hydrometallurgical processes in operation in
the United States, but the process has been
investigated on the research and development
level.
The primary mercury industry currently
employs little if any control technology to
reduce mercury emissions from the condenser
stack gases. As a result, few of the techniques
that are discussed as control methods have
actually been used in this industry. Cooling,
mist elimination, water scrubbing, and adsorp-
tion are control techniques that have been
used successfully in reducing mercury emis-
sions from similar gas streams and should be
applicable in reducing emissions from primary
mercury extraction plant effluents.
Chlor-Alkali Processing
Chlorine gas and alkali metal hydroxide
can be produced in electrolytic cells by the
diaphragm-cell process or the mercury-cell
process. In the diaphragm-cell process, an
asbestos diaphragm separates the anode from
the cathode. Chlorine gas is formed at the
anode and hydrogen gas and caustic are
formed at the cathode by the electrolytic
decomposition of a salt solution. Approxi-
mately 72 percent of the domestic chlorine is
produced by this process. One disadvantage of
this process, compared with the mercury-cell
process, is that the diaphragm cell produces a
low-grade caustic that must be concentrated
and purified. This process, however, produces
no mercury emissions.
Mercury is used as a flowing cathode in
the mercury-cell process. The mercury elec-
trolytic cell is composed of the electrolyzer
and the decomposer. In the electrolyzer
section, a salt solution, usually NaCl, flows
cocurrently with the mercury cathode. A high
current density is applied between the mer-
cury cathode and the carbon or metal anodes.
Chlorine gas forms at the anode and alkali
amalgam forms at the cathode. The amalgam
is separated from the brine in a discharge
end-box and enters the decomposer section,
where water is added. In the decomposer, the
amalgam becomes the anode to a short-
circuited graphite cathode. Hydrogen gas and
alkali metal hydroxide are formed in the
decomposer, and the amalgam is converted
back to mercury. The mercury is then re-
cycled to the inlet end-box, where it reenters
the electrolyzer. The major emissions of
mercury from this process occur with the
hydrogen gas, the end-box ventilation system,
and the cell room ventilation air.
After leaving the decomposer, the hydro-
gen gas stream contains substantial mercury
and water vapor. This stream is usually cooled
in one or two stages. The mercury that is
condensed from this cooling operation is
often removed by a mist elimination device
before the hydrogen is vented to the atmo-
sphere or burned as fuel. In some cases, the
hydrogen is subjected to additional treatment
such as passage through an adsorption bed or
scrubbing device.
The end-box ventilation air may consist
of air ventilated from the end-boxes, the
mercury pump sumps, and the water col-
lection tank. This gas stream may or may not
be saturated with mercury vapor, depending
on the individual plant. The stream is nor-
mally cooled with ambient water before it is
emitted to the atmosphere. Additional treat-
ment of the end-box ventilation system con-
sists of methods similar to those applied to
the hydrogen stream.
The cell room is ventilated in order to
cool the cell room and also to dilute the
concentration of mercury to protect the
xvi
-------
health of cell room workers. For a plant
producing 100 tons of chlorine per day, the
ventilation air flow rate for the cell room may
vary from 100,000 to 1,000,000 cubic feet
per minute. Although a low cell room mer-
cury concentration is usually maintained (50
to 100 micrograms per cubic meter), the high
volumetric flow rate can cause the cell room
ventilation air to be a major source of
mercury emissions from a chlor-alkali plant.
No applicable control techniques are cur-
rently available for the reduction of mercury
emissions from low-concentration, high-
volume gas streams. Meticulous housekeeping
and maintenance procedures should be em-
ployed to minimize the amount of mercury
vapor entering the cell room atmosphere.
Emission Control Techniques
Mercury emissions can be reduced by
treating effluent gases to remove mercury or
by converting to processes that inherently
produce fewer emissions. Conversion from
direct to indirect heating of ore, beneficiation
of ore, and conversion from pyrometallurgical
to hydrometallurgical processing are methods
that can reduce atmospheric mercury emis-
sions from primary mercury production facil-
ities. The conversion of a mercury-cell process
to a diaphragm-cell process will eliminate the
mercury emissions from chlor-alkali pro-
duction plants.
The cooling of a gas stream containing
mercury vapor below its mercury saturation
temperature will cause mercury vapor to
condense, thereby reducing the concentration
of mercury vapor in the gas. Some of the
condensed mercury remains in the gas stream
as a fine mist and can be removed with a mist
elimination device or a water scrubber to
increase the efficiency of the cooling tech-
nique.
Mercury vapor can be removed from a gas
stream by adsorption and chemical scrubbing
techniques. Molecular sieve and activated-
carbon beds can be used to adsorb mercury
vapor from a stream. Mercury vapor can also
be removed chemically by scrubbing the gas
with hypochlorite, depleted brine, or hot
concentrated sulfuric acid solutions.
Costs
Costs are presented for those control
systems applicable to both primary mercury
processing facilities and chlor-alkali plants.
The costs of control systems given for pri-
mary mercury facilities are for a processing
operation having a capacity of 100 tons of ore
per day. Control system costs for chlor-alkali
plants are based on the control of emissions
from both the hydrogen stream and the
end-box ventilation stream for a plant having
a capacity of 100 tons of chlorine per day.
A control system for a primary mercury
facility using cooling to 45° to 55°F and mist
elimination would cost between $86,000 and
$108,000, depending upon the type of mist
elimination device used. The cost of a similar
control system for a chlor-alkali plant is
estimated to be approximately $202,000.
Chemical scrubbing, preceded by at least
partial cooling and mist elimination, is a
control system applicable to both source
categories. Because of the expense, however,
this type of control would not be feasible for
existing primary mercury facilities. Costs for a
chlor-alkali plant having a capacity of 100
tons of chlorine per day are estimated to
range from $160,000 to $350,000.
A third class of control system uses
adsorption beds of either treated activated
carbon or a molecular sieve/adsorbent blend
and is applicable to both primary mercury
facilities and chlor-alkali plants. The cost of a
carbon bed system for a 100-ton-per-day
primary mercury facility is estimated to be
$66,000. A molecular sieve system would not
be feasible for the control of emissions from
this source because of the expense. The
capital investment for an adsorption bed
system for a chlor-alkali plant of 100-ton-per-
day capacity would range from $279,000 to
$349,000.
xvii
-------
CONTROL TECHNIQUES FOR MERCURY EMISSIONS
FROM EXTRACTION AND CHLOR-ALKALI PLANTS
1. INTRODUCTION
The hazardous nature of mercury has
been known for centuries. Mercury is hazard-
ous to public health chiefly because of its
interference with the central nervous system.
In addition, mercury is volatile, so that
mercury in the ambient air may approach
hazardous levels.
The scope of this document is specifically
limited to consideration of mercury emissions
and control techniques for those emissions in
the primary mercury processing industry and
in the mercury-cell chlor-alkali industry.
The objectives of this report are (1) to
present, in a concise manner, the essential
details of the operation of each industry; (2)
to identify the process steps that may pro-
duce atmospheric mercury emissions; (3) to
summarize the emission control techniques
and low-mercury-emission processes used in
or applicable to these industries; and (4) to
evaluate these techniques in terms of cost and
effectiveness.
The information used in the preparation
of this document was obtained from four
sources. An extensive review of the literature
was conducted with emphasis placed on re-
cent publications and reports. This material
was evaluated and considered together with
comments from management and trade-
association personnel and information devel-
oped by contractors. Plant visits provided
insight into current emission control practices
and applicable control techniques and pro-
cesses. Finally, a limited amount of source
testing was performed to determine plant
emission levels and to evaluate the effective-
ness of control techniques.
The information contained in this doc-
ument is intended to give an appraisal of the
emission control methods, techniques, and
processes currently being used in or potential-
ly adaptable to primary mercury extraction
plants and mercury-cell chlor-alkali plants.
1-1
-------
2. BACKGROUND INFORMATION
2.1 DEFINITIONS
2.1.1 Primary Mercury Extraction
Beneficiation — Methods by which low-grade
ore is processed into higher-grade ore or
concentrate.
Cinnabar — Mercury-bearing (in the form of
mercuric sulfide, HgS) mineral that is the
primary source of the metal.
Condenser stack gas — Vapor and particulate
matter emitted from the condenser stack
after having passed through a condenser
system.
Directly heated furnace — Furnace in which
ore is heated directly by hot combustion
gases.
Hoeing operation — Process whereby mer-
cury, dust, and soot collected in the
condenser system of a primary plant are
mixed with lime and alternately scraped
and pressed by mechanical or manual
means to cause coalescing of minute
mercury droplets.
Hydrometallurgical process — Procedure that
uses a water solution of various chemicals
to extract metal from its ore.
Indirectly heated furnace — Furnace in which
ore is heated indirectly by combustion
gases; that is, the combustion gases are
never in contact with the ore.
Particulate matter — Any material, except
uncombined water, that exists as a solid
or liquid in the atmosphere or in a gas
stream.
Primary mercury production — Production of
mercury metal from mercury-bearing ore.
Pyrometallurgical process — Procedure that
uses heat to extract metal from its ore.
Retort — Type of indirectly heated furnace.
2.1.2 Chlor-Alkali Plants
Amalgam — Alloy of mercury and another
metal, such as sodium or potassium.
Anode — Positive pole of an electrolytic cell.
Cathode - Negative pole of an electrolytic
cell.
Contact cooler — Tower in which a liquid is
used for direct contact cooling of a gas
stream.
Denuder or decomposer — Device into which
the sodium or potassium amalgam from
the electrolyzer flows continuously. With-
in this device, deionized water or other
chemicals are added, and the amalgam
becomes the anode to a short-circuited
iron or graphite cathode in an electrolyte
of sodium or potassium hydroxide solu-
tion. Hydrogen gas is formed and the
sodium or potassium hydroxide is in-
creased to a 50 percent solution. Regener-
ation of the mercury occurs simultane-
ously, and the mercury is recycled back
to the chlorine cell for continuous usage.
Diaphragm cell — Electrolytic device em-
ploying porous asbestos coating over the
cathode screen that separates the chlorine
gas evolved at the anode from the hydro-
gen gas evolved at the cathode.
Mercury-cell chlor-alkali electrolyzer — Elec-
trolytic device that uses a flowing mer-
cury cathode to make chlorine gas and
sodium or potassium amalgam.
Mercury knockout drum — Device with a
tangential top inlet for process gases
containing mercury vapor, mercury mist,
and water. As the gas spirals downward,
centrifugal forces separate some of the
mercury mist and water droplets from the
process gas stream. Mercury collects at
the bottom under a layer of water while
cleaned gas flows out of a pipe in the top
of the drum.
2-1
-------
Mercury mist eliminator — Device that uses
direct impingement or high centrifugal
forces for the removal of mercury mist
from process gas streams.
2.2 PROPERTIES OF MERCURY
Mercury is found in nature in many
forms, but the principal mercury ore is
cinnabar. Cinnabar in the pure form is red,
has a specific gravity of 8.1, and contains 86.2
percent mercury by weight. In the United
States, the average mined ore contains ap-
proximately 5 pounds of mercury per ton of
ore.
Mercury is the only common metal that
exists as a liquid at ordinary temperatures.1 It
is a heavy (molecular weight = 200.59),
silver-white, shining metal at normal temper-
atures but tarnishes at elevated temperatures
near its boiling point because of the forma-
tion of its oxide, HgO. Mercury combines
with many metals to form alloys called
amalgams. Mercury also is a fair conductor of
electricity and has a regular coefficient of
expansion. The latter properties make it
generally useful in thermometers, barometers,
and other instruments.
Some of the properties of elemental
mercury are summarized in Tables 2-1 and
2-2. Mercury is a unique metal because of its
high vapor pressure. This characteristic, in
combination with its toxicity, may create a
potential health hazard in some instances. The
vapor pressure of mercury for several temper-
atures is given in Table 2-3.
Table 2-1. ISOTOPIC ABUNDANCE OF
MERCURY1
Table 2-2. PHYSICAL PROPERTIES OF
ELEMENTAL MERCURY
1
Isotope, mol wt
Abundance, percent
204
202
201
200
199
198
196
6.8
29.8
13.2
23.1
16.8
10.0
0.15
Property
Avg mol wt
Density
Surface tension
Heat of fusion
Heat of vaporization
Melting point
Boiling point
Value and units
200
13.5955 g/ml at 0°C
484 dynes/cm vac 20°C
2.7 cal/g
13.985 kcal/gmole
-38.9°C
356.6° C
Table 2-3. VAPOR PRESSURE OF
MERCURY1
Temperature, °C Vapor pressure, mm Hg
-10
0
20
40
60
80
100
0.0000606
0.000185
0.001201
0.006079
0.02524
0.08880
0.2729
If mercury vapor is assumed to be an
ideal gas, the equilibrium concentration of
mercury vapor in a gas can be calculated using
vapor pressure data. Table A-l and Figure A-l
in the Appendix give this information.
2.3 ORIGIN, PRODUCTION, AND
USES OF MERCURY
2.3.1 World
Although there are 25 known mercury-
bearing minerals, the primary source of this
metal is cinnabar (mercuric sulfide).2 Other
economically important mineralogical species
of mercury include the sulfide minerals of
iron, arsenic, and antimony.
The gangue associated with cinnabar
deposits includes carbonate and silicate min-
erals such as calcite, chalcedony, dolomite,
opalite, quartz, and serpentine.2 The deposits
usually are shallow, extending downward to
depths of slightly less than 2500 feet. Almost
2-2
-------
all of the deposits are in areas of tertiary or
quaternary volcanic activity. The most
important deposits occur in Italy, Spain, the
U.S.S.R., Yugoslavia, China, the U.S., Canada,
Mexico, and the Philippines.
The annual world production of mercury
by country and the average London-based
price per flask from 1968 through 1970 are
given in Table 2-4.
Table 2-4. WORLD PRODUCTION OF MERCURY
BY COUNTRY FOR 1968 THROUGH 19703
(76-pound flasks3)
Country
Bolivia (exports)
Canada
Chile
China, mainland
Colombia
Czechoslovakia
Ireland
Italy
Japan
Mexicod
Peru
Philippines
Spain
Tunisia
Turkey
U.S.S.R.b
United States
Yugoslavia
Total6
Year
1968
134
5,700
513
20,000
362
116
53,317
5,084
17,202
3,132
3,544
56,943
309
4,670
45,000
28,874
14,794
259,694
1969
68
21,200
286
20,000
344
435
420
48,733
6,543
22,539
3,365
3,478
64,862
244
6,556
47,000
29,640
14,330
290,043
1970
12
24,400
380b
20,000
350b
2,000b
1,604C
44,382
5,907
30,269
3,400b
4,648
47,689
100
8,592
48,000
27,303
15,461
284,497
"Average price per 76-lb flask, London = $546.80 in 1968; $536.41 in
1969; and $411.45 in 1970.
bEstimate.
''Sales only.
^Official figures as reported by Statistical Office, Secretary of Industry
and Commerce, Mexico; overall production of mercury believed to be
much higher.
^otal is of listed figures only.
2.3.2 United States
The major mercury deposits in North
America are located in the western part of the
continent (see Figure 2-1). A summary of the
United States mercury statistics is given in
Table 2-5. These statistics indicate a consider-
able decrease in U.S. mercury production in
1971 over previous production, which
probably results from decreased demand and
lower prices. Production decreased from
27,303 flasks in 1970 to an estimated 17,445
flasks in 1971. During this same period, prices
fell from a 1970 average of $408 per flask to
a 1971 year-end price of $218 per flask.
California mines accounted for more than
85 percent of the domestic production of
mercury during the first 50 years of the
twentieth century.3 During recent years,
California and Nevada have been the major
producing states. The production in the
United States by states for 1969 and 1970 is
summarized in Table 2-6.
Mercury consumption in the United
States generally increased from 1950 to 1969
but declined significantly in 1970 and 1971.
Table 2-7 summarizes the consumption of
mercury in the United States, by use, from
1950 to 1971. The major mercury uses in
recent years have been in electrical apparatus,
electrolytic preparation of chlorine and
caustic soda, and mildew-proofing for paints.
The future use of mercury in the electrolytic
preparation of chlorine and caustic soda and
as a mildew-proofing for paints is expected to
be substantially reduced by recently proposed
Environmental Protection Agency control
actions. Figure 2-2 gives trends in production,
consumption, and price of mercury.
2.4 EMISSION MECHANISMS AND
GENERALIZED CONTROL PRINCIPLES
Elemental mercury has unusually high
volatility for a metal. This property can lead
to high emissions of mercury in process gas
streams and in room ventilation air.
2.4.1 Process Gases
Gases present in various processes in-
volving mercury tend to become saturated
with mercury vapor. These gases, if vented
without treatment, carry the mercury vapor
into the atmosphere. The amount of mercury
emitted in this manner is independent of the
kind of gas involved, but depends upon: (1)
the temperature of the system, (2) the degree
of saturation of the gas, and (3) the volumet-
ric flow rate of the gas.
These dependency factors suggest that
appropriate control procedures consist of: (1)
cooling the gas to condense mercury, (2)
reducing the mercury vapor concentration
2-3
-------
MARSH MOUNTAIN
* LEGEND
AMAJORDjSTRlcTORMINE
• DISTRICT OR MINE
ODISTRICT, MINE, OR OCCURRENCE;
PRODUCTION SMALL OR UNKNOWN
BELT CONTAINING MANY
SMALL DEPOSITS
MORTOI
CENTRAL OREGON.
BLACK BUTTE
BONANZA
ALTOONA
WILBUR SPRINGS. CLEAR LAKE. KNOXVILLE
MAYACMAS, GUERNEVILLE
NEW ALMADEN'
NEW IDRi
SAN LUIS OBISP
DOME ROCK MOUNTAINS
CUARENTO
SAIN ALTO
CANOAS
HUAHUAXTIA
CINNABAR
IDAHO ALMADEN
OPALITE. BRETZ. CORDERO
NEVADA BELT
* ARKANSAS
MAZATZAL MOUNTAINS
• TERLINGUA
NUEVO MERCURIO
r
^^-GUADALCAZAR
y?
Figure 2-1. Mercury deposits in North America.'
Table 2-5. UNITED STATES MERCURY STATISTIC^
Producing mines
Production, 76-lb flasks
Price,5 $
Value, S103
Exports, flasks
Re-exports
Imports:
For consumption
General
Stocks Dec. 31
Consumption
Year
1966
130
22,008
441.72
9,722
357
476
31,364
34,757
20,076
71,509
1967
122
23,784
489.36
11,639
2,627
475
24,348
23,899
18,277
69,517
1968
87
28,874
535.56
15,464
7,496
103
23,246
23,956
22,907
75,422
1969
109
29,640
505.04
14,969
507
108
31 ,924
30,848
22,692
77,372
1970
79
27,303
407.77
11,134
4,653
50
21,972
21,672
16,376
61,503
1971a
30
17,445
292.41 c
7,232
29,732
52,725
a1971 preliminary estimates.
Average price per flask. New York.
cYear-end price per flask = $218.
2-4
-------
Table 2-6. MERCURY PRODUCED IN THE UNITED STATES,
BY STATE FOR 1969 AND 19703
State
California
Idaho
Nevada
Oregon
Alaska, Arizona,
New York, and Texas
Alaska, Arkansas,
New York, Texas,
and Washington
Total
Producing mines
1969
72
1
24
4
8
109
1970
51
1
13
5
9
79
Flasks*
1969
18,480
1,012
8,165
43
1,940
29,640
1970
18,593
1,038
4,916
274
2,482
27,303
Value ,b$103
1969
9,333
511
4,124
22
979
14,969
1970
7,583
423
2,005
112
1,012
11,134
aFor 76-pound flasks.
bValue calculated at average New York price.
through chemical scrubbing or adsorption,
and (3) reducing the volumetric flow rate.
In order to estimate the quantity of
mercury emitted, equilibrium (that is, com-
plete saturation) is usually assumed. The
results represent the maximum emission of
mercury if the gas stream is not saturated, but
more frequently a minimum emission if the
gas stream is saturated and also contains
mercury particulates. The required calcu-
lations are presented in detail in the Appendix
together with a table of calculated results.
2.4.2 Ventilation Air
Mercury emissions can occur in ventila-
tion air if mercury is exposed to the circu-
lating air either as a result of uncovered
containers or through incomplete clean-up of
spills. In this case, the concentration of
mercury vapor is much lower than the satura-
tion value; but because of the large volumes
of ventilation air used, relatively large emis-
sions of mercury can occur. The amount of
mercury emitted, E, is the product of the
ventilation air volumetric flow rate and the
average mercury vapor concentration. The
following expression may be used:
E = 0.09x 10'6 xVxC
where: E = mercury emission, pounds per
day
V = ventilation volumetric flow rate,
cubic feet per minute
C = mercury vapor concentration,
micrograms per cubic meter.
The dependency factors cited in Section
2.4.1 indicate the appropriate approaches to
the control of mercury emissions in ventila-
tion air; namely, to reduce the volume of the
air or to reduce the concentration of mercury
vapor in the air. Because the ventilation air
volume is usually dictated by area cooling
requirements, reduction of concentration re-
mains the only recourse. The concentration
can be reduced by minimizing the sources of
exposed mercury through careful handling
and good housekeeping.
Frequently, the concentration of mercury
vapor in the ventilation air is not known. A
conservative estimate of this mercury emis-
sion may be made by assuming the vapor
concentration to be equal to the current
Threshold Limit Value (TLV) set by the
American Conference of Governmental Indus-
trial Hygienists (ACGIH), which is 50 micro-
2-5
-------
Table 2-7. MERCURY CONSUMED IN THE UNITED STATES3'4'5
(76-pound flasks)
Use
Agriculture3
Amalgamation
Catalysts
Dental preparations
Electrical apparatus
Electrical preparation
of chlorine and
caustic soda
General laboratory
Industrial and control
instruments
Paint
Antifouling
Mildew-proofing
Paper and pulp manufacturing
Pharmaceuticals
Redistilledb
Other0
Total identified uses
Total unidentified uses
Grand total d
Year
1950
4,504
192
2,743
1,458
12,049
1,309
646
5,385
3,133
—
—
5,996
7,600
4,200
—
—
49,215
1960
2,974
255
1,018
1,783
9,268
6,211
1,302
6,525
1,360
2,861
3,481
1,729
9,678
2,722
—
—
51,167
1965
3,116
268
924
1,619
16,097
8,753
1,119
4,628
255
8,211
619
418
—
15,402
73,560
—
73,660
1966
2,374
268
1,932
1,334
16,257
11,541
1,563
4,097
140
8,280
612
232
—
15,632
71,509
—
71,509
1967
3,732
219
2,689
1,359
14,610
14,306
1,133
3,865
152
7,026
446
283
—
12,568
69,517
—
69,517
1968
3,430
267
1,914
2,089
17,484
17.458
1,246
3,935
392
10,174
417
424
—
7,945
75,422
—
75,422
1969
2,689
195
2,958
2,880
18,490
20,720
1,936
6,655
244
9,486
588
712
9,134
76,657
715
77,372
1970
1,811
219
2,238
2,286
15,952
15,011
1,806
4,832
198
10,149
226
690
5,858
61,276
227
61,503
1971
1,478
996
1,871
16,646
12,252
1,357
3,906
414
8,192
2
668
2,292
50,074
—
52,725
alncludes fungicides and bactericides for industrial purposes and, prior to 1959, also includes pulp and paper manufacturing.
Redistilled" used in industrial instruments, dental preparations, and electrical apparatus.
c"0ther" includes mercury used for installation of chlor-alkali plants for 1963 and later dates.
All items do not add up to the total given, which has been increased to cover approximate total consumption.
-------
grams per cubic meter. For example, assuming
that the concentration does not exceed the
current TLV and using a ventilation flow rate
of 1 million cubic feet per minute of air, the
estimated cell room emission rate would be
4.5 pounds per day.
t/o
U.S. PERCENTAGE OF WORLD PRODUCTION
U.S. INDUSTRIAL CONSUMPTION
1950
Figure 2-2. Trends in U. S
for 1971 are preliminary.
1955
1960
1965
19701971
YEAR
;. production, consumption, and price of mercury.3 Values
2-7
-------
2.4.3 Particulate Emissions
In addition to vapor losses, mercury
emissions can occur in the particulate form as
elemental mercury mist, solid mercury com-
pounds, and mercury adsorbed on soot. Par-
ticulate emissions are frequently encountered
when a hot gas saturated with mercury vapor
is cooled. The mercury tends to condense in
the form of a mist that may be entrained in
the gas stream and carried to the atmosphere,
thus obviating the effect of the gas cooling
step.2 Amounts of particulate mercury emis-
sions cannot be readily predicted. Control
consists of collection and retention of the
particles involved by means of an entrainment
separator, often called a mist eliminator.
It will be seen in the following discus-
sions of each specific industry that there is a
commonality of principle inherent in control
practice but not necessarily a commonality of
application because of the system variables
encountered.
2.5 REFERENCES FOR SECTION 2
1. Weast, R.C. Handbook of Chemistry and
Physics (50th Ed.). Cleveland, Chemical
Rubber Company, 1970. p. B-31 to B-32,
B-129, B-262, B-491 to B^94, D-56,
D-139, F-6, F-23.
2. Pennington, J.W. Mercury — A Materials
Survey. U.S. Department of Interior,
Bureau of Mines. Washington, D.C. Infor-
mation Circular 7941. 1959. p. 9-10.
3. Cammarota, V.A., Jr. 1970 Bureau of
Mines Minerals Yearbook, Mercury (Pre-
print). U.S. Department of Interior,
Bureau of Mines. Washington, D.C. 1971.
p. 1,3-5,8.
4. U.S. Bureau of Mines Minerals Yearbook,
Mercury. U.S. Department of Interior,
Bureau of Mines. Washington, D. C.
1950-1968.
5. Mercury in the Fourth Quarter of 1971.
U.S. Department of Interior, Bureau of
Mines. Washington, D.C. February 22,
1972. 5 p.
2-8
-------
3. MERCURY EMISSIONS, CONTROL TECHNIQUES, AND CONTROL COSTS
3.1 PRIMARY MERCURY PRODUCTION
3.1.1 Emissions and Process Description
3.1.1.1 Rotary Furnaces
3.1.1.1.1 Process Description — Most of the
mercury produced in the United States is
extracted by the use of directly fired rotary
furnaces. A diagram of a typical rotary
furnace operation is shown in Figure 3-1-1
This type of furnace generally ranges from 2
to 7 feet in diameter and 24 to 140 feet in
length; the ore treatment capacity ranges
from 10 to 300 tons of ore per day.2 "6
Crushed mercury ore ranging in size
from 3/4 to 4 inches in diameter is fed into
the rotary furnace by means of a reciprocat-
ORE FROM
ing tube-type feeder known as a "shotgun"
feeder. The ore feed rate is adjusted with this
device to obtain ore temperatures of 1100° to
1600°F within the furnace.2"6 The proper
furnace retention time for a particular ore
may vary from 15 to 60 minutes. The ore
retention time is determined by the inclina-
tion slope of the furnace, which varies from
1/2 to 3/4 inch per foot, and by the rate at
which the furnace rotates, generally between
1/2 and 3-1/2 revolutions per minute.2"6 The
calcined ore is discharged from the furnace
into calcine or burnt ore bins, from which it is
subsequently transported to a waste dump.
The hot calcined ore is used in some cases to
preheat the combustion air or the feed ore.
STACK
GAS
EXPANSION TANK
FAN, DUST COLLECTOR,
AND DUST RECEPTACLE
COARSE
ORE
BIN
FEEDER
r CRUSHER
BELT CONVEYOR
"SHOTGUN"FEEDER
REFRACTORY LINING -'2
SEAL
RING
--^BURNER
BURNT-
ORE
BIN
Figure 3-1. Pyrometallurgical process for producing mercury.1
3-1
-------
Oil is usually burned as fuel with approx-
imately 50 percent excess air; however, if the
oil is sufficiently preheated and atomized, 35
percent excess air can be used.6'7 The rate of
oil use generally ranges from 7 to 10 gallons
of oil burned per ton of ore furnaced. If
natural gas is used, the excess air used can be
reduced to 25 percent because of the more
efficient combustion of this fuel.7
The hot combustion gases flow counter-
currently to the ore and heat the ore to
temperatures that volatilize the mercury. The
mercury-bearing vapors leave the furnace at
temperatures ranging from 450° to 600° F and
pass through one or more cyclone separators,
which remove most of the particulate matter.
The dust collectors are operated at 450° to
500° F to avoid condensation of the mercury
vapor.2 "6-8 A blower on the downstream side
of the cyclone maintains a slight suction on
the furnace to minimize leakage of hot
mercury-laden gases from the furnace.
After passage through the cyclone and
blower, the gas stream is usually divided and
introduced into sufficient banks of air-cooled
verticle U-tube condensers to control its flow
rate and temperature. The condenser pipes are
about 16 inches in diameter and from 20 to
40 feet in height.4 ~6'8 The condensers are
constructed of cast iron, mild steel, tile,
stainless steel, Monel,* or fiber glass, de-
pending on the sulfur or chloride content of
the ore and the subsequent corrosiveness of
the gases. The individual pipes are connected
at the top and bottom to adjacent pipes. The
lower ends of consecutive condenser pipes are
connected and sealed with water in a con-
tainer called a launderer, or in the case of
larger capacity plants, a condenser tank.
Condensed mercury, dust, and soot fall down
the condenser pipes into the launderer, or
into buckets that are submerged in the laun-
derer and placed under each connection.
Periodically, the condenser pipes are washed
down with a water spray to remove the
* Mention of commercial products or commercial
names does not constitute endorsement by the
Environmental Protection Agency.
mercury, dust, and soot adhering to the
insides of the pipes.
The mercury-bearing mud accumulated in
the launderer or in the buckets is periodically
collected, transferred to a hoeing table, mixed
with lime, and hoed manually or mechanically
to collect the mercury. During the hoeing
operation, the mud is scraped and pressed,
causing minute mercury droplets to coalesce
and form larger drops of mercury that flow
into a collection tank. The mud remaining
after the hoeing operation is completed is
processed in a retort furnace or is recycled
back into the main ore furnace.
The total length of a condenser system
depends on the volume of gas and the amount
of cooling desired. The condenser system
normally is designed to provide a temperature
of less than 110°F out of the stack.4-5-8
Although the condenser normally depends on
natural air convection for cooling, a common
practice is to spray water on the outside of
the hottest pipes in the condensing system,
particularly during the summer months.
After the gases leave the condenser sys-
tem, they are expanded into one or more
wooden tanks designed to maintain the pro-
per system draft.6 These wooden expansion
tanks also provide some additional cooling
and mist elimination. From these tanks, the
condenser gases pass into the stack from
which they are emitted to the atmosphere at
temperatures of 90° to 110°F.2 '6'8 A cooling
practice that is sometimes employed involves
spraying water into the wooden expansion
tanks to cool the condenser gases further.
The calculated combustion gas volumetric
flow rates for a typical directly heated fur-
nace employing propane or fuel oil with 0,
25, and 50 percent excess air are given in
Table 3-1.
As can be seen from Table 3-1, the
calculated volume of gases generated from
combustion and excess air for the two fuels
listed in the table are essentially equal. The
use of propane or natural gas as fuel would
offer a slight advantage because less soot
would be generated and a smaller amount of
3-2
-------
Table 3-1. CALCULATED VOLUMETRIC FLOW
RATES (scfm)a DUE TO COMBUSTION AND
EXCESS AIR FOR A 100-TON/DAY ORE RATE
Fuel
Fuel
Fuel oil
Propane
Excess air, %
0
800
900
25
1000
1100
50
1200
1300
a Standard conditions are 70° F and 29.92 in. Hg.
excess air would be necessary as a result of
more efficient combustion. The direct advan-
tages are reduced stack gas emissions and
particulate matter formation, which in turn
reduce both mercury vapor losses and mer-
cury losses due to adsorption of the metal on
soot.
Excess air should be kept to a minimum
as dictated by complete fuel combustion and
sulfur oxidation inasmuch as the mercury
emission rate is dependent on the stack gas
volumetric flow rate. Table 3-1 indicates that
in going from 5 0 percent to 25 percent excess
air furnace operation, a volumetric flow rate
reduction of 15 percent would result.
"Combustion gas," as used in Table 3-1,
is the gas, including excess air, that enters the
furnace immediately after combustion has
occurred. In actuality, other gases, such as
water vapor or sulfur dioxide, are driven off
the ore or are formed during combustion.
These additional gases increase the volume of
gas that flows out of the stack. Furnace
design calculations for directly fired furnaces
generally indicate a total stack volumetric
flow rate of 1500 to 1700 standard cubic feet
per minute (scfm) per 100 tons of ore per
day, assuming that an averaged silica carbon-
ate type mercury ore is being calcined.6
3.1.1.1.2 Emissions - The major emissions of
mercury from a primary mercury furnacing
operation occur from the condenser stack.
Other minor emission points are dust and
vapor emissions from the mining operation,
the furnace room ventilation air, the hoe table
ventilation air, and emissions from the hot
discharged ore. These minor emissions can be
minimized by good housekeeping and oper-
ating practices.
The mercury emissions from the con-
denser stack effluent can vary with the grade
and type of ore processed, and with variations
in plant operating practices. A typical con-
denser stack gas is described below:
1. Stack gas temperatures of 90° to
110°F.
2. Stack gas saturated with mercury
vapor.
3. Particulate mercury emission equal to
the mercury vapor emission. (This is
based on source testing results ob-
tained by the Environmental
Protection Agency at several mercury
extraction facilities.)9 -11
4. Stack gas volumetric flow rates of
1000 to 1600 standard cubic feet
per minute.
Stack emissions are calculated and presented
in Table 3-2.
Table 3-2. CALCULATED MERCURY EMISSIONS
FROM THE CONDENSER STACK OF
A PRIMARY MERCURY EXTRACTION
FACILITY
Emissions, Ib/day
Temperature, °F
90
110
At 1000 scfm
6.6
15.4
At 1600 scfm
10.5
24.7
A condenser stack gas flow rate of 1600
standard cubic feet per minute corresponds to
a 100-ton-per-day ore treatment rate; there-
fore, the estimated emission from a 100-ton-
per-day facility with a 110°F stack effluent
temperature is approximately 25 pounds per
day.
In actual emission sampling tests con-
ducted by the Environmental Protection
Agency at three primary mercury extraction
facilities, mercury emissions per 100-ton-per-
day ore treatment rate ranged between 18 and
59 pounds per day.9 "*1
3.1.1.2 Multiple-Hearth Furnaces
3.1.1.2.1 Process Description - Multiple-
hearth furnaces are not in common use in the
3-3
-------
United States for primary mercury produc-
tion; however, in Europe roughly 50 percent
of the primary mercury operations employ
these furnaces. A large Canadian mining oper-
ation is presently using a multiple-hearth
furnace to process a flotation concentrate.
The ore treatment prior to furnacing is
more elaborate for a multiple-hearth furnace
than for a rotary furnace. The ore feed to the
furnace can be concentrated by flotation or
other beneficiation processes, or it can be
furnaced directly as mined. Beneficiation of
ore is discussed in Section 3.1.1.3.
A multiple-hearth furnace consists of a
series of circular refractory hearths, placed
one on top of the other, that are usually
enclosed in a steel shell (Figure 3-2). A center
vertical shaft rotates arms that mix the ore by
moving it in a spiral path across each hearth.
The ore is fed to the top hearth, where it
moves across to drop-holes before falling to
the next hearth. The ore continues its travel
from hearth to hearth in this manner until it
is discharged from the bottom of the furnace.
Heat can be supplied alternatively by
combustion of the charge elements, com-
bustion of fuel in burners on certain hearths
(direct firing), combustion of fuel in a
separate combustion chamber (indirect
firing), and combustion of fuel in muffles
(indirect heating); or by heating with elec-
trical resistor elements (indirect heating). In
the most common application, hot com-
bustion gases flow countercurrently to heat
the ore to temperatures that will volatilize
mercury. The ore is heated continually from
the top hearth to the bottom hearth. The
temperature and atmosphere of each hearth in
a multiple-hearth furnace may be closely
regulated.7
The calcined ore is discharged from the
bottom of the furnace at a temperature of
approximately 1200°F. The mercury-
vapor-laden combustion gas is discharged
from the top of the furnace at temperatures
ranging from 500° to 600°F. The gas then
enters a cyclone where large particulate mat-
ter is removed. A fan on the downstream side
of the cyclone maintains a slight suction on
the furnace, minimizing vapor emissions from
the furnace. The condenser system for a
multiple-hearth furnace is essentially the same
as that used for a rotary furnace.
3.1.1.2.2 Emissions — The condenser stack
gas flow rate and temperature are approx-
imately the same for a multiple-hearth fur-
nace as for a rotary furnace; therefore, the
stack emissions of mercury are approximately
equal. Minor mercury emission points from a
multiple-hearth furnace operation are similar
to those previously described for a rotary-
furnace operation.
If an indirect method were used to heat
the multiple-hearth furnace, the volume of
gases produced would be much less, consisting
primarily of water vapor, a small amount of
air, and sulfur dioxide produced by the
oxidation of sulfur. This small volume of gas
would greatly reduce the condenser require-
ment, resulting in a large emission reduction.
Such furnaces are commercially available but
have not been used in the primary mercury
industry because of cost.
3.1.1.3 Beneficiation
3.1.1.3.1 Process Description — Convention-
ally, mined ore is subjected to pyrometal-
lurgical treatment without preliminary con-
centration; however, on the basis of differ-
ences in the physical properties of the mineral
species, for example, density and surface
characteristics, methods for preliminary con-
centration have been used in specific in-
stances. These include hand sorting, jigging,
tabling, and flotation.8 Flotation, when feas-
ible, is the most effective of these techniques.
Mercury recoveries of 80 to 90 percent can be
attained with this method, depending mainly
on the type of rock in which the cinnabar is
contained. 8«12-13 A typical mercury ore
flotation flow sheet is shown in Figure 3-3.14
In a flotation process, the ore is normally
subjected to a two-stage crushing operation
followed by grinding in a ball mill to approx-
imately 65 mesh.12 Finer grinding (100 mesh
3-4
-------
GAS OUTLET
STEEL SHELL-
FEEDING -
MECHANISM S
ADJUSTABLE
FEED KNIFE
CENTRAL SHAFT SECTION
Figure 3-2. Multiple-hearth furnace.7
3-5
-------
MERCURY ORE
SETTLING AND DEWATERING
BIN
HIGH-GRADE
MERCURY
CONCENTRATE
Figure 3-3. Flotation flow sheet for cinnabar ore.14
3-6
-------
or smaller) may be necessary to improve
yields in some ore types. The finely ground
ore is then passed through a jigging operation
in which the heavier mercury minerals are
separated by gravity and are removed as
concentrate. The lighter overflow material
from the jig is fed to a series of flotation cells
where air bubbles float the cinnabar by use of
flotation reagents into a concentrate. Con-
centrates containing from 200 to as high as
1000 pounds of mercury per ton of concen-
trate are produced. Concentrate is more
amenable to retorting, multiple-hearth
roasting, and hydrometallurgical processing
than to a directly fired rotary-furnace oper-
ation.
Flotation and other beneficiation pro-
cesses are not generally used in the United
States in the primary mercury industry. As a
general rule, the installed cost for a complete
flotation concentration system, exclusive of
electrical power and water provisions, is about
$3500 per ton of ore processed per day. The
operating cost is approximately $5.00 per ton
of ore processed.12'13 The low recovery
efficiency and the high initial investment and
operating cost have made this method un-
attractive in the United States.
The water required for flotation ranges
from 1.5 to 4.0 tons per ton of ore pro-
cessed.12-15 About 70 to 80 percent of this
water can be reclaimed if settling ponds are
used.12-15
Some advantages of flotation are listed
below:14
1. Furnacing plant operations that
employ flotation can be more flex-
ible because of smaller furnace ton-
nages.
2. Mercury emissions are reduced con-
siderably in directly heated furnace
operations since a lower ore treat-
ment tonnage results in a lower fur-
nace stack gas volume per flask of
mercury collected. The emission
reduction factor is estimated to be
roughly equal to the concentration
ratio.
3.1.1.3.2 Emissions — The flotation oper-
ation, unless properly designed, is a potential
source of water pollution. The air emissions
from flotation concentration operations are
small and may occur as dust emissions during
the crushing and grinding operations.
3.1.1.4 Retort
3.1.1.4.1 Process Description — A retort is an
indirectly heated furnace and is small in
comparison to directly heated furnaces.
Retorts are generally classified as pipe retorts
or D-retorts, depending on their size and
shape.16
The pipe retort consists of circular iron
pipes 8 to 12 inches in diameter and 7 to 9
feet in length. Several pipes may be collected
in banks with one fire box containing up to
12 pipes (Figure 3-4). These pipes can be
horizontally situated, in which case the con-
denser gases are taken from the rear of the
retort (Figure 3-4); or they can be inclined,
with the mercury vapor evolving from the
high side of the pipes (Figure 3-5). The
capacity of each pipe is on the order of 5
cubic feet of ore per charge. The furnacing
time varies from 12 to 24 hours per charge of
ore.
The D-retort, illustrated in Figure 3-6, has
a larger capacity than the pipe retort and has
a cross section resembling the letter D. The
ore charge for a D-retort varies from 5 to 10
cubic feet of ore. The furnacing times are
similar to those of the pipe retort described
above. Some D-retorts are operated under a
slight suction produced by a blower in the
retort stack. In another design, a small air
flow (10 cubic feet per minute) is blown
through the retort to move the mercury vapor
to the condenser system.
In operation, a retort is charged with
directly mined ore or concentrated ore, after
which the retort door is sealed with special
clay that prevents mercury vapor leaks. Lime
is mixed with the ore to aid in the oxidation
of sulfur to sulfur dioxide.16 The ore charge
is heated from 12 to 24 hours at temperatures
3-7
-------
SEALING
DEVICE
FRONT VIEW
Figure 3-4. Group horizontal pipe retort.16
of 1200° to 1400° F. At the end of this
heating period, the calcine is removed, and
another charge of ore is added.
The mercury-vapor-laden gas enters the
condenser system from the retort. The con-
denser system may be of an air-cooled or an
indirect or direct water-cooled type. If water
is available, an indirect water-cooled con-
denser is usually used. A typical water con-
denser consists of a jacketed 12-foot section
of 3- or 4-inch-diameter pipe. The condensed
mercury is usually collected under water and
in general does not require further purifi-
cation. In some cases mercury-bearing solids
collected in the condenser system are period-
ically mixed with lime and are hoed to collect
the mercury. Retort processing is not readily
adaptable to large mining operations because
of high labor requirements and low treatment
rates.
The stack gas volumetric flow rate from a
retort is very small because of the indirect-
heating method employed.11
3.1.1.4.2 Emissions — Stack emissions from a
retort are small because the stack gas flow
3-8
-------
STACK
CHARGE HOLE
WATER CONDENSER
SECTION A-A
A-I
SECTION B-B
Figure 3-5. Two-pipe, inclined pipe retort.^
rate is low. Possibly the largest emission
occurs during the opening and closing of the
retort door for removal of spent ore and
charging of ore concentrate.
Other emissions are from the retort room
ventilation air, from the hoeroom or table
ventilation air, and from the calcined concen-
trate. If good housekeeping practices are
followed, these emissions can be minimized.
3.1.2 Emission Reduction Resulting
from Process Changes
Reductions of mercury emissions can be
accomplished by adding a control device to
remove mercury from a gas or by using a
process that inherently produces lower emis-
sions. For new plants, a process that would
produce less mercury emissions would pro-
bably be employed. For existing plants, a
mercury control device might be added to the
existing process to reduce emissions; however,
in some cases, it may be more economical to
convert to a different process. The decision as
to which course of action to follow would
probably be determined by an economic
evaluation of both control methods. This
section will discuss emission reductions re-
sulting from process changes; Section 3.1.3
will discuss emission reductions resulting from
the application of control techniques.
3-9
-------
Hg
COLLECTION
!^-?>*ORE CHARGE "STr:?^
FRONT VIEW
SIDE VIEW
Figure 3-6. D - retort.16
3.1.2.1 Pyrometallurgical Processing
The largest mercury emission from a
mercury extraction facility using a directly
heated furnace occurs from condenser stack
gas emissions. These gases are saturated with
mercury vapor and also contain entrained
particulates. The major parameters governing
the emission rate are stack gas flow rate and
temperature. If either or both of these para-
meters are reduced, the mercury emissions
will also be reduced.
The temperature cannot generally be
reduced by a process change alone; a control
technique to cool the effluent would have to
be employed in most situations. This tech-
nique will be discussed in Section 3.1.3. For
the purposes of this discussion, the tempera-
ture will be assumed constant while the stack
gas flow rate is reduced by different pro-
cesses. A reduction in stack gas flow rate can
be accomplished by converting from a direct-
to an indirect-heating process or by bene-
ficiation of ore prior to furnacing.
3.1.2.1.1 Indirect Heating - Indirect heating
of ore, either concentrated or as mined, will
effect a reduction in mercury emissions from
the stack. The stack temperature generally
determines the concentration of mercury in
the stack gas; therefore, if the gas flow rate is
reduced, a proportional emission reduction
will occur at the same temperature. The retort
process and the multiple-hearth furnace using
muffle or electrical resistance heating can be
used. These processes are described in
Sections 3.1.1.2 and 3.1.1.4. The gases
evolving from the ore during roasting would
be the only gases produced and would consist
of moisture, sulfur dioxide, air, mercury
vapor, and small quantities of other gases. It is
estimated that the stack gas flow could be
reduced by 70 to 80 percent.
Because it is used for batch processing,
the retort is generally used only on a small
scale. The indirectly heated multiple-hearth
furnace requires a larger hearth area for the
same ore rate than the directly heated furnace
requires because of the lower heat transfer
efficiency inherent with indirect heating. The
larger area requirement would increase the
3-10
-------
capital investment and operating cost of an
indirectly heated furnace by a sizeable factor.
3.1.2.1.2 Beneficiation - Ore grade is im-
proved by the beneficiation processes of hand
sorting, jigging, tabling, and flotation. As an
example, if 100 tons of low-grade mercury
ore were concentrated by beneficiation
methods to 2 tons of higher-grade concen-
trate, the fuel needed to volatilize the mer-
cury would be reduced since 98 tons of
non-mercury-containing ore would not have
to be treated from ambient temperature to
1200°F. The lower fuel requirement would
result in a lower stack gas flow rate, thereby
resulting in a sizeable reduction in emissions.
The emission reduction factor is estimated to
be roughly equal to the concentration ratio.
In this example, the concentration ratio is
100/2, or 50. If both indirect heating and
beneficiation were used in the same process,
an even larger emission reduction would
result.
3.1.2.2 Hydrometallurgical Processing
Hydrometallurgical mercury ore pro-
cessing produces essentially no atmospheric
mercury emissions. The only potential emis-
sion is that of ore dust resulting from crushing
and grinding operations. Soluble mercury
compounds probably remain in the tailing and
leach solutions from the hydrometallurgical
operation, and these may create a water
pollution problem. Since virtually no data are
available on potential losses of mercury in the
solid residue and waste waters from sizeable
hydrometallurgical operations, it would be
difficult to estimate either the magnitude of
the water pollution problem or the cost of
alleviating it. Research and pilot studies have
indicated that water pollution problems can
be minimized if the operation is properly
designed.
3.1.3 Control Techniques
The primary mercury industry currently
employs little if any control technology for
reducing mercury emissions from the conden-
ser stack gases. As a result, few of the
techniques that are discussed as control
methods have been actually used in this
industry. Cooling, mist elimination, water
scrubbing, and the use of activated carbon are
control methods that have been used success-
fully in reducing mercury emissions from
similar gas streams and should be applicable
to reducing emissions from primary mercury
extraction plant effluents.
3.1.3.1 Directly Heated Furnaces
3.1.3.1.1 Cooling and Condensing - The basic
technique currently employed for minimizing
mercury emissions from the condenser stack
is the control of the temperature of the
effluent gases. A stack effluent temperature
range of 90° to 110°F will allow from 5.2 to
12.4 pounds of vaporous mercury to be
emitted to the atmosphere per day when the
stack flow rate is 1600 standard cubic feet per
minute. This is a typical stack flow rate for a
100-ton-per-day mercury extraction facility
and will be used for subsequent examples.
Assuming that the emission of particulate
mercury at least equals the amount lost as
vapor, the total loss of mercury from the
condenser stack would range from 10.4 to
24.8 pounds per day.
As the effluent temperature is lowered,
the mercury vapor content of the condenser
gas stream is decreased. This temperature
dependency is illustrated in Table 3-3 for five
volumetric flow rates and an assumed original
condenser outlet temperature of 110°F.
Cooling of the condenser gases can be
accomplished by the use of either direct or
indirect cooling techniques. Because of the
large particulate loading of the condenser gas
stream, direct cooling methods may possess
an advantage over indirect methods since
direct cooling aids in mist and particulate
removal. The use of direct cooling, however,
also introduces the necessity of water treat-
ment facilities and creates the possibility of a
water pollution problem.
If a supply of low-temperature water is
available for use as a cooling medium, the cost
of either a direct- or indirect-cooling system
can be substantially reduced. The majority of
3-11
-------
Table 3-3. CALCULATED VAPOROUS MERCURY EMISSIONS FOR
SELECTED CONDENSER STACK FLOW RATES AND TEMPERATURES
Condenser
volumetric flow
rate.
scfma
500
1500
2000
2500
3500
Condenser
exit temperature.
°F
110
80
70
60
55
110
80
70
60
55
110
80
70
60
55
110
80
70
60
55
110
80
70
60
55
Final
temperature
difference.
°F
0
30
40
50
55
0
30
40
50
55
0
30
40
50
55
0
30
40
50
55
0
30
40
50
55
Condenser
exit Hg vapor
concentration.
mg/m3 a
86.01
23.40
14.62
9.02
6.96
86.01
23.40
14.62
9.02
6.96
86.01
23.40
14.62
9.02
6.96
86.01
23.40
14.62
9.02
6.96
86.01
23.40
14.62
9.02
6.96
Hg vapor
condensed
by
cooling.
Ib/day
0
2.81
3.21
3.47
3.56
0
8.45
9.64
10.45
10.67
0
11.28
12.86
13.87
14.23
0
14.12
16.10
17.35
17.82
0
19.66
22.41
24.16
24.81
Hg vapor
atmospheric
emission.
Ib/day
3.87
1.06
0.66
0.40
0.31
11.61
3.16
1.97
1.22
0.94
15.49
4.21
2.63
1.62
1.26
19.39
5.27
3.29
2.04
1.57
27.01
7.35
4.60
2.85
2.20
Standard conditions, 70°F and 29.92 in. Hg.
the primary mercury extraction facilities do
not, however, have access to naturally occur-
ring low-temperature water, and some form of
chilling is required. If water is available but is
not of low enough temperature to cool the
effluent adequately, a chilled brine or Freon
refrigeration system could be employed to
produce low-temperature water. At those sites
where water is not available, a closed-loop
refrigeration system could be used.
In order to be effective, the cooling step
should be followed by some type of mist
elimination device.
3.1.3.1.2 Mist Elimination — The condenser
gas stream may contain mercury as mercury
vapor, mercury oxides, mercury mists, and
mercury adsorbed on soot and other partic-
ulate matter. Results of source tests by the
Environmental Protection Agency have in-
dicated that as much as 50 to 70 percent of
3-12
-------
f I-
WATER TO
INTERMITTENT SPRAYS [_ȣ
GAS IN --L
WATER OUT TO — t
BRINE SYSTEM
^
^_- ^~
\
r
71
:-j /^7X
1^
.^'.'
_^^.
^^5*C
~\ — ^ GAS OUT
^FIBER MIST ELIMINATOR
^ ELEMENTS
— WATER TO
l|— ' INTERMITTENT SPRAYS
J-^ WATER OUT TO
H BRINE SYSTEM
-, r LIQUID LIQUID SEPARATION
M^m SECTIONS
MERCURY OUT
TO RECOVERY
Figure 3-7. Mercury mist eliminator.17
the mercury that is emitted from the stacks of
primary extraction facilities is in the partic-
ulate form.9 "n Total mercury stack emis-
sions, therefore, may be two to three times
larger than the calculated vaporous emissions
presented in Table 3-3.
In one type of commercially available
mist eliminator, the mercury-particulate-laden
gases pass horizontally through a fiber bed
enclosed in screens. Clean gases exit from this
fiber bed, whereas the separated mercury
drains to the bottom of the mist eliminator.
Inertial impaction, Brownian movement, and
direct interception are the three basic
mechanisms utilized for mercury mist separ-
ation. In order to increase the rate of drainage
of the collected mercury from the fiber bed,
water or other liquid is intermittently or
continuously sprayed onto both sides of the
bed. The lower section of the eliminator
serves as a coalescer for the separation of the
collected mercury from the water. Figure 3-7
illustrates this device.
Since mercury mist eliminators are cur-
rently applied in the treatment of relatively
ciean gas streams (that is, low particulate
loading of contaminants other than mercury),
the effectiveness of this device for extracting
particulate mercury from a stream containing
high amounts of particulate matter is not
known. Additional particulate loading may
cause the efficiency of this technique to be
lower than normal. Use of additional flushing
water within the mist eliminator or instal-
lation of air precleaning equipment prior to
tiie mist elimination device may effectively
eliminate the problem.
The particulate mercury removal effi-
ciency of a mercury mist eliminator has been
estimated at 86 percent.18-19
3.1.3.1.3 Wet Scrubbing - Wet scrubbing
devices employ a variety of mechanisms to
collect particulate matter. Interception of
particulate matter by liquid droplets resulting
in a heavier dust-liquid agglomerate is the
most important of these mechanisms. A par-
ticle that has collided with a liquid droplet
resists separation because of van der Waals
forces. Particulate matter collected on a liquid
droplet in this manner can be efficiently
3-13
-------
removed from the gas stream by a centrifugal
collector. The particulate matter collection
efficiencies of these devices vary with energy
input and can extend over a wide range
depending on the design.
The liquid-solid slurry formed can be sent
to a settling tank either directly from the
device or from a subsequent centrifugal col-
lector. The settled scrubbing liquid may be
pumped back to the scrubber and reused. A
bleed stream must be taken off of the settling
tank to maintain a stable level in the tank
since particulate matter and moisture are
removed from the gas stream. The bleed
stream usually represents 3 to 5 percent of
the liquid-solid slurry and is a potential source
of water pollution. This stream can be dis-
charged into a secondary settling tank where
the mercury solids are separated, and the
liquid either discharged or given additional
treatment.
Scrubbing systems have several advan-
tages when compared to other particulate
matter collection devices:
1. The required capital expenditure is
normally lower for scrubbers than for
other types of gas-cleaning equip-
ment.
2. The collection efficiency of the
scrubber is flexible, depending on the
power input.
3. A large range of particulate sizes can
be collected.
4. There are no secondary dust pro-
blems since the disposal of collected
contaminants is a wet operation.
5. The cost of maintenance is low since
there are few or no moving parts.
6. Scrubbers are compact in size.
7. Since the collecting operation is wet,
there is virtually no limitation on the
gas stream humidity and tempera-
ture.
8. Wet scrubbers allow the simultaneous
collection of pollutants that are in
the gaseous, liquid, or solid form.
There are numerous types of wet scrub-
bers in use that remove particulates from a gas
stream, examples of which are the packed-bed
scrubber and the venturi scrubber.
Several variations of low-energy, low-
efficiency, packed-bed scrubbers have been
employed in the primary mercury industry to
remove mercury from gaseous streams. In one
domestic primary extraction facility, a system
has been used that consists of water sprays in
a packed redwood tank.20 Another extrac-
tion operation scrubs its condenser gases in a
similar packed-bed water scrubber.21 In most
situations, the low-energy scrubbing system
will prove inadequate for mercury removal
because of its low particulate removal effi-
ciency.
The venturi scrubber is a high-energy,
high-efficiency scrubber that has been used to
remove particulate matter from gas streams
similar to the condenser stack gas streams of
primary mercury extraction facilities. The
water necessary to scrub a 1600-standard-
cubic-foot-per-minute gas stream is estimated
to be between 5 and 9 gallons per minute,
depending on the gas loading and size distri-
bution of the particulate matter. It is ex-
pected that a particulate mercury collection
efficiency of 95 percent could be achieved
with a pressure drop of 20 to 30 inches of
water through the venturi scrubber and
cyclone.22 A suction fan located after the
cyclone is necessary to overcome this pressure
drop.
A venturi scrubber that produces suf-
ficient turbulence within the venturi to allow
cooling and condensing to occur is com-
mercially available.22 The scrubbing water
required for this situation would necessitate
cooling prior to injection. This type of venturi
scrubber would not require precooling of the
gas stream in order to achieve good collection
efficiencies.
3.1.3.2 Control of Indirectly Heated
Furnace Operations by Use of Treated
Activated Carbon
The use of either sulfur- or iodine-
impregnated activated carbon as a control
technique for the removal of mercury vapor
3-14
-------
from condenser eases of a retort opera-
tion should prove adequate. Some problems
could arise if substantial amounts of partic-
ulate mercury are present in the gas stream;
however, this situation can be corrected either
by preheating the gas stream or by using a
mist elimination device prior to the carbon
bed. Although activated carbon has not been
used specifically for the treatment of the
condenser stream of an indirectly heated
primary extraction facility, this technique has
been successfully used to treat similar process
streams. The mercury vapor is adsorbed by
the carbon and reacts with the impregnated
sulfur or iodine to form mercury compounds.
In order for treated activated carbon to
perform efficiently, the gas stream velocity
through the bed should be in the range of 20
to 40 feet per minute.23 This low gas velocity
is required to allow sufficient contact time
between the mercury and the treated carbon.
As previously mentioned, the particulate
loading of the gas stream, including both
particulate mercury and other solid and liquid
contaminants, must be low or the carbon bed
will plug and lose its efficiency. Destructive
distillation of the spent carbon appears
practical for recovering the adsorbed mercury.
Because of the low condenser gas flow
rates and the low particulate loading in a
retort operation, it is estimated that the
mercury vapor collection efficiency of the
treated activated carbon could approach 99
percent.23
3.1.4 Control Costs
Four basic methods for the removal of
mercury from the condenser stack gases of a
primary extraction facility have been
described. These four techniques are cooling
and condensing, mist elimination, wet scrub-
bing, and treated activated carbon. This
section will present capital and annual oper-
ating costs for each of the preceding control
techniques. All cost estimates are based on a
model mercury extraction facility of 100 tons
of ore per day capacity having a condenser
stack gas flow of 1600 standard cubic feet per
minute at 110°F. Equipment costs are based
on the use of titanium and titanium-clad
construction materials, which are required
because of the corrosive nature of the conden-
ser gases. All equipment costs were obtained
directly from various users and vendors of
control equipment.
The capital costs of specific systems are
itemized and listed in the Appendix, Section
A.3. The method employed for estimating the
capital requirement and the annual cost is
outlined in Section A.2.
3.1.4.1 Cooling and Mist Elimination
A control system that utilizes cooling to
55°F followed by partial mist elimination by
means of a knockout drum could reduce
emissions from the condenser stack gas stream
to 5.8 pounds per day. If indirect cooling
were used, the capital and annual operating
costs for the preceding system would be
$76,000 and $23,000, respectively. It is
estimated that the use of a direct cooler
would reduce the capital requirement to
$51,000. The annual operating cost for this
system would be $15,000. This estimate,
however, does not include the cost of water-
treatment facilities, which could be sub-
stantial.
The addition of a mist elimination device,
similar to the one described in Section
3.1.3.1.2, to the preceding control system
could reduce emissions to 1.7 pounds per day.
The capital and annual operating costs for
such a system utilizing indirect cooling would
be $108,000 and $32,000.
Table A-5 of the Appendix provides more
complete data on capital costs and expected
emissions. Figure 3-8 illustrates a manufac-
turer's estimate of the purchase cost of a mist
eliminator as a function of volumetric flow
rate.
3.1.4.2 Wet Scrubbing
The cost of controlling atmospheric mer-
cury emissions by means of a scrubbing
technique will vary considerably with the
type of scrubbing system employed.
The cost of a low-energy scrubber
3-15
-------
100,000
g 10,000
1,000
-i r mi i n-
100
1,000
GAS FLOW, acfm
10,000
Figure 3-8. Manufacturer's estimate of
purchase cost of a fiber pad type of mist
eliminator.
(wooden tank) similar to the type employed
by one domestic extraction plant is minimal.
It is estimated that the capital requirement
would be approximately $2000. This cost
does not allow for a chiller or for water
purification facilities; moreover, the effective-
ness of this system is questionable.
The estimated equipment cost of a
packed scrubbing tower constructed of car-
bon steel is about $0.30 per cubic foot per
minute of gas flow.24 Adjusting this price for
the use of a corrosion-resistant construction
material, such as fiber glass, and adding the
cost of a water pump and a blower to handle
a pressure drop of 3 inches of water, the
capital requirement for a packed tower of
1600 standard-cubic-foot-per-minute capacity
would be $5000. The capital and operating
costs for a control system utilizing a packed
tower and chilled water for direct cooling of
the condenser gases to 55°F have been
presented previously in Section 3.1.4.1. Insuf-
ficient data are available with which to
evaluate the efficiency of a packed tower for
the removal of mercury particulate.
The use of a venturi scrubber for removal
of mercury particulate has been discussed in
Section 3.1.3.1.3. It is estimated that a
control system using an indirect cooler to
cool the condenser gas stream to 55°F fol-
lowed by a venturi scrubber could reduce
emissions to 1.7 pounds per day. The capital
cost for this system would be $86,000. The
annual operating cost is estimated to be
$26,000. These costs assume a capital cost of
$12,000 for a standard venturi scrubber.
A second type of venturi scrubber, which
has the ability to cool the gas stream and
scrub simultaneously, can be purchased as a
complete unit for approximately $14,000 for
a facility treating 100 tons of ore per day.
The only additional costs that would be
incurred are for cooling the scrubbing water
and for two adequately sized settling tanks
for the scrubber discharge and the bleed
stream. The manufacturer of this system
considers that the packaging of this system
will allow a minimal installation cost. The
capital requirement for the complete system
is estimated to be $30,000. No information is
presently available to enable estimation of the
degree of cooling that can be attained.
3.1.4.3 Treated Activated Carbon
The use of treated activated carbon as a
control technique has been discussed in
Section 3.1.3.2. To be effective, this system
must be preceded by some type of mist
elimination device. A carbon bed system
preceded by a venturi scrubber could reduce
the stack gas emissions to 1.8 pounds per day
for a 100-ton-per-day facility. (This estimate
is based on a conservative carbon bed vapor
collection efficiency of 90 percent.) The
expected capital and annual operating costs
are estimated to be $66,000 and $20,000.
More complete cost data are presented in
Table A-5 of the Appendix.
3.1.4.4 Summary
Several control techniques and their re-
spective costs have been discussed in Sections
3.1.3 and 3.1.4. Table 3-4 presents a summary
3-16
-------
Table 3-4. SUMMARY OF PRIMARY MERCURY CONTROL TECHNIQUE COSTS
AND EXPECTED EMISSIONS IN CONDENSER STACK GAS STREAM
FOR 100-TON/DAY FACILITY
Control system
Cooling and mist elimination
(Section 3.1.4.1 and
Tables A-4 and A-5)
Wet scrubbing (Section 3.1.4.2
and Tables A-4 and A-5)
Treated activated carbon
(Section 3.1.4.3 and
Tables A4 and A-5)
Capital cost, $
76,000 to 108,000
86,000
66,000
Annual operating
cost, $
23,000 to 32,000
26,000
20,000
Expected
emissions, Ib/day
5.8 to 1.7
1.7
1.8
of both capital and operating costs for these
techniques together with expected emissions.
The capital costs of specific control systems
are itemized in Table A-5.
All costs have been based on a facility of
100 tons of ore per day capacity. Cost
estimates can be adjusted for other capacities
by using the following equation:
CA"C100
0.6
100 tons/day
where: CA = applicable control costs, equip-
ment or capital
CJQQ = control cost for a 100-ton-
per-day facility, equipment or
capital
P* = applicable extraction facility
capacity in tons of ore per
day.
3.1.5 Development of New Technology
This section contains information con-
cerning processes and control techniques that
are not considered to be "state-of-the-art" in
the primary mercury extraction industry.
Included in this section are processes and
techniques that have been tested at the
research and/or development levels but that
have not been sufficiently demonstrated at
full scale.
3.1.5.1 Hydrometallurgical Processing
3.1.5.1.1 Process Description - Over 90 years
ago, Volhard reported that an aqueous solu-
tion of sodium sulfide could be used for
dissolving cinnabar. Until recently, however,
the only hydrometallurgical technique in use
was that developed in 1915 for the recovery
of mercury from amalgams. In the late
1950's, the U.S. Bureau of Mines conducted
considerable research on the use of sodium
sulfide and sodium sulfide-sodium hydroxide
solutions for leaching mercury ores and con-
centrates.25 ~27 These investigations demon-
strated that cinnabar could be dissolved in
these solutions.
The great variations in composition of
ores, the cost of grinding ores finely enough
for effective leaching, the simplicity and
efficiency of pyrometallurgical processing,
and the cost of reagents have generally pre-
cluded the consideration of hydrometal-
lurgical processes for the recovery of mercury.
However, if flotation is used to concentrate
the ore, the cost of alkaline sulfide leaching
followed by electrolytic precipitation of the
mercury has been estimated to be about the
same as the cost of pyrometallurgical treat-
ment.
In 1970, U.S. Bureau of Mines investi-
gators reported that recoveries of 90 to 99
percent could be attained by electro-
3-17
-------
CLASSIFIER
CRUSHED ORE
DIGESTION
AGITATOR
'TAILS
MERCURY
Figure 3-9. Conceptual hydrometallurgical plant layout and flow diagram based on
laboratory data.28
oxidation of finely ground mercury ore in a
brine solution to form soluble mercuric salts,
followed by precipitation of the mercury with
an active metal dust such as zinc, iron, or
aluminum.28 The process is applicable to
low-grade mercury ores, and beneficiation is
not necessary. The most important para-
meters in the electro-oxidation process are
temperature, salt concentration, current
density, type of electrodes, electrode spacing,
treatment rate (amperes per ton of ore), and
particle size of ore. The ore must be ground
to 35 mesh or finer. In typical laboratory
experiments, 1 to 7 hours of electrolysis was
required at a 35 percent pulp density in a
brine solution containing 4 to 20 percent
sodium chloride.28 Power consumption
ranged from 10 to 50 kilowatt-hours per ton
of dry ore.
The tailings are discharged into a settling
pond. The concentration of mercury in the
tails ore is approximately 0.1 pound per ton
of solid tailing.13 The concentration of mer-
cury in the tails solution is approximately 1
part per million.13 Figure 3-9 shows a con-
ceptual plant layout and flow diagram based
on laboratory data.
The operating cost for the electro-
oxidation process is estimated to be $2 to $3
per ton compared to $5 to $8 per ton for a
furnacing operation.13 A feasibility study of
this process conducted by a large engineering
firm indicated that it was favorable for
low-grade ores in the 2- to 3-pound mercury
per ton range.13 The U.S. Bureau of Mines is
conducting pilot mill experiments in a 100- to
200-pound-per-hour pilot plant to quantify
power and reagent requirements. A report is
being prepared to give final details of this
research project.
3.1.5.1.2 Emissions — The use of hydrometal-
lurgical techniques for treating mercury ores
would minimize or eliminate the emission of
mercury to the atmosphere if the processing
3-18
-------
Table 3-5. CHLORINE CAPACITY AND PRODUCTION METHODS
IN THE UNITED STATES31'32
Year
1946
1956
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
Capacity,
short tons/day
4,012
10,300
14,697
15,503
16,404
17,245
18,939
21,216
23,238
25,124
28,276
29,131a
Percent of total installed capacity
Diaphragm
cells
88.6
81.6
76.2
74.1
72.2
71.2
69.7
69.8
68.1
69.2
69.6
69.8
Mercury
cells
4.3
12.4
18.5
20.8
23.0
24.2
26.5
26.7
28.6
27.9
27.2
27.2
Fused salt and
nonelectrolytic
7.1
6.0
5.3
5.1
4.8
4.6
3.8
3.5
3.3
2.9
3.2
3.0
a Preliminary.
plant were properly designed. To avoid the
emission of finely divided ore particulates, the
crushing and grinding circuit must provide
maximum control of the dust. It is probable
that soluble mercury compounds would
remain in the tailings from the leaching
operation even with careful countercurrent
washing. The potential water pollution pro-
blem created could be minimized or elimi-
nated by a properly designed tailings disposal
system.
3.1.5.2 Sulfuric Acid Scrubber
A foreign concern has used a sulfuric acid
scrubber to remove mercury and selenium
from its nonferrous smelter sulfur dioxide gas
stream.29-30 This concern states that this
system can reduce the mercury concentration
from about 60 milligrams per cubic meter to
about 200 micrograms per cubic meter in a
29,000 cubic-foot-per-minute gas stream. The
cost of this sulfuric acid scrubber is approx-
imately $500,000.
3.2 CHLOR-ALKALI PRODUCTION
3.2.1 History
Chlorine is produced almost entirely by
electrolytic methods from fused chlorides or
aqueous solutions of alkali-metal chlorides. In
the electrolysis of an aqueous solution of
potassium or sodium chloride, chlorine is
produced at the anode while hydrogen and
either potassium hydroxide or sodium
hydroxide are produced as a result of pro-
cesses occurring at the cathode. This requires
that the anode and cathode products be kept
entirely separate. Consequently, many
ingenious cell designs have been developed
and refined; all of these have been variations
either on the diaphragm cell or on a cell
which employs mercury metal as an inter-
mediate cathode. Historically, these two pro-
cesses were developed more or less in parallel.
In the United States, the mercury process was
an early leader but diminished to less than 5
percent of the installed chlorine capacity by
1946. From 1946 to 1968, the use of the
mercury cell grew toward 28 percent of the
total installed U.S. chlorine capacity. Since
1968, there has been a slight negative trend in
the use of mercury cells. In Europe and most
other parts of the world, the use of the
mercury cell predominates. Detailed data on
chlorine production by various production
techniques are given in Table 3-5 and shown
graphically in Figures 3-10 and 3-11. A listing
of the operating mercury-cell plants in the
U.S. is presented in Appendix A.4.
3-19
-------
90
-
LU eg
60
50
40
20
10
1946 1950 1960
YEAR
1970
Figure 3-10. Percentage of total installed
U.S. chlorine capacity for diaphragm and
mercury cells.31, 32
from the electrolyzer to the main brine
treatment section of the plant; it may also be
sent through a dechlorination step.
The chlorine gas product formed at the
anode is discharged from the electrolyzer for
further treatment. After cooling, the gas is
dried by scrubbing with concentrated sulfuric
acid. The spent acid from this drying step
contains most of the trace amounts of mer-
cury carried along by the wet chlorine gas.
After compression, the dry chlorine may be
used directly or may be subjected to a
liquefaction step.
The sodium amalgam flows continuously
from the electrolyzer through the outlet
end-box to the decomposer; there it is the
anode to a short-circuited graphite or metal
cathode in an electrolyte of sodium hydrox-
ide solution. The outlet end-box is a recep-
tacle, which is placed on the outlet of the
electrolyzer to provide a convenient means
for keeping the sodium amalgam covered with
an aqueous layer. The outlet end-box also
permits the physical separation of these two
streams. Purified water is fed continuously to
the decomposer. This water reacts with the
3.2.2 Emissions and Process Description
3.2.2.1 Mercury-Cell Process
3.2.2.1.1 Process Description — Since the
treatment of potassium chloride brines is
essentially analogous to that of sodium chlo-
ride brines, the latter will serve as the basis for
the following discussion. The basic process
flow sheet for the production of chlorine and
caustic soda is shown in Figure 3-12.
Purified and nearly saturated brine is fed
continuously from the main brine treatment
section into the inlet end of the electrolyzer,
where it flows between a stationary graphite
(or metal) anode and a flowing mercury
cathode. The inlet end-box is a receptacle
which is placed on the inlet end of the
electrolyzer to provide a convenient connec-
tion for the stripped mercury as it returns
from the decomposer (denuder). It also serves
to keep the incoming mercury covered with
an aqueous layer. The spent brine is recycled
o
2
-------
BASIC TREATMENT CHEMICALS
(SODA ASH, CAUSTIC LIME,
ACID, CaCL2, ETC.)
SOLID
NaCL FEED
CHLORINE
PRODUCT
CHLORINE
OTHER
1
MAIN
STREAM
RECYCLE
I
BRINE
DECHLORINATOR
SPENT BRINE
MAIN BRINE
SATURATION,
PURIFICATION, AND
FILTRATION
TREATED
BRINE
INLET
END-BOX-
END-BOX
VENTILATION SYSTEM'
AQUEOUS
LAYER
STRIPPED
AMALGAM
END-BOX
VENTILATION SYSTEM
1
COOLING,
DRYING,
COMPRESSION, AND
LIQUEFACTION
ELECTROLYZER
i
• OUTLET END-BOX
^.END-BOX
VENTILATION SYSTEM
AQUEOUS LAYER
WATER COLLECTION
SYSTEM
END-BOX
' VENTILATION SYSTEM
HYDROGEN GAS
BYPRODUCT
Hg PUMP
AQUEOUS LAYER
PROPRIETARY TREATMENT CHEMICALS INCLUDE PRECIPITATORS,
FLOCCULANTS, POLYELECTROLYTES, AND SIMILAR MATERIALS
DECOMPOSER
(DENUDER)
AMALGAM
I
CAUSTIC SODA
SOLUTION
PRODUCT
-CAUSTIC
SODA
Figure 3-12. Basic flow diagram for chlor-alkali
mercury-cell operation.8
3-21
-------
sodium amalgam and produces sodium hy-
droxide solution as well as by-product hydro-
gen gas. About 800 to 850 standard cubic feet
per minute of hydrogen is produced for each
100 tons of daily chlorine capacity: The
high-purity caustic soda generally leaves the
decomposer at a concentration of 50 percent
sodium hydroxide by weight.
The caustic soda solution from the
decomposer is usually sent to a filtration unit.
The solid waste material from the filter may
be processed to recover the mercury content
in a retort. The mercury recovered from this
retort operation is returned to the cell for
reuse.
Filtered caustic solution at a concentra-
tion of 50 percent by weight may be further
concentrated by evaporation to a 73 percent
by weight sodium hydroxide product. In
some instances, this material is heated to drive
off the remaining water in order to produce
anhydrous pellets or flakes of solid sodium
hydroxide.
The by-product hydrogen gas from the
decomposer may be vented to the atmo-
sphere, burned as fuel, or used as a feed
material for subsequent processing. This pro-
cess stream is saturated with mercury vapor.
In order to reduce mercury vapor emis-
sions to the cell room, an end-box ventilation
system applies suction to various sections of
the mercury cell operation. Such sections are
usually one or both end-boxes, the mercury
pump sump, and their water collection
systems.
A large quantity of ventilation air main-
tains the temperature and mercury vapor
concentration of the cell room at allowable
levels.
3.2.2.1.2 Emissions — In terms of direct
emissions of mercury to the atmosphere, the
major sources are:
1. The hydrogen by-product stream.
2. The end-box ventilation system.
3. The cell room ventilation air.
The hydrogen by-product stream leaving
the decomposer is saturated with mercury
vapor (2.3 grams of mercury vapor per cubic
meter at 210°F). If this stream were directly
discharged to the atmosphere without prior
cooling, an estimated 220 pounds of mercury
would be emitted for each 100 tons of
chlorine produced.
The minimum treatment known to be
used consists of cooling the hydrogen stream
to 110°F and partial mist elimination. For
hydrogen saturated with mercury vapor at
this temperature, the daily loss from the
stream following treatment would be 6.8
pounds of mercury vapor per 100 tons of
chlorine produced. The entrainment of con-
densed mercury in the hydrogen stream will
result in emissions in excess of the above
amount, which was calculated from known
values of the vapor pressure of mercury. At
least 86 percent of this additional loss of
mercury as mist can be eliminated in a
properly designed mercury entrainment sepa-
rator.18'19 The calculated combined emission
of mercury vapor and mercury mist, after
minimum treatment has occurred, is esti-
mated to be 50 pounds for each 100 tons of
chlorine produced.
Mercury and mercury compounds are
collected from the end-boxes, the mercury
pump sumps, and their water collection sys-
tems by the end-box ventilation system.
Mercury emissions from this system are
dependent on the percentage of mercury
saturation and the volumetric flow rate. The
volume, the degree of mercury saturation,
and, in turn, the resulting emissions from the
end-box ventilation system inherently depend
upon the age of the plant, upon the type and
specific configuration of the cells, end-boxes,
and decomposers, and upon the standard
operating procedures employed at a particular
location. For some new cell modifications
currently being installed, the volume of the
end-box ventilation system is small compared
with that of the hydrogen stream. In most
cases, however, the volumetric flow rate of
the end-box ventilation system may approach
or exceed the flow rate of the hydrogen
stream.8'19'33 For the purposes of this dis-
cussion and subsequent emission and cost
3-22
-------
calculations, the volumetric flow rate of this
system will be assumed to be equal to that of
the hydrogen stream. Preliminary results of
Environmental Protection Agency testing
indicate that the mercury emissions from an
untreated or inadequately treated end-box
ventilation system range from 2 to 15 pounds
for each 100 tons of chlorine pro-
duced.19'33-34
The cell room ventilation system serves
two major purposes. The primary function of
this air stream is to cool the cell room
environment, but it also provides a means of
reducing the cell room concentration to
within the recommended Threshold Limit
Value (TLV). The Threshold limit Value of
the American Conference of Governmental
Industrial Hygienists has recently been re-
duced from 100 micrograms per cubic meter
to 50 micrograms per cubic meter.
The volumetric flow rate of the cell
room ventilation stream varies from 100,000
to 1,000,000 cubic feet per minute for each
100 tons of daily chlorine capacity. A higher
flow of air may be needed in old plants,
where high mercury concentrations are more
likely to exist. On the basis of data obtained
from operating plants, it has been estimated
that mercury emissions from the cell room
ventilation system vary from 0.5 to 5.0
pounds per day per 100 tons of daily chlorine
., 8,19
capacity.
These emissions were calculated using
the TLV of 100 micrograms per cubic meter.
Figure 3-13 illustrates the daily mercury
emissions as a function of cell room volu-
metric flow rate when a plant maintains its
cell room air at a TLV of 50 micrograms per
cubic meter and at 80 percent of this TLV, or
40 micrograms per cubic meter.
Mercury enters the cell room atmosphere
as a result of a number of operations or
conditions, a few of which are listed below:
1. End-box sampling.
2. Removal of mercury butter from
the end-boxes.
3. Cell maintenance and rebuilding
operations.
4. Other maintenance work which ex-
poses the internal surfaces of pipes
and equipment.
5. Accidental spills of mercury.
6. Leaks from cells and mercury
pumps.
7. Cell failure and other unusual cir-
cumstances.
The number and variety of sources of mer-
cury emissions to the cell room air clearly
indicate that careful plant operation and good
housekeeping are essential in order to mini-
mize the amount of mercury emitted into the
cell room air. In Section 3.2.4.6, a list of
housekeeping practices for minimizing mer-
cury emissions to the cell room will be
recommended.
3.2.2.2 Diaphragm-Cell Process
3.2.2.2.1 Process Description — Chlorine and
caustic soda can be produced by the elec-
trolysis of brine in a diaphragm cell. In this
process, an asbestos diaphragm separates the
anode from the cathode. Chlorine gas is
formed at the anode and hydrogen gas and
caustic are formed at the cathode. One
disadvantage of this process is that the caustic
is of a lower grade than that produced by the
1000
M 800
600
o
a:
400
o
> 200
OF TOLERANCE
LIMIT VALUE
TOLERANCE LIMIT
VALUE, 50
12345
MERCURY EMISSIONS, Ib/day
Figure 3-13. Mercury emissions in cell room
ventilation air.
3-23
-------
mercury cell and must be concentrated and
purified for some uses.
3.2.2.2.2 Emissions - The diaphragm cell
produces no mercury emissions.
3.2.3 Emission Reduction Resulting from
Changing to Diaphragm-Cell Process
Reduction of mercury emissions can be
accomplished by adding a control device to
remove mercury from a gas stream or by
converting to a process that inherently pro-
duces lower emissions of mercury. This sec-
tion will discuss process changes; Section
3.2.4 will discuss emission reductions result-
ing from the application of control tech-
niques.
Replacement of the mercury-cell chlor-
alkali plant with a diaphragm-cell plant is an
effective, but expensive, method for eliminat-
ing all mercury emissions from the manufac-
ture of chlorine and sodium hydroxide.
The diaphragm-cell chlor-alkali plant
produces chlorine at a cost usually slightly
less than that of the mercury-cell chlor-alkali
plant, but the sodium hydroxide produced by
the diaphragm cell is only about 11 percent
sodium hydroxide by weight and is saturated
with sodium chloride. The sodium hydroxide
produced by the mercury cell is about 50
percent sodium hydroxide by weight. In order
for the caustic produced by the diaphragm
cell to be competitive with the mercury-cell
caustic, it must be upgraded. The process for
upgrading weak caustic from diaphragm-cell
chlor-alkali plants, while somewhat expensive
to install and operate, has been developed for
some time and has recently been improved.
The 11 percent sodium hydroxide from dia-
phragm plants can be concentrated to 50
percent in multiple-effect nickel evaporators
to obtain a 50 percent sodium hydroxide
solution with 1 percent sodium chloride
content.36 This material is somewhat turbid
when compared to 50 percent sodium hy-
droxide from mercury-cell chlor-alkali plants.
Impurities, such as sodium chloride and
trace metals, are often contained in the
concentrated diaphragm-cell caustic. Since
these impurities cannot be tolerated by cer-
tain industries requiring high-purity caustic
for processing, a proprietary process, which
reduces these impurities, has been developed
and is available for licensing. The process,
known as the DH process, utilizes anhydrous
ammonia in a liquid-liquid extraction opera-
tion to reduce the sodium chloride concentra-
tion in 50 percent diaphragm-cell caustic
solution from 1.0 to 0.025 percent.35 This
grade of 50 percent diaphragm-cell caustic is
nearly as pure as the 50 percent caustic
produced from mercury cells and should be
able to meet all or nearly all of the market for
a high-grade 50 percent sodium hydroxide.
The cost of producing the diaphragm-cell
caustic is somewhat higher than that of the
mercury-cell caustic. The costs of conversion
will be discussed in Section 3.2.5.1.
3.2.4 Control Techniques
Various techniques have been developed
for the control of mercury emissions from
chlor-alkali operations. These techniques ap-
ply to the by-product hydrogen stream and
the end-box ventilation system.
3.2.4.1 Cooling and Condensing
3.2.4.1.1 Hydrogen Gas Stream - Hydrogen
leaves the decomposer at 200° to 260°F and
passes into the primary cooler, where ambient
water is normally used in a shell-and-tube heat
exchanger to cool this stream to 90° to
110°F (Figure 3-14). The primary cooler
usually has a device, known as a mercury
knockout drum, for separating condensed
mercury from the hydrogen stream. After
passing through this initial drum, the hydro-
gen stream may be subjected to additional
cooling in a secondary cooler that uses chilled
water or brine as the cooling medium. The
condensed mercury from this secondary cool-
ing step is partially removed in a second
knockout drum. The temperature of the
hydrogen stream discharged from the second-
ary cooler ranges from 37° to 55°F. Because
of the rapid variation of mercury saturation
3-24
-------
H2 FROM DECOMPOSER
850 scfm AT 210°F
r—T
HgOUT
I
BASE SYSTEM
HgOUT
DIRECT OR INDIRECT
PRIMARY COOLER
110° F
KNOCKOUT
DRUM
HgOUT
_J
DIRECT OR INDIRECT
SECONDARY COOLER
HgOUT
45° F
KNOCKOUT
DRUM
SEAL TANK; CHECK VALVE
STOPS H2 BACK FLOW
H2 STREAM TO ADDITIONAL Hg CONTROL EQUIPMENT
OR H2 DISPOSAL
Figure 3-14. Cooling and condensing of
hydrogen stream.
concentration with temperature, the hydro-
gen stream temperature should be decreased
as much as possible in order to condense the
largest amount of mercury vapor.
A large percentage of the condensed
mercury may remain in the gas stream as a
mercury mist that is difficult to efficiently
separate from the hydrogen stream. One
approach to resolving this problem is the use
of a direct-contact cooler with chilled water
or brine instead of a shell-and-tube heat
exchanger. The chilled aqueous medium is
often in a closed loop with a shell-and-tube
heat exchanger that uses mechanical refrigera-
tion. The bleed-off liquor from the direct-
contact cooling system is sometimes treated
with chemicals such as sodium hydrosulfide
to remove the mercury; the clean liquor is
then recycled back into the closed loop. An
alternate method of treatment is to pump the
bleed-off stream into a waste-water settling
pond. The use of a direct-contact cooler has
the disadvantage of requiring a water treat-
ment system.
3.2.4.1.2 End-box Ventilation System - The
temperature of the end-box ventilation air
ranges from 160° to 180°F, and the air may
be 10 to 50 percent saturated with mercury
and nearly saturated with water. For calcula-
tion purposes, however, it is assumed that
both the end-box stream and the hydrogen
stream are saturated with mercury vapor. The
maximum mercury content of the process
streams is therefore considered. Some end-
box ventilation gases contain enough chlorine
to form mercuric chloride, which remains in
the gas stream as particulate matter. The
techniques used for cooling the end-box
ventilation air are similar to the methods used
to cool the hydrogen stream. However, the
presence of mercuric chloride in this gas
stream has led to a greater usage of direct-
contact coolers for cooling and particulate
matter removal. The high water content of
this system has limited the temperature to
which this gas stream can be cooled due to
the formation of ice crystals. The presence of
chlorine and mercury salt contaminants has
also required the use of more corrosion-
resistant construction materials such as
titanium and titanium-clad steel.
A technique used in industry to treat
end-box ventilation air is to cool the air
stream with a direct-contact packed tower
employing water as the coolant. This tower is
followed by an indirect cooler, which uses
40° to 68°F cooling water. The resulting
entrained particulate mercury is removed by a
mist eliminator.33 A mercury emission rate of
1 pound per day from the end-box ventilation
system for each 100 tons per day of chlorine
capacity has been estimated for such a
system.
3-25
-------
A considerable reduction in the capital
and operating costs of the end-box system can
be obtained by the reduction of the end-box
ventilation flow rate. This reduction can be
accomplished by the installation of leak-tight
covers on all cell end-boxes and by the
replacement of submerged mercury pumps
with in-line pumps.
3.2.4.2 Mist Elimination
There are two basic types of mist elimi-
nators that are commercially available to
remove mercury mist from gaseous streams.
One of these, which consists of fiber pads to
remove entrained mist by mechanisms of
impaction, interception, and Brownian move-
ment, is discussed in Section 3.1.3.1.2 and
shown in Figure 3-7, When this type of
eliminator is used for mercury control in an
end-box ventilation gas stream, backwashing
with water or a strong reducing agent may
become necessary to prevent plugging.
Another type of mist eliminator which is
being used in at least one application utilizes a
converging-diverging nozzle arranged so that
the gases being cleaned follow a curved path
and are acted upon by high centrifugal forces
in the throat area (Figure 3-15).36 These
forces are reported to cause the coalescence
PURE
GAS OUT
of mercury mist and entrapment of sub-
micron mercury mist upon the upper wall of
the divergent section. Separated particles are
washed away from the walls by sprayed
liquid. This mist eliminator is reported to
have an efficiency comparable to that of the
fiber pad type of mist eliminator. The entire
unit can be made of plastic or special alloys
which resist amalgamation. Figure 3-16 pre-
sents a typical control system utilizing a mist
elimination device which is applicable to
either the hydrogen or the end-box streams.
3.2.4.3 Chemical Scrubbing Techniques
3.2.4.3.1 Depleted Brine Scrubbing System -
Depleted brine scrubbing techniques have
been applied for mercury removal from
hydrogen and end-box ventilation gas streams
in only a few instances. The depleted brine
scrubbing technique uses the brine discharged
from a chlorine cell as a scrubbing liquor. This
depleted brine contains about 250 grams per
liter of sodium chloride and 0.6 to 0.9 gram
per liter of available chlorine; it has a pH of 2
to 4.37 This solution is used as the scrubbing
medium in a sieve plate tower or in a
packed-bed scrubber. Upon contact with the
brine scrubbing solution, mercury vapor and
mist' form soluble mercury complexes. The
IMPURE
GAS IN
MERCURY
Figure 3-15. Converging-diverging nozzle mist eliminator.36
3-26
-------
HgOUT
H2 STREAM FROM DECOMPOSER OR
END-BOX VENTILATION STREAM
850 scfm AT 180° TO 210° F
DIRECT OR INDIRECT
PRIMARY COOLER
I
110° F
HgOUT
KNOCKOUT
DRUM
HgOUT
DIRECT OR INDIRECT
SECONDARY COOLER
HgOUT
45° F
J_
KNOCKOUT
DRUM
HgOUT
MIST ELIMINATOR
GAS STREAM TO ADDITIONAL Hg
CONTROL EQUIPMENT OR DISPOSAL
Figure 3-16.
elimination.
Cooling, condensing, and mist
mercury is subsequently recovered by elec-
trolysis when the scrubbing medium is re-
turned to the mercury cell. Figure 3-17
presents a simplified flow sheet for a depleted
brine scrubbing system.
One application of a depleted brine
scrubbing system (System A) has been used
since the early 1960's for the removal of
mercury from the hydrogen stream. The
mercury concentration of the treated hydro-
gen has been reported to be approximately 85
micrograms per cubic meter. Mercury losses
from the treated hydrogen stream would,
therefore, be less than 0.01 pound per day on
a 100-ton-per-day chlorine basis. In this sys-
tem, the hydrogen gas is discharged from the
decomposer at 180°F. It is then pumped into
a direct-contact water scrubber where it is
cooled to 100°F. After partial separation of
the entrained mercury mist with a knockout
drum, the hydrogen is cooled to 55°F by a
direct-contact cooler using chilled brine in a
closed-loop system. The cooled hydrogen
H? STREAM FROM DECOMPOSER
OR END-BOX VENTILATION STREAM
850 scfm AT 180° TO 210° F
HgOUT
DIRECT OR INDIRECT
PRIMARY COOLER
HgOUT
1
110° F
J
KNOCKOUT
DRUM
HgOUT
DIRECT OR INDIRECT
SECONDARY COOLER
45° F
HgOUT
MIST ELIMINATOR
SCRUBBING
LIQUOR
DEPLETED
BRINE
SCRUBBER
DILUTE
CAUSTIC
SCRUBBING
LIQUOR
DILUTE
CAUSTIC
SCRUBBER
T
DILUTE
CAUSTIC
GAS STREAM TO DISPOSAL OR USAGE
Figure 3-17.
system.
Depleted brine scrubbing
3-27
-------
then flows through a mist eliminator and a
depleted brine scrubbing tower. A final alka-
line scrubber is used to remove entrained
chlorine and acid.
A second application of depleted brine
scrubbing utilizes a system similar to the one
discussed above, but deletes the alkaline
scrubber (System B). With this system, mer-
cury emissions from the combined process
streams are reported to be about 0.6 to 0.9
pound of mercury per day on a 100-ton-per-
day chlorine basis.
3.2.4.3.2 Hypochlorite Scrubbing System - A
second type of chemical scrubbing technique,
which has been recently developed, uses a
sodium hypochlorite solution as a scrubbing
liquor. In one application of this technique,
the scrubbing liquor consisted of equal molar
amounts of sodium chloride and sodium
hypochlorite. The scrubbing medium report-
edly required a narrow pH control for opti-
mum mercury removal efficiency. This nar-
row range of pH was difficult to maintain,
and as a result, the sodium hypochlorite
scrubber was converted to a depleted brine
scrubber.
A second application of a sodium hypo-
chlorite scrubbing technique, which is re-
ported to have solved the problem of pH
control, is available for licensing.38 This
system employs a dilute solution of sodium
hypochlorite with a large excess of sodium
chloride over the stoichiometric quantity. The
mercury removal efficiency of this system is
maintained over a wide enough pH range to
make control possible.38 This system has
been employed successfully at two sites to
date. A mercury collection efficiency of 95 to
99 percent has been reported.39
Figure 3-18 presents a general schematic
flow sheet of the hydrogen stream for a
control system using a hypochlorite scrubber.
It is estimated that mercury emissions for the
combined hydrogen and end-box streams
would range from 0.2 to 0.8 pound per day
for a 100-ton-per-day plant.
HgOUT
HgOUT
HgOUT
HgOUT
SCRUBBING
LIQUOR
Hg, ENTRAINED
SCRUBBING
SOLUTION
H? STREAM FROM DECOMPOSER
850 scfm AT 210° F
DIRECT OR INDIRECT
PRIMARY COOLER
I
110° F
1
KNOCKOUT
DRUM
DIRECT OR INDIRECT
SECONDARY COOLER
I
45° F
I
KNOCKOUT
DRUM
SCRUBBING
LIQUOR
HYPOCHLORITE
SCRUBBER
MIST ELIMINATOR
I
GAS STREAM TO DISPOSAL OR USAGE
Figure 3-18. Hypochlorite scrubbing system.
3.2.4.4 Treated Activated Carbon
Control systems containing either sulfur-
or iodine-impregnated activated carbon are
being utilized by several mercury-cell chlor-
-alkali plants for reduction of the mercury
concentration in the hydrogen stream. In
3-28
-------
these systems, the mercury vapor is adsorbed
by the carbon and chemically reacts with the
sulfur or iodine to form mercury compounds.
If properly designed, this technique can
reduce the mercury concentration in the
hydrogen stream to 5 to 10 micrograms per
cubic meter.23 The hydrogen at the inlet to
the treated activated carbon bed usually has
had 90 percent of its mercury content
removed by primary and secondary cooling
followed by efficient mist elimination. Figure
3-19 is a schematic flow diagram for a typical
hydrogen stream employing treated activated
carbon. The treated activated carbon can
adsorb from 10 to 20 percent of its weight in
mercury before it requires replacement.
Destructive distillation of saturated activated
carbon in retorts appears practical for recover-
ing the adsorbed mercury.
This technique should also be applicable
to mercury removal from the end-box ventila-
tion system.
3.2.4.5 Molecular Sieve
The molecular sieve control technique
utilizes a sieve-adsorbent blend to adsorb the
mercury contained in gas streams. A molecu-
lar sieve system currently available for treat-
ment of the hydrogen gas stream of chlor-
alkali plants is the PuraSiv—Hg System. The
designer of this system guarantees a reduction
of the mercury concentration level to 0.50
milligram per cubic meter.36 This concentra-
tion corresponds to a hydrogen stream emis-
sion of 0.04 pound of mercury per 100 tons
of daily chlorine capacity.
Figure 3-20 illustrates a simplified
PuraSiv-Hg System for the hydrogen stream
of a mercury-cell chlor-alkali plant. 3
Hydrogen, laden with mercury vapor and
mist, is passed through a secondary cooler and
a mist eliminator. The gas stream then passes
through one of two adsorption beds, both of
which contain a proprietary sieve-adsorbent
blend. Eighty to 90 percent of the treated
hydrogen gas, containing 0.50 milligram per
cubic meter mercury or less, is vented to the
atmosphere, combusted in a burner system, or
used in a subsequent production operation. 36
The remainder of the controlled hydrogen
H2 STREAM FROM DECOMPOSER
850 scfm AT 210° F
HgOUT
PRIMARY
COOLER
I
110° F
COMPRESSOR,
20 psi
SEAL TANK,
PREVENTS)
H2 BACK FLOW
ACTIVATED CARBON
BED, 4000 Ib,
6 mo LIFE
-X-
BY-PASS ACTIVATED
CARBON BED
4000 Ib
CLEAN H2
Figure 3-19. Activated carbon bed system.
3-29
-------
HgOUT
HgOUT
HgOUT
HgOUT
H2 STREAM FROM DECOMPOSER
850 scfm AT 210<> F
PRIMARY COOLER
I
110° F
f
KNOCKOUT DRUM
RECYCLE/
REGENERATION STREAM
SECONDARY
COOLER
HgOUT
450 F
PETERSEN
SEPARATOR
ADSORPTION
BED
ADSORPTION
BED
600° F
H2 TO DISPOSAL
OR USAGE
Figure 3-20. Process flow sheet for a two-
bed molecular sieve system.
stream is heated to 600° F and used as a
recycle-regeneration stream for removing
entrapped mercury for the second adsorber
bed. After passing through the second
adsorption bed, this gas stream, with its high
mercury concentration, passes through a
cooler and is combined with the incoming
mercury-laden hydrogen stream from the
primary cooling section. The two adsorption
beds alternate in function; while one bed is
removing mercury from the gas stream, the
other bed is being regenerated.
Since the mercury adsorption characteris-
tics of the end-box ventilation gas stream are
similar to those of the hydrogen stream, the
molecular sieve system has potential applica-
tion for the control of mercury emission from
the end-box ventilation system. Special
materials of construction would be required
in most situations because of the more
corrosive nature, due to chlorine and mercury
salt contaminants, of this gas stream.
3.2.4.6 Housekeeping Practices
The following housekeeping practices for
minimizing the various mercury emissions
within the cell room are recommended.40
Adherence to these recommended practices
will result in a sizeable reduction of the
mercury vapor concentration in the venti-
lation effluent from the cell room.
1. Chlorine cells and end-box covers
should be installed, operated, and
maintained in a manner to minimize
leakage of mercury and mercury-
contaminated materials.
2. Daily inspection should be made by
operating personnel to detect leaks,
and immediate steps to stop the leaks
should be taken.
3. High housekeeping standards should
be enforced, and any spills of
mercury should be promptly cleaned
up either mechanically or chemically
or by other appropriate means. Each
cell room facility should have avail-
able and should employ a well-
defined procedure for handling these
situations.
4. Floor seams should be smoothed over
to minimize depressions and to
facilitate washing down the floors.
3-30
-------
5. All floors should be maintained in
good condition, free of cracking and
spalling, and should be regularly
inspected, cleaned, and, to the extent
practical, chemically decontami-
nated.
6. Gaskets on denuders and hydrogen
piping should be maintained in good
condition. Daily inspection should be
made to detect hydrogen leaks, and
prompt corrective action should be
taken. Covers on decomposers, end-
boxes, and mercury pump tanks
should be well maintained and kept
closed at all times except when
operation requires opening.
7. Precautions should be taken to avoid
all mercury spills when changing
graphite grids or balls in horizontal
decomposers or graphite packing in
vertical decomposers. Mercury-
contaminated graphite should be
stored in closed containers or under
water or chemically treated solutions
until it is processed for reuse or
disposed.
8. Where submerged pumps are used for
recycling mercury from the decom-
poser to the inlet of the chlorine cell,
the mercury should be covered with
an aqueous layer maintained at a
temperature below its boiling point.
9. Each submerged pump should have a
vapor outlet with a connection to the
end-box ventilation system. The
connection should be under a slight
negative pressure so that all vapors
flow into the end-box ventilation
system.
10. Unless vapor-tight covers are pro-
vided, end-boxes of both inlet and
outlet ends of chlorine cells should
be maintained under an aqueous
layer maintained at a temperature
below its boiling point.
11. End-boxes of cells should either be
maintained under a negative pressure
by a ventilation system or be
equipped with fixed covers which are
leak tight. The ventilation system or
end-box covers should be maintained
in good condition.
12. Any drips from hydrogen seal pots
and compressor seals should be
collected and confined for processing
to remove mercury, and these drips
should not be allowed to run on the
floor or in open trenches.
13. Solids and liquids collected from
back-flushing the filter used for alkali
metal hydroxide should be collected
in an enclosed system.
14. Impure amalgam removed from cells
and mercury recovered from process
systems should be stored in an
enclosed system.
15. Brine should not be purged to the
cell room floor. Headers or trenches
should be provided when it is
necessary to purge brine from the
process. Purged brine should be
returned to the system or sent to a
treating system for mercury removal.
16. A portable tank should be used to
collect any mercury spills during
maintenance procedures.
17. Good maintenance practice should be
followed when cleaning chlorine
cells. During cleaning, all cells should
have any mercury surface covered
continuously with an aqueous med-
ium. When the cells are disassembled
for overhaul maintenance, the bed
plate should be either decon-
taminated chemically or thoroughly
flushed with water.
18. Brine, alkali metal hydroxide, and
water-wash process lines and pumps
should be maintained in good con-
dition, and leaks should be mini-
mized. Leaks should be corrected
promptly, and in the interim, the
leaks should be collected in suitable
containers rather than allowed to
spill on floor areas.
3.2.5 Control Costs
Costs in this section are developed for a
model plant of 100 tons per day of chlorine
capacity. Many actual equipment costs were
received by communication with several
mercury-cell chlor-alkali plant operators and
with various vendors of equipment and
proprietary processes. The various capital
3-31
-------
costs are itemized and documented in the
Appendix, Section A.3, for specific systems
of control equipment. This section will
classify control systems and illustrate their
respective capital and annual costs. The
methods of cost estimation are presented in
Section A. 2.
3.2.5.1 Conversion to Diaphragm-Cell Chlor-
Alkali Plant
The technology and problems involved in
converting a mercury-cell chlor-alkali plant to
a diaphragm-cell chlor-alkali plant are dis-
cussed in Section 3.2.3. A recent paper
indicates that the conversion cost will be
approximately $12,000 per daily ton of
chlorine capacity;41 thus, about $1,200,000
of capital investment would be required for
the conversion of a plant producing 100 tons
of chlorine per day. Table A-9 in the
Appendix presents the conversion cost for a
100-ton-per-day plant. Estimates range from
$3,700,000 to $8,000,000 if the 11 percent
diaphragm cell sodium hydroxide is upgraded
and purified to the quality of the caustic
produced by the mercury cell.
3.2.5.2 Cooling and Mist Elimination
A control system utilizing primary
cooling and partial mist elimination by means
of a knockout drum is considered to be the
minimum existing technology practiced at
most domestic plants. This control system is
applicable to both the hydrogen and the
end-box ventilation streams. The capital cost
and annual operating cost for this system are
$49,000 and $15,000, respectively, if applied
to both streams. Expected emissions from the
combined hydrogen and end-box streams
when treated by this system are listed in
Table A-8, which contains emission and cost
data for several control systems. A complete
breakdown of each system's cost is presented
in Table A-7.
The costs of most heat exchanger
equipment for the hydrogen system were
calculated on the basis of stainless steel as the
construction material. Costs were adjusted for
the end-box ventilation system to reflect the
higher cost of titanium and titanium-clad
materials necessitated by the more corrosive
nature of this stream. Several metals are listed
below in order of increasing resistance to the
corrosion effects of the end-box ventilation
gases:
1. Mild steel.
2. Alloy 316 stainless steel.
3. Titanium-clad steel.
4. Solid titanium.
It has been estimated that the addition of
a secondary cooler, a knockout drum, and a
mist elimination device to the base system, as
illustrated in Figure 3-16, could reduce the
combined emission of the hydrogen and
end-box streams to 3.4 pounds per day for a
100-ton-per-day plant. The capital and annual
operating cost for this system would be
$202,000 and $60,000 for the treatment of
both process streams.
3.2.5.3 Chemical Scrubbing
Figure 3-17 illustrates a depleted brine
scrubbing system, which is applicable to both
the hydrogen and the end-box ventilation
stream. Table A-7 in the Appendix gives a
breakdown of the estimated costs for a plant
producing 100 tons per day of chlorine.
Estimated emissions and total system costs
for plants of 100, 250, and 5QO tons per day
are listed in Table A-8. System costs range
from $160,000 to $350,000 for a 100-ton-
per-day plant. Annual operating costs are
estimated to be $48,000 to $105,000.
A diagram of a hypochlorite scrubbing
system, which is applicable to both the
hydrogen and end-box stream, is presented in
Figure 3-18. Estimated emissions for a
100-ton-per-day plant are 0.2 to 0.8 pound
per day for both process streams. Capital cost
and annual operating costs for such a system
would be approximately $226,000* and
$68,000. More complete data on capital cost
and expected emissions are provided in Tables
A-6, A-7, and A-8.
3.2.5.4 Treated Activated Carbon
Figure 3-19 illustrates a control system
that has been successfully employed on all or
part of the hydrogen streams from several
domestic mercury-cell chlor-alkali plants.
Table A-7 in the Appendix lists the
breakdown for the costs of the various pieces
of equipment required. The total capital
investment for a plant producing 100 tons of
*Does not include licensing fee.
3-32
-------
chlorine per day is estimated to be $279,000;
the operating cost is estimated at $83,000 per
year. These costs are for the application of
treated activated carbon on both the hydro-
gen and end-box ventilation gas streams.
3.2.5.5 Molecular Sieve
The molecular sieve system, which is
illustrated in Figure 3-20, can be applied to
both the hydrogen and end-box stream. By
treating both streams, it is estimated that
total emissions could be reduced to 0.08
pound of mercury per 100 tons of daily
chlorine capacity. Total capital investment for
a 100-ton-per-day plant is estimated to be
$349,000. Annual operating cost is expected
to be $105,000.
3.2.5.6 Summary
The approximate costs of mercury
control systems and the associated pieces of
equipment have been developed and are
presented in this section and Section A.3.
Since a 100-ton-per-day chlorine capacity has
been used for all cost estimations, the
following equation is presented for developing
cost estimates for other capacity plants.
CA~C100
0.6
,100 tons/day
where: C^ = applicable control cost; equip-
ment, capital, or operating
CIQQ = control cost for 100-ton-per-
day plant; equipment, capitol,
or operating
PA = applicable plant chlorine
capacity in tons per day
Table 3-6 is a summary of control
systems costs. As a basis for comparison of
the cost of various control systems, the
capital requirement for the construction of a
grass-roots mercury-cell chlor-alkali plant is
utilized as a common denominator. An
estimate of this capital requirement is
$100,000 for each daily ton of chlorine
capacity, or $10,000,000 for a 100-ton-per-
day plant.
All costs that have been discussed thus far
have not taken one possible contingency into
consideration. This contingency is the addi-
tional cost of procuring proprietary equip-
ment, systems, or associated engineering.
3.2.6 Development of New Technology
3.2.6.1 Ion Exchange Process
A Japanese ion exchange process has been
licensed for domestic use for the removal of
mercury from both air and water. The
designer claims that this process will reduce
the mercury concentration in a gas stream to
10 micrograms per cubic meter.4 2 There are
no known domestic applications of this
process on either the hydrogen or the end-box
ventilation streams of mercury-cell chlor-
alkali plants.
3.2.6.2 Sulfuric Acid Scrubber
As previously discussed in Section
3.1.5.2, a foreign nonferrous smelting oper-
ation has employed a sulfuric acid scrubbing
system to remove mercury and selenium from
its sulfur dioxide gas stream.29'30 Current
information indicates that this method has
not been employed in any mercury-cell
chlor-alkali plant in the world for control of
mercury emissions from the hydrogen and
end-box ventilation gas streams.
3.3 REFERENCES FOR SECTION 3
1. Pennington, J.W. Mercury—A Materials
Survey. U.S. Department of Interior,
Bureau of Mines. Washington, D.C. Infor-
mation Circular 7941. 1959. p. 33.
2. Tinn, R.K. and W.F. Die trick. Mining and
Furnacing Mercury Ore at the New Idria
Mine, San Benito County, California. U.S.
Department of Interior, Bureau of Mines.
Washington, D.C. Information Circular
8033. 1961. p. 23-28.
3. Tickes, M.R. Mining, Processing, and
Costs—Idaho Almaden Mercury Mine,
Washington County, Idaho. U.S. Depart-
ment of Interior, Bureau of Mines. Wash-
ington, D.C. Information Circular 7800.
1957. p. 15-24.
4. Roy, S.L., Jr. Personal communications
from operators of a mercury roasting
company. June 3, 1971. OAP-SSPCP-
SDID,* U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
*Office of Air Programs, Stationary Source Pollution
Control Programs, Standards Development Implemen-
tation Division.
3-33
-------
Table 3-6. SUMMARY OF CHLOR-ALKALI CONTROL
COSTS AND EXPECTED EMISSIONS FOR COMBINED
HYDROGEN AND END-BOX VENTILATION STREAMS
FOR A 100-TON/DAY FACILITY
Control system
Diaphragm cell
conversion (Section
3.2.5.1, Table A-9)
Cooling and mist
elimination
(Section 3.2.5.2,
Figure 3-16, and
Tables A-7 and A-8)
Chemical scrubbing
(Section 3.2.5.3,
Figures 3-17 and
3-18, and Tables
A-7 and A-8)
Treated activated
carbon (Section
3.2.5.4, Figure
3-19, and Tables
A-7 and A-8)
Molecular sieve
(Section 3.2.5.5,
Figure 3-20, and
Tables A-7 and A-8)
Capital
cost, $
3,700,000 to
8,000,000
201,000
1 59,000 to
349,000
278,000
349,000
Annual
operating
cost, $
1 ,000,000 to
2,400,000
60,000
48,000 to
105,000
83,000
105,000
Annual
operating cost
as % of grass
roots plant cost
11.0 to 24.0
0.6
0.5 to 1.1
0.8
1.0
Estimated emissions
from combined
H2 and end-box
streams, Ib/day
0.0
3.4
0.02 to 0.6
0.3
0.1
5. Alpiser, F.M. Personal communications
from operators of a mercury roasting
company. May 5, 1971. OAP-SSPCP-
SDID,* U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
6. Roy, S.L., Jr. Personal communications
from Mr. S.M. Fopp, Gordon I. Gould
and Company, San Francisco, Calif. May
1972. OAP-SSPCP-SDID,* U.S. Environ-
mental Protection Agency, Research
Triangle Park, N.C.
7. Roy, S.L., Jr. Personal communications
from Mr. E.A. Lado, Nichols Engineering
and Research Corporation, New York,
N.Y. June 1972. OAP-SSPCP-SDID,*
U.S. Environmental Protection Agency,
Research Triangle Park, N.C.
Basis for National Emission Standards for
Mercury. Battelle Memorial Institute.
Columbus, Ohio. Environmental
Protection Agency Contract No. EHSD
71-33. July 1971. p. 11-13,50-53.
El Paso Natural Gas Mercury Mine,
Weiser, Idaho. Roy F. Weston, Inc., West-
chester, Penn. Environmental Protection
* Office of Air Programs, Stationary Source Pol-
lution Control Programs, Standards Development
Implementation Division.
3-34
-------
Agency Contract No. 70-132, Task Order
No. 2. November 19, 1971.
10. New Aim ad en Mining and Chemical Com-
pany Mercury Mine. New Almaden, Cal-
ifornia. Roy F. Weston, Inc., Westchester,
Penn. Environmental Protection Agency
Contract No. 70-132, Task Order No. 7.
June 1972.
11. Sonamo Mines. Inc., Mt. Jackson Mercury
Mine, Gurneville, California. Roy F.
Weston, Inc., Westchester, Penn. Environ-
mental Protection Agency Contract No.
CPA 70-132, Task Order No. 7. June
1972.
12. Osag, T. Personal communications from
Mr. F. Seeton, The Denver Equipment
Division, Joy Manufacturing Company,
Denver, Colo. May 1972. OAP-SSPCP-
SDID,* U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
13. Osag, T. Personal communications from
Dr. B.J. Schemer, U.S. Department of
Interior, Bureau of Mines, Reno Metal-
lurgy Research, Reno, Nev. May 1972.
OAP-SSPCP-SDID,* U.S. Environmental
Protection Agency, Research Triangle
Park, N.C.
14. Modern Mineral Processing Flow Sheets
(2nd Ed.). Denver, The Denver
Equipment Division, Joy Manufacturing
Company, 1962. p. 110-113.
15. Roy, S.L., Jr. Personal communications
from operators of a mercury roasting
company. December 16, 1971. OAP-
SSPCP-SDID,* U.S. Environmental Pro-
tection Agency, Research Triangle Park,
N.C.
16. Van Bernewitz, N.W. Occurrence and
Treatment of Mercury Ore at Small
Mines. U.S. Department of Interior,
Bureau of Mines. Washington, D.C. Infor-
mation Circular 6966. 1937. p. 16-24.
17. Brink, J.A., Jr., C.N. Dougald, T.R.
Metzger, and E.D. Kennedy. Preventing
the Production of Airborne Pollutants
from Chemical Processes. Environmental
Pollution Management. London. Decem-
ber 1971. p. 79-84.
18. B.F. Goodrich Chemical Company Chlor-
Alkali Plant, Calvert City, Kentucky. Roy
F. Weston, Inc., Westchester, Penn. Envir-
onmental Protection Agency Contract
No. CPA 70-132, Task Order No. 3. May
1972.
19. Georgia Pacific Chlor-Alkali Plant, Bel-
lingham, Washington. Roy F. Weston,
Inc., Westchester, Penn. Environmental
Protection Agency Contract No. CPA
70-132, Task Order No. 2. November
1971.
20. Thompson, G. Personal communications
from Battelle Memorial Institute, Colum-
bus, Ohio. November 1971.
OAP-SSPCP-SDID,* U.S. Environmental
Protection Agency, Research Triangle
Park, N.C.
21. Nicholson, L.V. and G. Brown. Environ-
mental Control at Cominco's Pinchi Lake
Operations. (Presented at Canadian In-
stitute of Mining and Metallurgy Annual
Meeting. Kamloops, British Columbia.
October 24, 1970.) p. 3.
22. Osag, T. Personal communications from
Mr. J. Gallagher, Koertrol Corporation,
Shutte and Koerting Company, Durham,
N.C. May 1972. OAP-SSPCP-SDID,* U.S.
Environmental Protection Agency, Re-
search Triangle Park, N.C.
23. Osag, T. Personal communications from
Mr. F. Cunniff, Pittsburgh Activated Car-
bon Division, Calgon Corporation, Pitts-
burgh, Penn. May 1972. OAP-SSPCP-
SDID,* U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
24. Imperato, N.F. Gas Scrubbers. Chem.
Eng. 75(22): 152-155, October 1968.
25. Erspamer, E.G. and R.K. Wells. Selective
Extraction of Mercury and Antimony
from Cinnabar—Stibnite Ore. U.S. De-
partment of Interior, Bureau of Mines.
Washington, D.C. Report of In-
vestigations 5243. 1956.
26. Town, J.W., R.F. Link, and W.A.
Stickney. Caustic Sulfite Leaching of
Mercury Products. U.S. Department of
Interior, Bureau of Mines. Washington,
D.C. Report of Investigations 5748.
1961.
27. Stickney, W.A. and J.W. Town. Hydro-
metallurgical Treatment of Mercury Ores.
* Office of Air Programs, Stationary Source Pol-
lution Control Programs, Standards Development
Implementation Division.
3-35
-------
(Presented at the Annual Meeting of the
American Institute of Mining Engineers,
March 1961.)
28. Scheiner, B.J., R.E. Lindstrom, D.E.
Shanks, and T.A. Henrie. Electrolytic
Oxidation of Cinnabar Ores for Mercury
Recovery. U.S. Department of- Interior,
Bureau of Mines. Washington, D.C.
Technical Progress Report 26. 1970. lip.
29. Rastas, J., E. Nyholm, and J. Kangas.
Mercury Recovery from SO2-Rich
Smelter Gases. Eng. and Mining J. p.
123-124, April 1971.
30. Kangas, J., E. Nyholm, and J. Rastas.
Smelter Gases Yield Mercury. Chem. Eng.
p. 55-57, September 1971.
31. MacMullin, R.B. Diaphragm Versus
Amalgam Cells for Chlorine-Caustic Pro-
duction. Chemical Industries. July 1947.
32. Chlorine Capacity, Chlorine Production
Route Percentages. New York, The
Chlorine Institute, Inc., 1956-1971.
33. Alpiser, P.M. Personal communications
from operators of a chlor-alkali company.
May 6, 1971. OAP-SSPCP-SDID,* U.S.
Environmental Protection Agency, Re-
search Triangle Park, N.C.
34. Diamond Shamrock Corporation Mercury
Cell Chlor-Alkali Plant, Delaware City,
Delaware. Roy F. Weston, Inc., West-
chester, Penn. Environmental Protection
Agency Contract No. CPA 70-132, Task
Order No. 3. June 1972.
35. Roy, S.L., Jr. Personal communications
from Mr. J. Clapperton, PPG Industries,
Inc., Pittsburgh, Penn. May 1972. OAP-
SSPCP-SDID,* U.S. Environmental Pro-
tection Agency, Research Triangle Park,
N.C.
36. Roy, S.L., Jr. Personal communications
from Mr. W. Miller, Union Carbide Cor-
poration, Tarrytown, N.Y. May 1972.
OAP-SSPCP-SDID,* U.S. Environmental
Protection Agency, Research Triangle
Park, N.C.
37. Roy, S.L., Jr. Personal communications
from Mr. R.J. Quentin, BASF Wyandotte
Corporation, Wyandotte, Mich. May
1972. OAP-SSPCP-SDID,* U.S. Environ-
mental Protection Agency, Research Tri-
angle Park, N.C.
38. British Petroleum Company. Canadian
Patent 867293 (1971).
39. Cell Systems Keep Mercury from Atmo-
sphere. Chem. and Eng. News. p. 14-15,
February 1972.
40. Housekeeping Procedures for Cell Rooms
of Mercury Cell Chlor-Alkali Plants. De-
veloped cooperatively by The Chlorine
Institute, Inc., New York, N.Y., and
OAP-SSPCP-SDID,* U.S. Environmental
Protection Agency, Research Triangle
Park, N.C. Febraury 1972.
41. Porter, D.H. and J.D. Watts. Economic
Aspects of Converting a Chlor-Alkali
Plant from Mercury Cells to Diaphragm.
(Presented at the American Institute of
Chemical Engineers National Meeting,
Houston, Texas, March 3, 1971.)
42. Roy, S.L., Jr. Personal communications
from Mr. L. Portnoy, Crawford and
Russel, Inc., Stamford, Conn. May 1972.
OAP-SSPCP-SDID,* U.S. Environmental
Protection Agency, Research Triangle
Park, N.C.
* Office of Air Programs, Stationary Source Pol-
lution Control Programs, Standards Development
Implementation Division.
3-36
-------
APPENDIX
A.1 CALCULATION OF EQUILIBRIUM
CONCENTRATION OF MERCURY VAPOR
IN A GAS AND RESULTANT LOSSES IN
PROCESS STREAMS OR IN VENTILATION
AIR
The weight of mercury, W, contained per
unit volume of gas at equilibrium can be
calculated from the ideal gas law, PV = nRT,
or, in this instance,
W/V = PM/RT
where: W = weight of Hg, mg
V = volume of gas, m3
P = equilibrium vapor pressure of Hg,
mm Hg at temperature of satu-
rated gas stream
M = molecular weight of Hg = 200.6
x 103 mg/mole
T = absolute temperature, °K
R = gas constant = 0.06237 (m3)
(mmHg)/(°K)(mole)
thus W/V (mg/m3)=1.09 x 104 x P at
standard conditions of 70°F and 29.92 inches
ofHg.
The equilibrium concentration of mer-
cury is plotted as a function of Fahrenheit
temperature in Figure A-l. These values can
be used to estimate the mercury vapor losses
to be expected in a mercury-saturated gas
stream as a function of the condenser oper-
ating temperature. For this calculation, it is
assumed that approximately 300 cubic meters
of hydrogen is produced and vented per ton
of chlorine produced. Some representative
calculated values are given in Table A-l. The
mercury vapor pressure data used were taken
from the Handbook of Chemistry and
Physics.1 The equilibrium concentration of
mercury can also be used to estimate the
order of magnitude of mercury emissions that
might be associated with the venting of the
end-box ventilation air. In this application the
mercury in the end-box is covered by an
aqueous layer to prevent rapid evaporation of
the mercury; hence the end-box air may not
100
_ 80
^ 6°
Is 40
| 20
t 1
O
cc
LU
z 6
O 4
o H
rv
2
s 1
1 §i
| 0.4
S" 0.2
0.1
20 40 60 80 100
TEMPERATURE, °F
120
Figure A-1. Equilibrium concentration
of mercury vapor in air as a function of
temperature.
be saturated with mercury vapor. If, however,
a mean air temperature, a degree of satu-
ration, and a ventilation air flow rate are
assumed, the mercury emission rate, E, in
pounds per day can be calculated from the
equation:
E = 0.09x 1Q-6 x VxC,
where: E = mercury emission rate, Ib/day
V = air flow rate, cfm
C = mercury concentration, Aig/m3
For example, assuming an end-box ventilation
flow rate of 1000 standard cubic feet per
minute, a mean air temperature of 100°F
(equilibrium concentration of mercury = 57
milligrams per cubic meter), and a 50 percent
saturation of the air, the mercury emission
rate would be:
A-l
-------
Table A-1. MERCURY VAPOR LOSSES IN A HYDROGEN STREAM3
Condenser
temperature
°C
40
30
25
20
14
4
0
-10
°F
104
86
77
68
57
39
32
14
Mercury
concentration.
mg/m3b
67
30
23
13
7.8
3.0
2.0
0.66
Mercury loss/ton CI2
g
20
9.1
6.8
4.0
2.4
0.90
0.60
0.20
Ib
0.040
0.020
0.015
0.008
0.005
0.002
0.0014
0.0004
Mercury loss/day
for 100-T plant
9
2000
910
680
400
240
90
60
20
Ib
4.4
2.0
1.5
0.80
0.51
0.20
0.14
0.04
aAssume 300m3 of hydrogen per ton of chlorine.
bAt 70°F and 29.92 in. Hg.
E - 0.09 x 10-6 x 1000 x (0.50 x 57,000)
= 2.6 Ib/day.
The calculation can be repeated for any
number of assumed conditions. When the
results obtained for an arbitrary series of
conditions are plotted as emission rate versus
ventilation air flow rate, a family of straight
lines is obtained, as shown in Figure A-2.
Actual ventilation flow rates used in the
operations that were surveyed ranged from
400 to 2000 cubic feet per minute. As
documented in the body of this report,
mercury emissions in the end-box ventilation
air have been noted to vary from 2 to 15
pounds per day for each 100 tons per day of
chlorine produced. The temperatures of un-
treated end-box ventilation air streams ranged
from approximately 100° to 160°F and had
mercury vapor saturations of about 10 to 80
percent.
The above equation may also be used to
calculate mercury emissions in building ven-
tilation air. The mercury vapor concentration
in this case is very low. If the actual value is
not known, the maximum emission rate may
be estimated by assuming the concentration
to be equal to the Threshold Limit Value;
that is, 50 micrograms per cubic meter. The
volumes encountered in the cell room venti-
lation air ranged from 100,000 to 1,000,000
cubic feet per minute. A few calculated
emission rates for arbitrary air flow rates in
these ranges are given in Table A-2.
A.2 METHODS OF ESTIMATING CON-
TROL COSTS
A.2.1 Equipment Costs
The equipment costs used in this doc-
ument are the free-on-board (f.o.b.) charges
for either a specific piece or system of control
equipment. Unless otherwise indicated, the
equipment costs are based on the use of
stainless steel as the construction material.
When more corrosion-resistant materials such
as titanium or titanium-clad steel are required,
the equipment costs are estimated to be 25
Table A-2. CALCULATED MERCURY VAPOR
EMISSION RATES IN CELL ROOM
VENTILATION AIRa
Airflow, cfm
1,000
2,500
5,000
10,000
100,000
250,000
500,000
1,000,000
Mercury emission rate,
Ib/day
0.0045
0.011
0.023
0.045
0.45
1.15
2.3
4.5
aAssume a mercury concentration of 50 jug/m3.
A-2
-------
= 100°F at 20%
T =100° F at 10%
200 400 600 800 1000 1200 1400 1600 1800 2000
AIR FLOW RATE FOR END-BOX VENTILATION, cfm
Figure A-2. Mercury emission rate as a function of
end-box ventilation air flow rate at arbitrary con-
ditions of air temperature (T) and percentage saturation (%).
percent greater than the cost based on stain-
less steel construction.
All equipment costs were obtained
through communications with vendors and
users of said equipment. In most situations,
averages of f.o.b. estimates from vendors and
users were selected. These f.o.b. estimates
serve as a basis for the calculation of other
costs, such as the fixed-capital requirement
and the annual operating cost.
A-3
-------
A.2.2 Fixed-Capital Requirement
The fixed-capital requirement for a piece
or system of control equipment includes the
following:2
1. Purchase price.
2. Cost of installation.
3. Instrumentation and control.
4. Piping.
5. Electrical equipment and materials.
6. Engineering and supervision.
7. Construction expenses.
8. Contingency.
In this list, the cost of installation includes
costs for labor, foundations, supports, plat-
forms, construction, and other items directly
related to the erection of the purchased
equipment.2
In order to estimate the fixed-capital
requirement, the following assumptions were
made. All cost estimates are presented as a
percentage of the equipment cost based on
stainless steel construction.3
1. Additional buildings, service facil-
ities, and site preparation are un-
necessary.
2. The cost of installation equals the
equipment cost.
3. Process piping, electrical equipment,
and instrumentation are in accord-
ance with the minimum require-
ments. The combined cost of these
items is equal to 40 percent of the
equipment cost.
4. Engineering expenses are approx-
imately 30 percent of the equipment
cost.
5. Construction overhead adds an
amount equal to 20 percent of the
equipment cost. This estimate is for
applications involving nonexplosive
gas streams. For control systems in-
volving the hydrogen gas stream of
chlor-alkali plants, the construction
overhead is 80 percent of the equip-
ment purchase cost.
6. Plant maintenance personnel provide
most of the construction force.
7. An amount equal to 30 percent of
the equipment cost is added to
account for contingencies.
If the preceding components of the fixed-
capital requirement are added to the equip-
ment purchase cost, estimates obtained of the
fixed-capital requirement range from 320 to
380 percent of the equipment cost. These
estimates are for nonexplosive and explosive
gas streams, respectively.
For applications requiring the use of
titanium or titanium-clad steel construction
material, the estimated fixed-capital require-
ment is increased by an amount equal to 25
percent of the stainless steel equipment cost.
The estimated fixed-capital requirement for
titanium equipment would therefore range
from 345 to 405 percent of the equipment
purchase cost.
The fixed-capital estimation method pre-
sented above may tend to result in high
estimates because the equipment costs are
based on stainless steel construction.4 The
accuracy of the estimation method is approx-
imately 30 percent.
The costs presented in this document
were fixed on January 1, 1972. Costs for later
dates can be estimated by using appropriate
indices.
A.2.3 Annual Operating Costs
The annual operating costs are estimated
to be 30 percent of the estimated capital
requirement.3 This estimating procedure is
based on accepted standard practice in the
chemical process industry and should provide
an ample allowance. The estimate includes
allowances for labor and supervision, mainten-
ance, payroll overhead, operating supplies,
indirect costs, and capital charges at 18
percent per year.3
A.3 CONTROL EQUIPMENT COSTS AND
EFFICIENCIES
Control efficiencies and equipment costs
are itemized in Tables A-3 through A-9.
A.4 GEOGRAPHIC LOCATION OF
CHLOR-ALKALI PLANTS THAT USE MER-
CURY CELLS
The chlor-alkali plants that use mercury
-------
cells are listed in Table A-10. A map showing
these locations is presented in Figure A-3.
A.5 REFERENCES FOR APPENDIX
1. Weast, R.C. Handbook of Chemistry and
Physics (50th Ed.)- Cleveland, Chemical
Rubber Company, 1970. p. D-139.
2. Peters, M.S. and K.D. Timmerhaus. Plant
Design and Economics for Chemical
Engineers (2nd Ed.). New York, Mc-
Graw-Hill, 1968. p. 96-97.
3. Basis for National Emission Standards for
Mercury. Battelle Memorial Institute.
Columbus, Ohio. Environmental Pro-
4.
tection Agency Contract No. EHSD
71-33. July 1971. p. A-l to A-4 and E-l
to E-2.
Gallagher, J.T. Rapid Estimation of Plant
Costs. Chem. Eng. 74(26):89, 1967.
5. Porter, D.H. and J.D. Watts. Economic
Aspects of Converting a Chlor-Alkali
Plant from Mercury Cells to Diaphragm.
Hooker Chemical Corp., Niagra Falls,
N.Y., and Blaw-Knox Chemical Plants,
Inc., Pittsburgh, Penn. (Presented at
American Institute of Chemical Engineers
National Meeting. Houston. March 3,
1971.)
A-5
-------
Table A-3. ESTIMATED COLLECTION
EFFICIENCIES OF CONTROL EQUIPMENT
Equipment
Heat exchanger and
knockout drum
Mist eliminator
Venturi scrubber
Depleted brine
scrubbing system
(system A)
Depleted brine
scrubbing system
(system B)
Hypochlorite scrubbing
system
Activated carbon
system
Molecular sieve
system
Table A-4. EQUIPMENT COSTS FOR PRIMARY
MERCURY EXTRACTION FACILITIES
AT 100-TON/DAY CAPACITY
Estimated collection
efficiency
80% (particulate removal)
86% (particulate removal)
95% (particulate removal)
Exit concentration of
85 Atg/m3 (particulate
and vapor removal)3
Exit concentration of
n
4.4 mg/m (particulate
and vapor removal)3
95 to 99% (particulate
and vapor removal)
90% (vapor removal)
Exit concentration of
n
0.5 mg/m (vapor
removal)3
Equipment
Secondary cooler (indirect)
Chiller
Knockout drum
Mist eliminator
Venturi scrubber
Carbon bed system
Purchaje cost
(itainlesi steel), S
8,300
14,000
600
9,500
3,700
15,500
Purchase cost
(titanium), $
10,400
—
750
11,900
4,600
19,400
Capital cost
(titanium). $
28,600
44,900
2,100
32,800
12,800
53,500
Table A-5. SYSTEM COSTS FOR PRIMARY
MERCURY EXTRACTION FACILITIES
WITH 100-TON/DAY CAPACITY
Concentrations will be reduced to this level,
regardless of loadings, according to the equipment
manufacturers.
System
Cooling and
partial mist
elimination
Total
Cooling and
mist elimina-
tion
Total
Wet scrubbing
and carbon bed
Total
Cooling and wet
scrubbing
Total
Equipment
Secondary cooler (indirect)
Chiller
Knockout drum
Secondary cooler (indirect)
Chiller
Knockout drum
Mist eliminator
Venturi scrubber
Carbon bed system
Secondary cooler (indirect)
Chiller
Venturi scrubber
Capital
costs, $
28,600
44,900
2,100
75,600
28,600
44,900
2,100
32,800
108,400
12,800
53,500
66,300
28,600
44,900
12,800
86,300
Emissions,
1b/day
5.8
1.7
1.8
1.7
Table A-6. EQU]PJVlEI\rrCOSTS FOR MERCURY-CELL
CHLOR-ALKALI PLANTSWITH 100-TON/DAY CAPACITY
Equipment
Primary cooler
(indirect)
Secondary cooler
(indirect)
Chiller
Seal tanks
Knockout drum
Blower
Mist eliminator
Petersen separator
Depleted-brine
scrubbing system
(system A)
Depleted-brine
scrubbing system
(system B)
Hypochlorite
scrubbing system
Carbon bed system
Molecular sieve
system
Purchase cost
(H2 stream), $
5,700
5,700
9,600
1,200
400
400
6,500
3,800
20,800
15,500
6,000
10,600
23,000
Capital cost
(H2 stream), $
21,700
21,700
30,700
4,600
1,500
1,500
24,700
14,400
79,000
58,900
22,800
40,300
87,400
Purchase cost
(end-box stream), $
7,100
7,100
9,600
-
500
500
8,300
4,800
26,000
19,400
7,500
13,300
28,800
Capital cost
(end-box stream), $
19,700
19,700
30,700
-
1,400
1,400
22,400
13,100
71,800
53,500
20,700
36,600
79,400
A-6
-------
Table A-7. CONTROL SYSTEM COSTS FOR CHLOR-ALKALI PLANTS
TREATING HYDROGEN AND END-BOX VENTILATION
STREAMS BASED ON 100-TON/DAY CAPACITY
System
Partial cooling and mist elimination
(see Figure 3-14 for basic system)
Total
Cooling and mist elimination
(Figure 3-16)
Total
Cooling, mist elimination, and
chemical scrubbing (Figure 3-17)
Total
Cooling, mist elimination, and
chemical scrubbing (Figure 3-17a)
Total
Cooling, mist elimination, and
chemical scrubbing (Figure 3-18)
Total
Cooling, mist elimination, and
activated carbon (Figure 3-19)
Total
Cooling, mist elimination, and
molecular sieve (Figure 3-20)
Total
Equipment
Primary cooler
Knockout drum
Seal tanks
Base system
Secondary cooler
Chiller
Knockout drum
Mist eliminator
Base system
Secondary cooler
Chiller
Mist eliminator
Depleted-brine scrubbing system
(system A)
Mist eliminator
Depleted-brine scrubbing system
(system Bb)
Base system
Secondary cooler
Chiller
Knockout drum
Hypochlorite scrubbing systemc
Petersen separator
Base system
Secondary cooler
Chiller
Knockout drum
Mist eliminator
Carbon bed
Base system
Secondary cooler
Chiller
Blower
Petersen separator
Molecular sieve system6
Capital cost
(H2 stream), $
21,700
1,500
4,600
27,800
27,800
21,700
30,700
1,500
24,700
106,400
27,800
21,700
30,700
24,700
79,000
183,900
24,700
58,900
83,600
27,800
21,700
30,700
1,500
22,800
14,400
118,900
27,800
21,700
30,700
1,500
24,700
40,300
146,700
27,800
21,700
30,700
1,500
14,400
87,400
183,500
Capital cost
(end-box stream), $
19,700
1,400
21,100
21,100
19,700
30,700
1,400
22,400
95,300
21,100
19,700
30,700
22,400
71,800
165,700
22,400
53,500
75,900
21,100
19,700
30,700
1,400
20,700
13,100
106,700
21,100
19,700
30,700
1,400
22,400
36,600
131,900
21,100
19,700
30,700
1,400
13,100
79,400
165,400
aSystem does not include alkaline scrubber.
b Includes cost of primary cooler and compressor.
cDoes not include licensing fee.
dIncludes preheater, two absorption beds, and 600°F H2 heater.
A-7
-------
Table A-8. CONTROL SYSTEM COSTS AND EMISSION RATES FOR COMBINED
HYDROGEN AND END-BOX VENTILATION STREAMS
System
Partial cooling and mist
elimination (see Figure
3-14 for base system)
Cooling and mist elimination
(Figure 3-16)
Cooling, mist elimination,
and chemical scrubbing
(Figure 3-17) (system A)
Cooling, mist elimination,
and chemical scrubbing
(Figure 3-17) (system B)a
Cooling, mist elimination,
and chemical scrubbing
(Figure 3-18)b
Cooling, mist elimination,
and activated carbon
(Figure 3-19)
Cooling, mist elimination,
and molecular sieve
(Figure 3-20)
Production rate, tons/day
100
250
500
100
250
500
100
250
500
100
250
500
100
250
500
100
250
500
100
250
500
Capital cost, $
49,000
85,000
129,000
202,000
350,000
531,000
350,000
607,000
919,000
160,000
277,000
420,000
226,000
392,000
594,000
279,000
483,000
733,000
349,000
605,000
917,000
Emissions, Ib/day
99.6
249.0
498.0
3.4
8.5
17.0
0.02
0.05
0.10
0.6
1.5
3.0
0.2
0.5
1.0
0.28
0.70
1.40
0.08
0.20
0.40
"System does not include alkaline scrubber.
bDoes not include licensing fee.
A-8
-------
Table A-9. CAPITAL.INVESTMENT REQUIRED TO CONVERT FROM MERCURY TO DIAPHRAGM
CELLS AT PLANT WITH 100-TON/DAY CAPACITY4-3
Site preparation including grading, sewers, fire protection, and road and
fence changes
New building for cells
Cells including headers, anodes, bus, cell rebuilding facilities, and rectifier
changes
Weak-liquor handling and storage
Weak-liquor evaporation plant including salt handling
New boiler to supply steam for weak-liquor evaportation plant
Caustic purification
Engineering and construction supervision
Contingency
Dismantling of old mercury cell facility
Credit for mercury recovered
Total cost
Capital investment, $
230,000
500,000
1,200,000
190,000
1,600,000
800,000
1,693,000
1,552,000
625,000
150,000
(540,000)
8,000,000
Assumptions: (1) Diaphragm cell would be installed in a new building in order to permit continued operation of old
plant until diaphragm cells are on-line; (2) Diaphragm cell caustic would be purified to give quality equivalent to
mercury cell liquor.
A-9
-------
Table A-10. MERCURY-CELL CHLOR-ALKALI PLANTS
IN THE UNITED STATES
State and city
Alabama
Le Moyne
Mclntosh
Mobile
Muscle Shoals
Delaware
Delaware City
Georgia
Augusta
Brunswick
Illinois
East St. Louis
Kentucky
Calvert City
Calvert City
Louisiana
Lake Charles
St. Gabrile
Geismar
Maine
Orrington
New York
Niagara Falls
Niagara Falls
Syracuse
Syracuse
North Carolina
Pisgah Forest
Acme
Ohio
Ashtabula
Tennessee
Charleston
Texas
Dear Park
Point Comfort
Washington
Bellingham
Longview
West Virginia
Moundsville
New Martinsville
Wisconsin
Port Edwards
Producer
Stauffer Chemical Co.
Olin Corp.
Diamond Shamrock Chemical Co.
Diamond Shamrock Chemical Co.
Diamond Shamrock Chemical Co.
Olin Corp.
Allied Chemical Corp.
Monsanto Co.
B.F Goodrich Chemical Corp.
Pennwalt Corp.
PPG Industries, Inc.
Stauffer Chemical Co.
BASF Wyandotte Corp.
Sobin Chlor-Alkali, Inc.
Hooker Chemical Corp.
Olin Corp.
Allied Chemical Corp.
Allied Chemical Corp.
Olin, Ecusta Operations
Allied Chemical Corp.
Detrex Chemical Industries, Inc.
Olin Corp.
Diamond Shamrock Chemical Co.
Aluminum Co. of America
Georgia-Pacific Corp.
Weyerhaeuser Co.
Allied Chemical Corp.
PPG Industries, Inc.
BASF Wyandotte Corp.
Year
built3
1965
1952
1964
1952
1965
1965
1957
1962
1966
1967
1969
1970
1964
1967
1961
1960
1946
1953
1947
1963
1963
1962
1938
1966
1965
1967
1953
1958
1967
Cell type
De Nora 22 x 5
OlinES
De Nora
De Nora 24 x 2M
De Nora 18 x 4
Olin E11F
Solvay V-100
De Nora 18x6
DeNora24H5
Uhde 30 m2
De Nora 48H5
Uhde 30m2
Uhde 30m2
De Nora 24H5
Uhde 20 m2
Olin E11F
Solvay Process SD 12
Solvay S60
Sorensen
Solvay V-200
Olin E11F
Olin E11F, E812
DeNora 18 SGL
De Nora 24 x 5
De Nora 18x4
DeNora 14TGL&24H5
Solvay S60
Uhde 20 m2
De Nora 24H5
aRefers to year chlorine production started at location.
A-10
-------
Figure A-3. Location of mercury-cell chlor-alkali plants in the United States.
-------
SUBJECT INDEX
Activated carbon, 3-14, - 3-16, 3-28, 3-29,
3-32 - 3-34, A-6 - A-8
B
Beneficiation, 3-4, 3-6, 3-7
Chemical scrubbing, 3-19, 3-26 - 3-28, 3-32
- 3-34, A-6, A-8
Chlor-alkali production, 2-1, 3-19 - 3-34, A-l
- A-4, A-6 - A-9
Control costs, 3-15 - 3-19, 3-31 - 3-33, A-2
- A-4, A-6
Chlor-alkali production, 3-31 - 3-34, A-2
- A-4, A-6 - A-9
Activated carbon, 3-32 — 3-34, A-6 —
A-8
Chemical scrubbing, 3-32 — 3-34,
A-6, A-8
Conversion to diaphragm cell, 3-32,
3-34, A-9
Cooling, 3-32, 3-34, A-6 - A-8
Mist elimination, 3-32, 3-34, A-6
Molecular sieve, 3-33, 3-34, A-6, A-8
Extraction, 3-15 - 3-19, A-2 - A-4, A-6
Activated carbon, 3-16, A-6
Cooling, 3-15,3-16, A-6
Mist elimination, 3-15, 3-16, A-6
Sulfuric acid scrubber, 3-19
Wet scrubbing, 3-15, A-6
Control techniques, 2-3 - 2-8, 3-11 - 3-15,
3-19, 3-24-3-31,3-33, A-6
Chlor-alkali production, 3-24 - 3-31,
3-33, A-6
Activated carbon, 3-28, 3-29, A-6
Chemical scrubbing, 3-26 - 3-28, A-6
Cooling and condensing, 3-24 — 3-26,
A-6
Housekeeping practices, 3-30, 3-31
Ion exchange process, 3-33
Mist elimination, 3-26, 3-27, A-6
Molecular sieve, 3-29, 3-30, A-6
Sulfuric acid scrubber, 3-33
Extraction, 3-11 - 3-15, 3-19, A-6
Activated carbon, 3-14, 3-15, A-6
Cooling and condensing, 3-11, 3-12,
A-6
Mist elimination, 3-12, 3-13, A-6
Sulfuric acid scrubber, 3-19
Wet scrubbing, 3-13, 3-14, A-6
Cooling and mist elimination, 3-11 — 3-13,
3-15, 3-16, 3-24 - 3-27, A-6 - A-8
D
Diaphragm-cell chlor-alkali process, 3-23,
3-24, 3-32, 3-34, A-9
E
Emissions, 2-3 - 2-8, 3-1 - 3-9, 3-19 - 3-24,
3-34, A-l, A-2, A-8
Chlor-alkali production, 3-22 — 3-24,
3-34, A-l, A-2, A-8
Diaphragm-cell process, 3-24, 3-34
Mercury-cell process, 3-22, 3-23, 3-34
Extraction, 3-1 - 3-9, A-l, A-2
Beneficiation, 3-7
Multiple-hearth furnaces, 3-4
1-1
-------
Retort, 3-8
Rotary furnaces, 3-3
Emissions reduction by process changes, 3-9
-3-11,3-24
Chlor-alkali production (diaphragm-cell
process), 3-24
Extraction, 3-9 - 3-11
Hydrometallurgical processing, 3-11
Pyrometallurgical processing, 3-10,
3-12
Extraction (primary mercury), 2-1, 3-1 —
3-19, A-l -A-4, A-6
N
New technology, 3-17 - 3-19, 3-33
Chlor-alkali production, 3-33
Extraction, 3-17 - 3-19
Hy drome tallurgical processing, 3-17
-3-19
Sulfuric acid scrubber, 3-19
H
Housekeeping practices (Chlor-alkali pro-
duction), 3-30, 3-31
Hydrometallurgical processing, 3-11, 3-17 —
3-19
Processes, 3-1 - 3-10, 3-17 - 3-24
Chlor-alkali production, 3-20 - 3-24
Diaphragm-cell process, 3-23
Mercury-cell process, 3-20 — 3-22
Pyrometallurgical processing, 3-10, 3-12
R
Ion exchange process, 3-33
Retort, 3-7 - 3-10
Rotary furnaces, 3-1 — 3-3
M
Mercury (properties, origin, production, use),
2-1 - 2-4, 2-7
Molecular sieve, 3-29, 3-30, 3-33, 3-34, A-6 -
A-8
Multiple-hearth furnaces, 3-3 — 3-5
Sulfuric acid scrubbing, 3-19, 3-33
W
Wet scrubbing, 3-13 - 3-15, A-6
1-2
U. S. GOVERNMENT PRINTING OFFICE: 1973 746768/4125
------- |