EPA-650/2-74-100
OCTOBER 1974
Environmental Protection Technology Series
I
55
\
5SS
UJ
C3
-------
EPA-650/2-74-100
PROCESS MODIFICATIONS FOR CONTROL
OF PARTICULATE EMISSIONS
FROM STATIONARY COMBUSTION,
INCINERATION, AND METALS
by
R. Nekervis, J. Pilcher,
J. Varga Jr. , B. Gonser, and J. Hallowell
Battelle, Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-1323
Task 9
ROAP No. 21ADK-017
Program Element No. 1AB012
EPA Project Officer: G. J. Foley
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S . ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
October 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
iii
ABSTRACT
The report summarizes the state of process modifications relative
to the control of fine particulate emissions from stationary combustion
sources (electric utilities and industrial processes); municipal incinera-
tors; iron and steel plants; ferro-alloy plants; and nonferrous metal
smelters (zinc plants, copper smelters, aluminum reduction cells). This
study is to uncover modifications to conventional practices or new uncon-
ventional practices which appear to improve the control of fine particulate
emissions in these five areas. Modifications to conventional stationary
combustion sources considered include ash fluxing, SO addition to flue gas,
staged combustion, use of fuel additives, fly-ash agglomeration, solvent
refining, and flue-gas recirculation. Unconventional systems studied include
fluidized bed, coal gasification, and submerged combustion. For incinerators,
combined fuel-refuse firing, gas cooling, and pyrolysis methods are considered,
Emphasis for iron and steel plants is given to the bottom-blowing oxygen
process (Q-BOP). Modification of the conventional reverberatory smelting
procedure and the introduction of hydrometallurgical methods are discussed
for copper, and the chloride electrolytic (ASP) process by ALCOA is considered
for aluminum. Each process is considered with respect to its stage of
development, availability or acceptability by industry, efficiency in reduc-
ing emissions, and environmental impact.
-------
iv
TABLE OF CONTENTS
Page
1. Introduction 1
2. Summary 2
2.1 Stationary Combustion 4
2.1.1 Electric Utilities 4
2.1.2 Industrial Processing and Steam Generation .... 6
2.2 Municipal Incinerators 6
2.3 Iron and Steel Plants 7
2.4 Ferroally Furnaces 8
2.5 The Primary Nonferrous Metals Industry ..... 9
2.5.1 Zinc Roasting, Sintering, and Distillation .... 9
2.5.2 Copper Roasting, Matte Smelting and Converting . . 9
2.5.3 Aluminum Reduction Cells 10
3. Stationary Combustion, Particulate Control Combustion
Modification 11
3.0 Introduction 11
3.1 Electric Utilities 11
3.1.0 Background; the Problem of Fine Particles 11
3.1.0.1 New Combustion System Modifications .... 13
3.1.0.2 Ash Fluxing 13
3.1.0.3 Fluidized-Bed Combustion 14
3.1.0.3.1 Status 18
3.1.0.4 Staged Combustion 18
3.1.0.5 Recirculation of Flue Gas 19
3.1.0.6 Gasification 19
-------
TABLE OF CONTENTS
(Continued)
Page
3.1.0.7 Modification of Fly Ash Resistivity .... 21
3.1.0.8 Agglomeration of Ash Particles 22
3.1.0.9 Advanced Power Cycles 22
3.1.1.1 Modification of Particle Size 23
3.1.1.2 Solvent Refined Coal 23
3.1.1.3 Low Excess Air 23
3.1.1.4 Electrochemical Oxidation of Coal 24
3.1.1.5 Fuel Additives 24
3.1.1.6 Improved Control Methods for Fine
Particulates 25
3.1.1.7 Electrical Control of Particulates
From Flames 25
3.2 Industrial Processing and Steam Generation 27
3.2.0 Introduction 27
3.2.1 Process Modifications 27
3.2.1.1 Burne'r Design 27
3.2.1.2 Flue-Gas Recirculation 28
3.2.1.3 Two-Stage Combustion 28
3.2.1.4 Other Combustion Modifications 29
3.2.1.5 Improved Service and Maintenance 30
3.2.1.6 Modifications to Reduce PNA 30
4. Municipal Incinerators 32
4.0 Introduction 32
4.1.0 Process Modifications 32
-------
vi
TABLE OF CONTENTS
(Continued)
Page
4.1.1 Combined Firing 33
4.1.2 The CPU-400 33
4.1.3 Water-Walled Incinerators 34
4.1.4 Pyrolysis Process 35
4.1.5 Advanced Concepts Involving Heat Recovery ... 37
4.1.6 Electron-beam Irradiation 38
4.2 Predictions for the Year 2000 38
5. Iron and Steel Plants 39
S.I Open Hearth Furnace 39
5.1.1 Hydrocarbon Additive to Lancing Operation ... ^0
5.1.1.1 State of Development -
Hydrocarbon Addition ^
5.1.1.2 Availability to Industry ^
5.1.1.3 Degree of Effectiveness ^
5.1.1.4 Environmental Effects
5.1.1.5 Use of Liquid Oxygen and Liquid
Hydrocarbon Injection ..... • • •
5.2 EOF Furnace ................. .... 45
5.3 Q-BOP Process .................... ^5
5.3.1 State of Development - Q-BOP Process ..... 45
5.3.2 Degree of Effectiveness ............ ^'
5.3.3 Environmental Effects ............. ^'
5.4 Electric Arc Furnace ................. 48
5.4.1 Preheating and Melting with
Oxygen-Fuel Burners ............. ^9
-------
vii
TABLE OF CONTENTS
(Continued)
5.4.1.1 State of Development 49
5.4.1.2 Degree of Effectiveness 49
5.4.1.3 Environmental Effects 50
5.4.2 Electric-Arc Furnace--Scrap Charge
Compatibility 50
5.4.2.1 State of Development 50
5.4.2.2 Availability to Industry 50
5.4.2.3 Acceptance by Industry 51
5.4.2.4 Degree of Effectiveness 51
5.5 Metallurgical Coke Ovens 51
5.5.1 Particulates from Charging Coke Ovens 51
5.5.2 Particulates from Pushing Coke 52
5.5.3 Particulate Emissions During Quenching .... 52
6. Ferroalloy Furnaces 54
7. Process Modifications For Particulate Control
In the Primary Nonferrous Metallurgical Industry 55
7.1 Zinc Roasting, Sintering, and Distillation 55
7.1.1 Roasting 55
7.1.2 Sintering 56
7.1.3 Reduction and Distillation 56
7.2 Copper Roasting, Matte Smelting, and Converting ... 56
7.2.1 Pyrometallurgical Modifications 57
7.2.1.1 Flash Smelting 57
7.2.1.1.1 Electric Furnaces 59
-------
viii
TABLE OF CONTENTS
(Continued)
Page
7.2.1.2 Continuous Smelting 60
7.2.1.3 Miscellaneous Pyrometallurgical
Developments 63
7.2.2 Hydrometallurgical Copper Recovery 68
7.2.2.1 Stanford LCPR Process 68
7.2.2.2 Anaconda's "Arbiter" Ammonia 70
Leach Process ..... 70
7.2.2.3 Sherritt-Gordon Process 72
7.2.2.4 Cymet Process 73
7.2.2.5 Duval Corporation and Other Processes . 73
7.3 Aluminum Reduction Cells 75
7.3.1 New Aluminum Reduction Processes 79
8. References 82
I
APPENDIX A
CONTROL OF FINE PARTICULATE EMISSIONS IN CONVENTIONAL
COPPER SMELTING PRACTICE A-l
-------
IX
LIST OF TABLES
Page
Table 1. Emissions from Primary Aluminum Industry--1970 .... 80
Table 2. Emissions from an Uncontrolled Prebake 80
Potline 80
Table 3. Fluoride Removal Efficiencies of Selected
Primary and Secondary Controlled Systems 80
Table A-l Annual Copper Production of Smelters and
Refineries in the U.S A-2
Table A-2 Control Practice in Copper Smelting A-10
Table A-3 Composition of Typical Emissions From
Reverberatory Stacks A-13
Table A-4 Compositions of Atmospheric Emissions From
Converter Operations in Primary Copper Industry . . A-15
LIST OF FIGURES
Figure 1. Simplified Fluidized-Bed Combustion Boiler Concept . . 15
Figure 2. Atmospheric Fluidized-Bed Combustion Power Plant ... 16
Figure 3. Pressurized Fluidized-Bed Combustion Power Plant ... 17
Figure 4. Statistical Distributions of Dust Loadings During the
Operation of a 225 net ton Openhearth Furnace
Using Oxygen Lancing and Oxygen + Propane Lancing . 42
Figure 5. Relationship Between Dust Loadings of a 225 net ton
Openhearth Furnace Operating with Oxygen Lancing
and Oxygen -f Propane Lancing ............ 43
Figure 6. OEM Process Vessel, U.S. Patent 3,706,549 ...... 46
Figure 7. Outokumpu Furnace .................. 58
Figure 8. INCO Flash Furnace .................. 58
Figure 9. Noranda Process ................... 61
-------
LIST OF FIGURES
(Continued)
Figure 10. Schematic view of Mitsubishi's
Semicommercial Plant 64
Figure 11. Bureau of Mines Autogenous Smelting 66
Figure 12. Flowsheet for the lime-concentrate-pellet-roast
Process with Direct Electrowinning 69
Figure 13. Anaconda Arbiter Plant—Block Flow Diagram 71
Figure 14. Schematic Flowsheet--Cymet Process 74
Figure 15. Schematic Drawing of a Prebaked Anode Cell 76
Figure 16. Schematic Drawing of a Horizontal Stud Soderberg
Aluminum Reduction Cell 77
Figure 17. Schematic Drawing of a Vertical Stud Soderberg
Aluminum Reduction Cell 78
Figure A-l Generalized Diagram of Conventional Smelting
Flowsheet A-3
Figure A-2 Cutaway View Showing Key Features of a
Conventional Reverberatory Furnace A-6
-------
1. Introduction
The purpose of this study is to summarize in one report the
state of technology relating to the control of fine particulate emissions
resulting from process modifications in the following areas (1) stationary
combustion as it relates to electric utilities and industrial processing,
(2) municipal incinerators, (3) iron and steel plants, (4) ferroalloy
plants, and (5) nonferrous metal smelters (zinc plants, copper smelters,
and aluminum reduction cells). The report is intended to serve as a brief-
ing document for persons who are not well versed in these areas.
Fine particulate emissions may be defined as those particles
smaller than 2 microns diameter. The sizes that will deposit in the lungs
(called "respirable dust") vary from submicron (O.Olp, and smaller) to about
8 microns. A high percentage of the larger particles (2 to 8u) that enter
the bronchial tubes is deposited by the mechanism of inertial impaction,
and a large fraction of the smaller particles (less than 0.1 u) is deposited
by means of diffusion. However, for the in-between sizes 0.1 to 2.0 y,),
only a small fraction is retained in the respiratory tract because neither
mechanism of deposition is efficient in this size range. Hence, the very
fine particles (less than 0.1 LL) which are not readily removed by collectors
designed in the conventional fashion are of growing concern because of their
possible adverse effects on health and reduction of atmospheric visibility.
Also, our ability to measure and characterize very fine particulates is
limited.
The literature reflects this state of affairs. In our study on
the effect of process modifications in the five areas mentioned in the
first paragraph above, we found that quantitive data on changes in par-
ticulate emissions as a result of either a modification or even a replace-
ment of a conventional process was not only not reported, but appeared to
be largely ignored. Accordingly only qualitative data are provided in
substantiation of claims of beneficial effects on the amount, size and
properties of fine particulate emissions. In almost all cases the degree
of reduction of fine particulate emissions could not be ascertained with
any degree of precision. This necessiated some changes in the format of
the report as originally conceived.
-------
In identifying process modifications that may have beneficial
effect on the control of fine participates, the approach was to determine
the mechanism responsible—agglomeration or growth of participates, altera-
tions of their properties, or reduction in mass. Insofar as possible,
information on the stage of development, availability and acceptance by
industry, and environmental side effects, if any, of both modifications
and the more unconventional system replacements have been covered.
There was an unusual number of new modifications in copper smelt-
ing technology which necessitated including background information on the
current status of emission control practices in conventional U.S. primary
copper smelters. This is provided in Appendix A.
2. Summary
This report summarizes the state of technology relating to bene-
ficial effects on fine particulate emissions that come about through new
process modifications in five major areas:
(1) stationary combustion
(a) electric utilities
(b) industrial processing
(2) municipal incinerators
(3) iron and steel plants
(4) ferroalloy plants
(5) nonferrous metal smelters
(a) zinc plants
(b) copper smelters
(c) aluminum reduction plants.
Summarizing results that relate to all five areas, the 1-2 percent
of fine particles that escapes present day high efficiency collectors are
practically all smaller than 1 micron. Quantitative information on the
size distribution, chemical properties, and the effects of these fine
particulates is lacking, probably because of the difficulties in their
measurement.
While quantitative data on fine particulate emissions from con-
vaitional stationary combustion equipment, municipal incinerators, iron
-------
and steel plants, ferroally plants, and nonferrous metal smelters are
very sparse, similar quantitive data relating to process modifications in
these areas are nonexistent. Even the qualitative data on the degree of
fine particulate emission from these current modifications in practice are
not definitive.
Bag filters are the most effective method for removing particles
smaller than 2 microns. Some types of scrubbers and many electrostatic
precipitators can also be effective for the collection of particles under
2 microns. The technology for particulate emission control has changed
little over the past 20-40 years. Characterization of very small particles
has been neglected because they are difficult to sample, classify, and
analyze. This difficulty is reflected in the dearth of quantitative data
on the effectiveness of process modification in reducing fine particulate
emissions.
Fine particulates are formed largely as a result of metals in the
coal ash being vaporized by the intense heat of combustion. Upon cooling,
these metallic vapors condense to form a multitude of very fine spherical
particles. A few small particles are formed also from polynuclear aromatic
compounds present in the coal which show up as particulate POM (Polycyclic
Organic Matter). Process modifications to circumvent or minimize particle
formation by these mechanisms would be helpful.
Our study indicates that generally submicron particle abatement
can be achieved more economically through either process modification or
outright substitution rather than through the installation of new, larger
and more efficient precipitators. To achieve an increment of improvement
from 99.0 to 99.5 percent efficiency in an electrostatic precipitator
requires a great increase in the size of the collecting apparatus with a
corresponding great increase in cost. Wet scrubbers are being designed as
a means of capturing particulate matter as well as S02 but these are
still in the early stages of development. Venturi scrubbers work well
down below 1.0 .
-------
2.1 Stationary Combustion
2.1.1 Electric Utilities
There are a number of modifications in conventional combustion
practices, as well as some totally new unconventional systems, being
considered by the electrical utility industry which appear to exert a
beneficial effect on the control of fine particulate emissions. Some
of these may have questionable side effects. None in the following
listing have as yet been studied in a quantitive way to determine the
amount, size distribution, and characteristics of fine particulate
emission. They are as follows:
(1) Ash fluxing to capture the ash in a molten slag,
thereby decreasing the dust loading in the flue
gas.
(2) Adding SO in controlled amounts to achieve an
optimum resistivity of the fly ash particulates
so as to improve their collection in electro-
static precipitation equipment; also altering
the burning rate, amount of excess air, coal
particle size and rank to achieve the same
result.
(3) Utilizing staged combustion (that is burning
fuel in a primary zone with less than stoichio-
tnetric amounts of ai'r, partially cooling the
combustion products, then completing combustion
in a second zone with additional air) to reduce
the emission of NO and particulates by reducing
3C
temperature. The degree of beneficial effect on
the control of particulates depends on the fusion
characteristics of the ash.
-------
(4) Utilizing fuel additives to reduce total participate
emissions (the possibility of toxicity of the newly
created emissions makes their use questionable for
the present).
(5) Agglomerating the fly ash in the combustion zone by
extending residence time to produce tackiness and
inducing high turbulence or pulsating flows to
promote agglomeration (this approach is still very
much in the concept stage).
(6) Solvent refining (reconstituing coal) to achieve low
ash and sulfur contents and thus resolve the particulate
problem (the cost of this is excessive).
(7) Recirculating flue gas, i.e., diluting the fuel air
mixture in the combustion zone with cooled flue gases,
has been found to contribute to lower emissions of
smoke when firing fuel oil. Application of this
technique is still in the early stages of investigation.
Unconventional systems being considered include
(1) Fluidized-bed combustion systems which use relatively
large particles of coal in comparison to the pulverized
coal used in conventional combustion systems; while
more particulates are generated, the larger size leads
to greater collection efficiencies with existing
collectors. It appears that high-pressure fluidized
combustion systems offer even greater collection
efficiencies.
(2) Gasification of coal can be considered as modified
staged combustion. While fluidized-bed combustors
have not as yet been applied to boilers, there are
a number under development for gasification of coal
to produce pipeline gas or liquid fuels.
(3) Carrying the gasification of coal a step further,
there are the submerged combusion processes. One
aspect of this concept is to burn pulverized coal
in a bath of molten salts or. coal-ash slag
-------
to capture sulfur and fly ash in the molten salt or
slag. In a related process, crushed coal is injected
into molten iron which dissolves the fixed carbon and
sulfur constituents, the volatile constituents of the
coal are driven off and cracked to H? and CO. Air
introduced into the molten iron produces a "Bessemer"
reaction which oxidizes the carbon to CO, Ash of the
coal becomes a supernatant slag which is fluxed with
enough lime to extract sulfur from the iron and hold
it in the slag which is drawn off continuously.
(Development work on these processes is in progress,
but no details have been publicly released.)
2.1.2 Industrial processing and Steam Generation
The above conclusions on modifications of stationary combustion
units such as flue-gas recirculation and two-stage combustion apply to the
industrial processing area also.
Poly-Nuclear-Aromatic (PNA) compound emissions into the atmosphere
are generated principally by commercial and residential coal-fired units.
Oxidation during the combustion process will destroy PNA and other hydro-
carbons. There is a dearth of data identifying the conditions required
for the oxidation of PNA, and modifications of the combustion processes
to reduce PNA emissions.
2.2 Municipal Incinerators
(1) Tests on combined firing of municipal refuse (20
percent) and fossil fuels (80 percent) in steam
generating boilers indicate that the particulate
emissions can be controlled to the degree obtain-
able when firing fuel alone if modifications in
the electrostatic collection system are made to
accommodate the higher stack volume.
-------
(2) Water-walled incinerator furnaces permit the cooling of
gases to a temperature which permits the operation of
flue-gas cleaning equipment. While this substantially
reduces the particulate emissions over that of conven-
tional incinerators, there is a corrosion problem
associated with deleterious salts and gases that are
present in all municipal incinerators.
(3) Pyrolysis processes have the potential of reducing
particulate emissions; quantitative data on this
aspect of these processes have not been published.
2.3 Iron and Steel Plants
A major change in the oxygen-blowing process in steel making
from blowing at the surface of a melt in a vessel (the EOF or BOP process)
to bottom-blowing oxygen with a hydrocarbon through the melt (U.S. Steel
Corporation's Q-BOP and related processes) results in a quieter bath with
reduced fumes and larger particulate sizes in the emissions; dusting is
reported to be 1/3 to 1/5 that of the EOF process; in addition the Q-BOP
and related processes provides slightly higher rates of production and
slightly better ultization of scrap than does the EOF process,
Installed or planned capacity of Q-BOP vessels in the U.S. was
about 9,000,000 tons annually as of March, 1973. U.S. 1970 production
of steel was 150,000,000 net tons, of which the basic oxygen furnace (BOF)
accounted for 55 percent (Q-BOP production is included in these BOF
statistics), the open hearth accounted for 26 percent, and the electric
furnace, 18 percent. There is potential for explosive growth in the
Q-BOP and related processes. Of the 21 open hearth shops in operation,
only ten are modern plants with extensive air pollution control, three are
scheduled to close, and eight will be replaced. The Q-BOP would appear
to be a logical replacement for these, but present economic factors and a
muddied patent situation make this choice unclear. Since the production
-------
rates and degree of scrap utilization of the Q-BOP vessels are only
marginally higher than that of the EOF vessels, there is no pressure on
the EOF shops to convert.
The use of hydrocarbons in combination with oxygen during
lancing open-hearth steel has been found to reduce the emission of fumes
and the grain loadings of exhaust gases when tested on a laboratory scale
ten years ago. Except for a few plant trials since, the results of which
were not reported, there has been no further interest in the process mod-
ification; it has not been accepted by U.S. industry. On the other hand,
the Soviets are experimenting with the use of liquid oxygen-liquid hydro-
carbons in the lancing of open hearth steels.
The reduction of particulate emissions from coke-oven operations
by agglomeration of the particulates, or their reduction, by process
modification, does not appear to be imminent, owing to the hard, angular
characteristics of the carbon particulates. New processing equipment, such
as pipe-line charging and other modifications of charging methods will
probably be able to achieve the required reductions in emissions during
charging. New equipment, presently under evaluation under production
conditions, may be the answer to the reduction or even elimination of
emissions during the pushing and quenching operations.
2.4 Ferroally Furnaces
The only modification in ferroally furnace practice which
resulted in improvement in the control of particulate emission uncovered
in this study was a study on the use of pelletized ore concentrates.
These reduced particulate emission by 42 percent in comparison to that
of fine ore concentrates.
-------
2.5 The Primary Nonferrous Metals Industry
2.5.1 Zinc Roasting, Sintering, and Distillation
Electrolytic zinc plants, present and being constructed, have
negligble loss of particulates other than in general handling of concen-
trates and calcines or sinter. The two pyrometallurgical zinc plants use
good practice in keeping the escape of particulates to the atmosphere at
a minimum, and no modifications have been suggested for improvement in
the basic steps of their processes.
2.5.2 Copper Roasting, Matte Smelting, and Converting
Progress in controlling the extent of emissions of fine particles
in the copper smelting industry appears to be substantial. There are two
general approaches; one is to modify the conventional reverberatory
smelting procedure, the other is to replace the pyrometallurgical methods
with hydrometallurgical methods. The goal in both cases is first to
reduce or eliminate the emission of sulfur oxide and concommitantly to
reduce stack losses of particulates.
Modification of reverberatory converter procedures has involved
the introduction of continuous smelting, flash smelting, and flash roasting-
electric furnace combinations, all of which seek a reduction in gas flow and
an enrichment of S0» in the effluent gases to permit recovery of sulfur
as sulfuric acid, liquid SO , or elemental sulfur, which in turn necessi-
tates cleaner off-gases. An exception to the new smelter modification
to recover SO. is a limestone scrubbing treatment to control sulfur. It
is only moderately efficient in removing SCL, but in combination with a
preceding dry-gas cleaning system, it is excellent for removing par-
ticulates.
Hydrometallurgical methods substantially eliminate the emission
of particulates to the atmosphere. A number of new processes are being
developed; among these are pressurized ammonia leaches and chloride leaches.
One of the new processes, Anaconda's "Arbiter" Ammonia Leach Process, has
-------
10
reached the commercial stage. The only particulate control involved is
in handling concentrates and lime and in preventing tailings from becoming
wind blown.
2.5.3 Aluminum Reduction Cells
Aluminum reduction cells which use prebaked anodes,and account
for 59 percent of total U.S. production of aluminum, produce less voliti-
zation of pitch and less fouling of the emission control system than do
Soderberg reduction cells which use baked-in-place anodes. However, the
separate anode baking furnace requies an emission control system.
Alcoa's aluminum-chloride electrolytic process (the ASP Process)
has reached the pilot plant stage. Press releases on the process indicate
less dusting than is the case with the conventional Hall aluminum reduction
cells. No cryolite (Na Al Fe,) will be required. Fluorides, and the
working of Al 0 into the bath will be eliminated. Emphasis, however, is
J O
on the savings of energy (as much as 30 percent less energy than the con-
ventional Hall process) and on the fact that "scarce and costly cryolite"
will no longer be required.
There are a number of new processes aimed at exploiting new
sources of alumina, such as alunite, the kaolin clays, and laterite.
These do not affect materially the amount or type of emissions from the
reduction cells.
-------
11
3. Stationary Combustion, Participate Control Combustion Modification
3.0 Introduction
Theburning of coal, more than any other fuel, impacts on the
ambient air because of the emission of particulates: fly ash, fine
particules of ash; small, but nevertheless important amounts, of organics
as vapor or fine droplets; and sulfuric acid mist, in the amount of 1 or 2
percent of the sulfur equivalent in the coal. Thirty-two percent of the
industrial particulate emissions in the U.S. are from coal burning.
In the burning of pulverized coal, the characteristics of the
particulate emissions depend on many factors. Among these are the heat
liberation rate in the furnace, the composition of the coal (especially
the ash content), the degree of pulverization of the coal, and the amount
of excess air. In addition, the basic parameters of time, temperature and
turbulence, plus the reactivity of the coal influence the quantity and
characteristics of particulate emissions.
Polycyclic organic matter (POM) which is emitted in particulate
form from a vast number of stationary combustion sources will be considered
part of this study. Particulate POM consists of a variety of chemical
entities; however, it is common practice to use benzo (a?) pyrene as an indi-
cator of other POM owing to the demonstrated carcinogenicity of benzo (o)
pyrene and the relatively large amount of published data on it. ^ Relatively
high levels of POM have been measured in coal-burning equipment.
This section of the report considers practical combustion process
modifications that are likely to improve the situation with regard to par-
ticulate emission.
3.1 Electric Utilities
3.1.0 Background; the Problem of Fine Particles
Over 80 percent of the potential fly ash from U.S. coal-fired
power plants is now being captured by means of large electrostatic pre-
cipitators, and most new power plants now have improved collectors guaranteed
-------
12
to remove 98 to 99 percent of the fly ash leaving the furnace. However, the
small amount (only 1 or 2 percent) which escapes these high-efficiency collect-
ors, is practically all smaller than 1 micron. The environmental impact
of these extremely fine particles has been largely ignored because quantitative
information on the particles and their effects is so difficult to measure.
However, their effect in making stack plumes highly visible because of the
light-scattering effect of 0.4 to 0.8 micron particles, though hard to
quantify, is often readily apparent to the community. Furthermore, such
plumes add to the haze so common in metropolitan areas.
There is increasing concern about these fine particles which are
not adequately controlled with existing equipment. Often, their physical
and chemical properties are not known. However, it is recognized that these
fine particles, by nature of their size, are the principal contributors to
adverse health effects, visibility reduction, and soiling of surfaces. A
review panel of the National Research Council has identified the importance
of fine particles and reports that current practices for evaluating control
techniques as "tonnage-collection figures and weight-removal efficiencies
are inadequate to delineate the entire particle-emission problem". They
further state that.... "small particles....may continue to limit visibility
and may affect health even when presently uncontrolled sources of particulate
matter are equipped with the best collection devices currently available".
However, to obtain improved efficiencies in electrostatic precipitation
from 99 to 99.5 percent with today's technology involves a great increase
in size with a corresponding great increase in cost. Accordingly, alternative
strategies for the control of fine particulates are being sought.
One of the most obvious alternatives would be to keep the ash
in the furnace. A characteristic of conventional pulverized coal-fired
dry-bottom furnaces is that 50 to 70 percent of the ash leaves the furnace
as fine particles suspended in the hot (300 F) exhaust gas. Although some
of the ash particles become liquid during combustion they are cooled and
solidified before they reach the walls of dry-bottom furnaces. In cyclone
and conventional slagging or wet-bottom furnaces some of these particles
are liquid or sticky when they reach the walls where they combine to form
-------
13
a viscous slag layer which slowly drains down to a slag tap. The cyclone
furnace retains about 90 percent of the ash as slag. However, because of the
intense combustion in the cyclone resulting in high flame temperatures, some
of the ash is vaporized. When this vapor condenses later in the system,
large numbers of extremely fine spherical particles are formed by the La Mer
effect which defy conventional collecting systems,
3.1.0.1 New Combustion System Modifications
Other alternative modifications to existing combustion systems
are being considered. For example, fluidized combustion systems use
relatively large particles of coal in comparison to the pulverized coal
used in conventional combustion systems and this use of larger coal particles
generates larger particulates in the combustion gas. Although there are
more particulates generated by this system, the larger size leads to a
greater collection efficiency with existing collectors. It appears that
operating fluidized-bed combustion systems at high pressures leads to even
greater collection efficiency. Among other new developments that may
influence fine-particle generation are combustion-system modifications
like staged combustion and flue-gas recirculation that are intended for
NO control, and submerged combustion in molten salts. These and others
X
are discussed individually below:
3.1.0.2 Ash Fluxing
As pointed out earlier, conventional slag-tap and cyclone
furnaces convert much more of the ash into molten slag than do the more
common dry-bottom pulverized-coal-fired boiler furnaces. Decreasing the
viscosity of coal-ash slags will tend to increase the amount of ash con-
verted to slag, thereby decreasing the dust loading in the flue gas.
Fusibility of coal ash varies widely. However, the ash from some coal will
not form slag in the cyclone. The addition of inexpensive fluxes such as
limestone (CaO) and mill scale or iron ore (Fe20.) can lower slag viscosity
appreciably.
-------
14
3.1.0.3 Fluidized-Bed Combustion
Burning crushed coal in a fluidized bed of limestone has
demonstrated merit as a means of reducing gaseous pollutants (SO and NO ).
X X
Figure 1 shows a simplified fluidized-bed combustion boiler concept.
Research is under way on atmospheric as well as on elevated pressure
(10 atmospheres) fluidized-bed combustors. Development work for OCR and
EPA by Pope, Evans and Robbins has lead to the announcement ' ' of an OCR
contract for the design, construction and operation of a 30 megawatt
coal-fired atmospheric-pressure, fluidized-bed boiler at the Rivesville
Station of Monongahela Power at Fairmont, West Virginia. Several organiza-
tions are engaged in bench and pilot plant scale development of pressurized
fluidized-bed boilers, including conceptual designs of 30 to 635 MW plants.
These studies include sorbent regeneration and sulfur recovery, as well as
reduction of SO , NO , trace materials, and particulate emissions.
X X
In an atmospheric fluidized-bed combustion power plant, the
electric power is generated by a steam turbine as shown in Figure 2. In
a pressurized combustion system, additional power is generated by a gas
turbine which is driven by the high pressure hot-combustion gas as shown
in Figure 3.
Fluidized-bed combustors have not as yet been applied commercially
in boilers, but there are at least five fluidized-bed reactors currently
under development for gasification of coal to produce pipeline gas and/or
(45)
liquid fuels. ' These activities include Battelle"s program for the
Office of Coal Research employing the Union Carbide coal gasification
process aimed toward production of synthesis gas; the self-agglomerating
fluidized-bed concept employed involves production of a clean gas from
the combustor and a circulating agglomerated ash burden to supply thermal
energy to the fluidized-bed gas generator.
Pressurized fluid-bed combustion (the process that appears to
be preferred for electric utility use) appears to be less complex than
combustion through coal gasification. Pressurized operation is preferred
because of the significant decrease in the size of the plant over atmospheric
operation. For large utility plants now contemplated, the multiplicity of
atmospheric boilers needed would appear to be prohibitive.
-------
15
Convection
Section
•53*—
Walls
Raffle ,.--•
Primary
Cyclone
Secondary
Particulate Removal
VJ
Hest Recovery
Section
Exhaust
Water
Walis
;--r:--'.-7-- , /i.','.,
'~»"*-.r~""-~—— ••. . —fv ' v
A;r
l.i.-fte
Ash; Per
Pra'iaaiei'. ^L-i^.i; hc-atcr
or tfehtvisr Soil
Distribulor Plw'
Pr?35urc:
Ccal Sue;
Air Flow:
Temper a lure:
1 - 25 at;n
pf - 1/4 in.
2-15 ft/sec
1400 - 1900°F
Surfauo: Water Wiills, Horizontal, avid
Vartica) Tubsi if Bc.3
Snlfui removal: »';aO t S02 + >iO2 -* CaSO4
FIGURE 1. SIMPLIFIED FLUIDIZED-BED COMBUSTION BOILER CONCEPT
-------
16
Paniculate Removal — r-^
E,w,,o,
Superhtwtet
Gv(t:'i>:Y"o:
Air -«,-.» -^-•T"!1
Force D'iitt
Fail
J
,
-
._. .,
—
1
...!T»» p
1
— — - ~, ^~^-
. _.J C
"V, f V"™"
Lime Coal
l_
^v.
# r i
11
v rH V
* %,;ipi
f j
. —
Reheater
h...
l\ '•';
'!"
©-H";
%s
— «S7-
/-""
^x
jier Pi
Induced Draft Fan — j i —
1
Lconomizer — >
nnr
joi r°r.i
Rej^iierjijor
[JO
J
!
i
1
\
\
^
ir.p
|
ff-U
Stsd
Sulfur
r<: ewe-;')'
Unit
r ,
-!
MM.W.WM«njU.^
/ i
Stiltons A •!<
FIGURE 2. ATMOSPHERIC FLUIDIZED-BED COMBUSTION POWER PLANT
-------
17
Cyclone
, ITS.-
i
1
rr>.
-r-W'.w''**— *^\_/-<
^~«.rf-\_X^^«»^''S^->^*"
D C
_J-.B.
\
—
T I
ri
Lime
i
"t1"
Coal
/
g^
Sorb^nt
Rtf;eii(:r
-------
18
3.1.0.3.1 Status
The fluidized-bed combustion system offers the potential for
burning high-sulfur and low rank coals having a wide range of properties.
Since fuel-bed temperatures are low, slagging problems are minimized.
Heat release and heat transfer rates to tubes immersed in the bed of
fluidized solids are considerably higher than those in conventional boilers.
Furthermore, it is possible that fluidized-bed combustion can be advanced
to a commercial reality with less effort and time than many of the gasi-
fication methods now under development.
The major disadvantages of the fluidized-bed approach are en-
gineering rather than fundamental in nature. Problems currently requiring
solution relate to solids and gas handling, feeding (especially in pressure
operations) and limitations on control of load changes. Additional
problems are concerned with efficiency of utilization of sorbent, bed
particle attrition and elutriation, and incomplete combustion of coal.
However, in combined cycle operations, there is the fundamental problem of
developing an adequate cleaning process to remove particulates. An elaborate
reinjection system to handle carry-over of unburned combustibles probably
will be required.
A fluidized bed of noncombustible particles (not consisting of
lime or magnesia), arranged so that particles of coal are burned in close
contact with inert particles to which they are transferring thermal energy,
offers some promise for particulate control. This is a highly complicated
system and much is lacking in an understanding of the mechanism of combustion
in fluidized beds that would be required to achieve optimum conditions in
large central-station power plants.
3.1.0.4 Staged Combustion
Staged combustion involves burning fuel in a fuel rich primary
zone with less than stoichiometric amounts of air (90-95 percent), partially
cooling the combustion products, and finally supplying additional air and
completing combustion in a secondary zone. The earliest downfired furnaces
for firing pulverized coal employed essentially a two-stage combustion
process which suggests that two-stage combustion with pulverized coal
-------
19
may be feasible. Two-stage combustion has been demonstrated, at power
plant scale, as a successful method of reducing the emission of NQX by
reducing temperatures. Depending upon the fusion characteristics of the
ash, this process may also have a beneficial effect on control of
particulates.
3.1.0.5 Recirculation of Flue Gas
The recirculation of flue gas to reduce the peak temperature has
been demonstrated on a small scale and has been found to decrease the forma-
tion of NO . However, its effect on particulate emissions is not known.
Factual data are lacking regarding the volume of flue gas that should be re-
circulated, its composition and temperature, and the points at which it should
be injected into the furnace.
3.1.0.6 Gasification
Gasification of coal can be considered as a modified combustor
in which the first stage, the gasifier, is operated rich. Babcock and
Wilcox are currently studying one variation in which ash- and sulfur-free
gas is produced and burned in a combustion boiler and gas turbine cycle.
Other unconventional systems which involve submerged combustion/gasifica-
tion are as follows:
The ATC Molten-Iron Gasification Process, The concept of
Applied Technology Corporation's JtBolten-iron coal gasification process is
designed to gasify high-sulfur coal with the use of a molten-iron bath in
which the sulfur is separated and recovered in usable form while producing
an essentially sulfur-free, low-Btu gas, suitable for combustion under
steam boilers as a second stage of combustion.
The heart of the process is a molten-iron bath with a depth of
3-4 feet contained in a cylindrical, refractory-lined vessel called a
combustor. ' Crushed coal is injected into the iron bath through a
submerged lance to a depth of about 25 inches. This depth has been found
sufficient to allow time for the volatile constituents of the coal to be
driven off and cracked to H^ and CO and for the fixed carbon to be dissolved
completely in the iron. The sulfur contained in the coal is also dissolved
in the iron.
-------
20
Simultaneously, preheated air is injected through air lances
which need to be submerged to a depth of only 5-6 inches. The oxygen of
the air produces a Bessemer-type reaction which preferentially oxidizes
the dissolved carbon to CO. By balancing the coal injection rate with the
air injection rate, the carbon content of the iron would be maintained in
the range of 3-4 percent.
The ash of the coal, predominantly Al 0 and SiO , floats to
the surface of the iron to form a slag. Lime, as limestone, is injected
with the coal to flux the coal ash and to provide enough CaO to extract
the sulfur from the iron and hold it in the slag as CaS. Sulfur-laden
slag is drained continuously from the combustor to make way for new slag.
The slag is treated hot with superheated steam to produce elemental sulfur,
H2S, and SO . The latter gases are sent to a Claus reactor and converted
to elemental sulfur. The desulfurized slag is disposed of or part of it is
recycled through the combustor if more slag volume is needed.
The gas generated is composed mainly of CO, H«, and N_ and has
3 Li
a heating value of 160 Btu/ft . It comes off the combustor at about 2500 F
and is conducted directly to a steam boiler where it is burned with secondary
air to CO and HO. A portion of the off-gas is burned in a stove to preheat
the primary air injected into the combustor. It is reported that, under
proper operation, the SO content of the off-gas will be less than 60 ppm.
Status. Experimental and development work, under the sponsor-
ship of EPA, has been carried on in the Pittsburgh Laboratories of ATC,
(8)
since 1970. It was recently reported that ATC has received a contract
to develop design criteria for a 50 to 100 MW power generating unit based
on their process. ATC has reported that preliminary cost studies have
shown the process to be competitive with conventional coal-fired plant.
EPA is also considering some engineering studies for application to power plant
which include more detailed cost estimated.
Submerged Combustion in Molten Salts. The concept of submerged
combustion of coal in a molten bath of basic salts is another variation on
the theme of simultaneously providing a sink for slag, sulfur, and trace
metal contaminants of the coal at the combustion site. Currently the use
of molten salts for this purpose is being investigated by two organizations:
(1) Atomics International is investigating the partial combustion of coal
-------
21
(9)
with production of low-Btu gas by submerged combustion ; and (2) M. W.
Kellogg is studying the potential for production of high-Btu gas by a
similar process. '
The Atomics International (AI) process is an extension of their
previous work with the use of a molten-salt scrubber for the removal of
sulfur oxides from flue gas. Both Al and Kellogg use sodium carbonate
as the liquid medium at about 1800 F. The major difference between the
two processes is that the Kellogg process employs steam as a source of
hydrogen to produce a high Btu gas. The Kellogg gasifier is run at 1200
psia to enhance the hydrogenation of the CO produced by the combustion
process. The AI process on the other hand involves simple partial combustion
of the coal to carbon monoxide. In both systems, the sulfur is retained in
the ash/molten-salt mix. A portion of the contaiminated salt is withdrawn
and regenerated with the production of hydrogen sulfide or sulfur as a
byproduct.
Status. Both systems are still in the early investigative
stages. Many problems, such as amounts of sodium carbonate make up
required, corrosion problems, and the percentage of sulfur retained in the
melt remain unanswered.
The Superslagging Combustor. The concept of burning crushed or
pulverized coal in a bath of molten coal-ash slag is another example of
(12)
submerged combustion of coal incorporating sulfur removal. The two
aspects of the Superslagging Combustor that offer an incentive for consid-
eration of the concept as a coal combustion method are (1) capture of coal
sulfur by the slag to provide hot, low-sulfur combustion gases to a boiler,
and (2) capture of fly ash by the slag that could reduce emissions of fine
particulates and perhaps eliminate the need for an electrostatic precipita-
tor. This concept, being investigated at Battelle-Columbus has not been
tested experimentally and has received only limited analysis to date.
3.1.0.7 Modification of Fly Ash Resistivity
The ability of an electrostatic precipitator to capture particles
of fly ash depends in a large measure on the electrical resistivity of the
fly ash. When the electrical resistivity exceeds 2 x 10 ohm-centimeters,
-------
22
the normal voltage gradients in a precipitator are upset and the collection
efficiency drops sharply. Controlled amounts of SO may be added to the
flue gas to alter the resistivity of the fly ash particles so as to improve
collection of the fine particulates. Also, by modifying the combustion process
by such procedures as changing the rate of burning, particle size, amount
of excess air and rank of coal, the resistivity of the ash, and conse-
quently the efficiency of electrostatic precipitators can be improved.
3.1.0.8 Agglomeration of Ash Particles
Ash particles could be more easily removed from the combustion
gases if agglomeration of the fly ash particles could be achieved. In
order to adhere on contact the particles must be sticky which imples some
lower temperature limit at which agglomeration can occur. High turbulence
or pulsating flows may promote agglomeration of the stick particles. Also,
a longer residence time in the flame should encourage agglomeration,
3.1.0.9 Advanced Power Cycles
Concepts other than those presently in central-station power
plants undoubtedly will be developed over the next 30 years to convert
the energy in fuels into electricity. Magnetohydrodynamics and fuel cells
offermost promise of the so-called direct-energy-conversion processes,
but both have serious shortcomings. More attractive at present are advanced
power cycles wherein two or more energy-extraction systems are combined in
a single power plant.
Typical of such systems is the combined cycle burning of a gas-
eous or liquid fuel in a pressurized -boiler furnace followed by expansion
Of the hot products of combustion through a gas turbine. The gaseous fuel
could be natural gas, but another likely fuel is a hot sulfur-free producer
gas made from coal, or a fuel gas produced from residual fuel oil. A suit-
able gasification scheme including a hot gas-desulfurizing step of the type
described above is required. In addition to higher overall thermal efficiency,
which can be expected to exceed 50 percent, such cycles have the great
advantage of essentially complete removal of sulfur and particulates from
the exhaust gas.
-------
23
Many types of advanced power cycles can be devised, including
those in which volatile matter in the coal is distilled off and only
the residual char is burned in the power plant. Such systems have been
investigated for many years, but without the present-day incentive of
emission control. The need for "clean" power is renewing that interest^13}
Power plants based on such cycles should emit substantially less pollutants
than current plants.
3.1.1.1 Modification of Particle Size
The size of the ash discharged from a conventional furnace is
affected primarily by the size of the coal fired. The finer the grind
the finer the ash. Unfortunately, from the particle control standpoint,
fine grinding promotes combustion which makes the resulting ash harder
to catch. Also, high temperature flames increase ash vaporization and
subsequent condensation into fine particles. The full effect of the
particle size of pulverized coal on particulate control requires more
investigation.
3.1.1.2 Solvent Refined Coal
Solvent refined coal (SRC) is a reconstituted coal, low in
ash and sulfur' . The use of SRC in place of high sulfur and high ash
fuel is a marked modification of conventional methods of burning coal.
The particulate emission problem would be resolved, but cost is a factor
that would present a serious obstacle.
3.1.1.3 Low Excess Air
If pulverized coal could be burned with low excess air rather
than with 15 percent excess air, as is common today in large boiler
furnaces, the problem with particulate emissions would be eased. Advan-
tages are that the volume of the flue gas would be reduced appreciably
and more effective control of particulates would be possible, and higher
efficiency would be achieved (meaning less thermal pollution). To accomplish
this may require improved burners, finer sized coal, and higher turbulence
levels in the furnace.
-------
24
In a recent investigation of particulate emissions from oil-
fired residential heating units, conducted by Battelle for the EPA, two
units (a warm-air furnace and a boiler) were fired in the laboratory
while smoke and filterable particulate emissions were measured at several
excess air levels for both cyclic and steady-state runs. In addition,
particle-size distributions were measured during runs on the boiler to
determine if particle-size variations might help explain the lack of
correlation between smoke and particulate emissions, based on earlier field
measurements.
It was determined that particulate emissions varied with excess
air in the same manner as smoke varies with excess air. Also, correlations
between smoke and particulate emissions appeared practical for given units
firing at specific operating conditions. However, the data did not suggest
a general correlation between smoke and filterable particulate emissions.
Particle size distributions indicated that over 80 percent of the particles
were below 1.0 micron and the distributions were nearly uniform for all
runs, so that particle size did not explain differences between smoke and
particulate emissions.
3.1.1.4 Electrochemical Oxidation of Coal
Fuel cells using coal as the fuel may offer a long-range solution
to the need for supplying large quantities of electrical energy without
excessive emissions. The extent to which particulate emissions would be
generated is uncertain and would depend on details of the process which
are not now known.
3.1.1.5 Fuel Additives
The U.S. Environmental Protection Agency completed in 1971
a study of the use of fuel additives to control air pollution from distill-
ate oil-burning systems. Each candidate was analyzed for elemental com-
position to provide a basis for testing. A standard screening procedure
was established to test the effect of each additive on emissions from
fuel oil combustion. Screening tests were carried out on all distillate
soluble additives and the most promising additives were subjected to a
rigorous examination.
-------
25
Fuel additives were found not to be a promising way of reducing
air pollution from distillate oil combustion. A majority of the additives
tested had no beneficial effects on air pollutant emissions; in fact, some
additives even increased total particulate and NO emissions. Several of
A
the metal containing additives, e.g., Ferrocene, Cl-2, and Fuelco SO
3, re-
duced total particulate emissions; however, the unknown toxicity of new
emissions they create makes their use questionable. Further, there is
evidence that for distillate oils, burner modifications are a more suitable
route to air pollution control.
3.1.1.6 Improved Control Methods for Fine Particulates
The only control method now being tried for subtnicron particle
abatement is the application, in new installations, of larger and more
efficient precipitators. Experience thus far indicates that efficiencies
of 99.5 percent and above will achieve invisible plumes except for the
sulfuric acid mist that appears almost immediately as a bluish-white haze
when the small amount of sulfur trioxide in the emerging gases combines
with the moisture in the ambient air to form ^SO^ mist. Assuming that
SO abatement processes become feasible so that H SO mist is eliminated,
then achievement of fly ash emission levels low enough to produce invisible
exhaust plumes will be feasible at a price.
3.1.1.7 Electrical Control of Particulates from Flames
A radical departure from conventional submicron particle
abatement methods is the application of electric fields to charged dis-
persions so as to alter particle trajectories so that fully formed particles
may be caused to deposit in specific places or to burn up if they are
combustible. Also, the rate of particle formation, size, and concentra-
tion may be modified by applying electric fields to the region in which
they are formed
It has been demonstrated that the entire process of burning sprays
and particulate dispersions can be controlled electrically, all the way
from electrical dispersion and charging, over mixing with air inducted by
ion pumps, to burning controlled by electrical fields.
-------
26
The attachment of flame ions, which would otherwise recombine
unprofitably, to particulate pollutants to allow their subsequent manipu-
lation by electric fields appears possible for all particulates. For
species whose boiling point lies below the flame temperature, it is
necessary to use a second flame as an ion source in the cooler part of
the stream.
The concept has only recently been introduced, but already the
theory of predicting rates of charging and hence particle trajectories
in fields, in terms of their growth histories has been established, and
predictions of optimum conditions can be expected to follow
-------
27
3.2 Industrial Processing and Steam Generation
3.2.0 Introduction
This subsection is concerned with process modifications with
the objective of controlling particulate emissions from industrial processes
such as cement and lime production, glass melting and brick and ceramic
production, and from industrial steam generation and space heating.
Metallurgical processes are covered in Sections 5, 6, and 7.
Most pollutants from industrial processing sources are not
subject to reduction by modification of the combustion process. For
example, the particulates include those from thermal processing in which
the finest fractions of the processed materials or ash escape with the
combustion gases; such particulate emission would normally be reduced by
means of dust collectors rather than by combustion modifications.
The amount of particulate POM formed will vary greatly. Efficient,
controlled combustion favors very low POM emissions, whereas inefficient
burning results in high emissions. Hand-stoked residential furnaces
account for most of the particulate POM emitted.
3.2.1 Process Modifications
Energy-conversion devices for industrial processing and steam
generation are used primarily to convert the chemical energy in fuel to
thermal energy in the form of steam, hot water, or warm air. Possibili-
ties for reducing particulate emissions by process modifications are
discussed in the following paragraphs.
3.2.1.1 Burner Design
Burner design is not a well developed scientific discipline
but has been, and still is, largely an art. Most burner designs are
arrived at by trial and error and hardware-oriented development, rather
than by direct application of a body of scientific knowledge of the
subject of combustion. This is not too surprising, considering the
complexity of the combustion process. However, as the combustion process
is understood more fully and as complete mathematical models describing
both the physical and chemical aspects of the process are constructed,
-------
28
there should be improved understanding of the relationship between par-
ticulate emissions and combustion conditions. As a result, burners and
furnaces may eventually be designed to assure minimum generation of fine
particulates.
3.2.1.2 Flue-Gas Recirculation
Flue-gas recirculation involves diluting the fuel-air mixture
in the combustion zone with cooled flue gases and,thereby, reducing flame
temperature rise and peak gas temperature. Utilizing cooled flue gas
as the diluent does not adversely affect combustion efficiency as neither
the mass nor the temperature of exhaust products need increase.
Studies ^ 'have shown that recirculated flue gas can also
contribute to lower emission of smoke when firing fuel oil.
Application of flue-gas recirculation to small and intermediate
size combustion equipment is limited by knowledge gaps in the following
areas:
• Effect on emission levels of variables such
as temperature of the recirculated gas, point at
which the gas is injected, and quantity of re-
circulated gas
• Corrosion and deposits associated with recirculation
« Control of recirculated flue gas quantity in equip-
ment operating at variable firing rates
• Start-up problems.
3.2.1.3 Two-Stage Combustion
Two-stage combustion involves burning fuel in a fuel-rich
primary zone, partially cooling the combustion products, and, finally,
supplying additional air and completing combustion in a secondary zone.
Thus, two-stage combustion makes possible low peak gas temperatures,
because the combustion products are cooled before combustion is completed
and oxygen levels are low in the highest temperature region—the primary
combustion zone. Beinstock, Amsler, and Bauer 'found that, when
pulverized coal was burned in a power plant with 5 percent excess air in
the primary zone and 17 percent excess air added at the proper location
downstream, NO emissions did not increase above levels obtained when
X
-------
29
burning with 5 percent excess air and no downstream air addition. Two-
stage combustion has not been demonstrated as practical on smaller
combustion units.
Application of two-stage combustion to small- and intermediate-
size combustion equipment is limited because of gaps in knowledge of:
• Effect on emissions of important combustion
variables such as primary-zone fuel-air ratio}
temperature drop between the primary and the
secondary zones, and secondary zone fuel-air ratio.
• Possible corrosion associated with reducing con-
ditions in the primary zone.
* How to design combustion equipment to achieve
effective two-stage combustion (i.e., how to achieve
necessary mixing in the secondary zone, how to achieve
burnout of incompletely burned products formed in the
primary zone, what the residence-time requirements are
for primary and secondary zones, and how to incorporate
sufficient cooling between zones without increasing
equipment size and emissions of smoke, C0, and HC).
3.2.1.4 Other Combustion Modifications
Other techniques to achieve combustion at lower temperatures
and possibly reduce particulate emissions are (1) catalytic- or
surface-combustion burners, (2) radiant-heating devices, and (3) com-
bustors that operate at high excess air. These techniques have not been
developed to the point of practical application to heating boilers or
furnaces.
Surface combustion has been used successfully on a number of
gas-fired infrared or surface heating units. These units operate with
combustion zone temperatures of about 1800 F to 2000 F and, although no
specific data are available, these units are likely to have low NO
A
emission levels. Surface combustion of gaseous fuels could be incorpor-
ated into residential heating units with little difficulty, although
-------
30
practical combustion of fuel oil has not been demonstrated for long-term
operation. Potential problems with fuel oil include achieving complete
combustion, achieving uniform fuel feed across the burner face, and
preventing deposits on the burner surface.
3.2.1.5 Improved Service and Maintenance
Although only scattered data on burner servicing and maintenance
(18
are available ' , it is known that lack of proper burner adjustment
and/or maintenace can result in poor combustion ^ "and higher levels of
pollutant emissions. This is especially true with regard to pollutants
associated with incomplete combustion of the fuel (combustible particulate,
CO, and HC) .
Medium- to large- size combustion units are most likely to receive
proper maintenance under service contracts or by on- site operators. There-
fore, reasonably good combustion performance of these units should be
realized. However, emissions such as particulate POM that might be over-
looked by service personnel still may not receive proper attention.
Education and proper instrumentation would assist in this area.
3.2.1.6 Modifications to Reduce PNA
PNA (Poly-Nuclear Aromatic) compounds, which appear in particulate
POM (Polycyclic Organic Matter) comprise only a small fraction of total
particulate emissions, but they are significant as some of them are
carcinogenic .
Over 90 percent of all PNA emissions from energy-conversion
combustion processes is attributed to commercial and residential coal-fired
sources utilizing fixed-bed combustion. Although small coal- fired units
are disappearing from use, their combustion with regard to PNA emissions
is significant.
PNA compounds may be present in the coal itself or may form
(21 22)
during the combustion process v ' . However, there appears to be no
data in the literature which identify PNA as being present in coal. Diehl,
(23)
DuBruel, and Glenn x ' observed variations in PNA emissions when firing
different coals on a pulsating-grate stoker. However, they also observed
-------
31
such variations when operating a pulverized-coal-fired boiler with one
fuel and under essentially constant conditions. Consequently, they were
not able to relate PNA emissions to coal, and the source of PNA remains
uncertain.
The destruction of PNA, as for any other hydrocarbons, should be
by oxidation during the combustion process. The literature appears to
contain no data which would identify any unique conditions required for
oxidation of PNA ' . However, the fact that PNA compounds are emitted
suggests incomplete combustion of these compounds.
PNA emissions decrease as unit size increases. Larger units
provide longer residence times in the flame and, generally, higher average
gas temperatures--so that conditions contributing to more nearly complete
combustion prevail. Also, fixed-bed combustion units (with poorer mixing)
tend to emit larger quantities of PNA than do spreader stokers or pulverized-
coal-fired units.
It is possible that, with a better knowledge of how PNA occurs
or forms and of the conditions promoting its oxidation, modifications of
the combustion process to reduce PNA emissions could be identified.
-------
32
4. Municipal Incinerators
4.0 Introduction
Approximately three percent of the 4.3 billion tons of solid waste
(24)
generated in the United States annually is municipal solid waste (MSW)v .
This means that about 130 million tons of MSW, high in organic content, is
generated per year comprised mainly of residential, commercial, and some
industrial wastes. This waste can be utilized as an economical and ecological
fuel to help relieve the "energy crisis" and, at the same time, achieve
considerable recovery of certain natural resources such as metals and glass,
provided particulate emissions can be controlled.
In 1968, about 90 percent of MSW was disposed of by landfill, with
(25)
only 9 percent being incinerated . Waste requiring disposal is increasing
at the rate of 7 million tons per year, whereas suitable areas available for
economic disposal by landfill is rapidly dwindling. The alternative is to
incinerate an increased percentage of this material. A result would be more
air pollution unless improved burning processes are developed and adopted.
The fact that incinerators can be located near population centers and that
they require less land areas in comparison to landfills enhances their
atrractiveness to urban planners. Incineration is expected to increase in
popularity as improved methods, especially for particulate control and resource
recovery, are developed. The use of MSW as a source of energy should be a big
plus factor.
4.1.0 Process Modifications
A number of process modifications have been proposed and several
are under investigation (some at the pilot plant stage) to control emissions
of particulate and, at the same time, utilize some of the heat generated
and recover certain natural resources such as metals and glass. These will
now be discussed individually.
-------
33
4.1.1 Combined Firing
In a current task order study for the Emission Standards and
Engineering Division, OAQPS of EPA, Battelle-Columbus is compiling information
on the combined firing of municipal refuse and fossil fuels in steam-
generating boilers. The study includes a survey to determine the planning
status of new and retrofit combined- firing installations in the United
States, an engineering discussion of the technology of combined firing,
estimation of the emissions anticipated from such sources, descriptions of
control devices for these emissions, and the impact of combined firing on
fossil fuel usage and on solid waste disposal situations in the United States.
The only extensive application of combined firing at present is
the demonstration project in St, Louis which is sponsored by EPA, Union
Electric and the City of St. Louis. Test data on emissions are incomplete
at this time. However, it is expected that, for a nominal firing ratio of
80 percent pulverized coal and 20 percent prepared municipal refuse, the
particulate emissions can be controlled by electrostatic precipitators,
Codified somewhat to accommodate the higher stack gas volume produced, so
that final particulate emissions will be similar to those produced when
firing coal alone.
4.1.2 The CPU-400
Combustion Power Company, Inc., developed the CPU-400 system which
recovers energy from the combustible MSW in the form of electric power through
the use of a gas turbine powered electric generator. Simultaneously, the
CPU-400 will concentrate, separate, and recover valuable noncombustible
materials for recycling. Claims are that CPU-400 produces about five percent
of the electric power needs of the community supplying the solid waste.
Income from the sale of electric power and recylced materials should result
in a substantial reduction in the cost of solid waste disposal.
The pilot plant of the CPU-400 was developed in Menlo Park, California,
This plant is a complete system capable of consuming 80 tons per day of
-------
34
combustible solid waste while producing 1,000 kilowatts of electric power
and separating steel, aluminum, glass, and other inerts from the incoming
material. The CFU-400 system will be composed of three identical fluid bed
combustor/gas turbine modules. Each 3,000 kilowatts module will have a
capacity on the order of 150 tons per day. The modular design of the
CPU-400 will allow plants processing from 150 to 1,500 tons per day to be
constructed and will provide redundancy to assure reliable disposal
capability. During a visit to Menlo Park by a Battelle staff member on
/ n f \
February 6, 1974 , he observed that, when burning sawdust, the system
works well, the stack is clear, and the dust loading to the turbine and
to the stack is extremely low. Turbine erosion and corrosion do not appear
to be problems. However, there is a problem of deposition of A.l?0~ on
turbine blades. This is attributed to the melting of aluminum foil which
enters with the garbage. A particle bed collector is being considered as
a final stage for removal of these Al 0_ particles. In cleaning the
combustion products do not meet turbine requirements, environmental
standards are met.
4.1.3 Water-Walled Incinerators
The use of pollution-control devices requires that the incinerator
furnace gases be cooled to permit operation of the flue-gas cleaning equipment.
An attractive way to do this is to absorb much of the heat by water contained
in furnace wall tubes and in convection-pass tubes.
In addition to greatly decreasing the size requirements for pollution-
control equipment and fans, the water-wall incinerator has several attractive
features. First, there is a gainful use of the heat energy available in the
refuse. Second, the furnace throughput can be increased because of the rapid
absorption of heat. Third, wall slagging problems often encountered in
refractory-type construction are absent.
The technology of the water-wall incinerator was developed first in
Europe where it is in fairly extensive use. Corrosion problems have been
reported, however, in some instances. The first operating water-wall unit on
-------
35
this continent was at the Navy Public Works Center in Norfolk, Virginia,
beginning in 1969. Units are now operating in Montreal, Braintree,
Massachusetts, Harrisburg, and Northwest Chicago. In all locations except
Braintree, the steam is being wasted for lack of local demand. It is
anticipated that this type of construction will be used more and more in the
future.
In March, 1969, research was started at Battelle-Columbus on a
grant program supported by the Solid Waste Management Office, EPA, which was
aimed at determining the cause and extent of fireside metal wastage in
incinerators and devising methods of alleviation. In March, 1971, work was
started on a supplemental program aimed at obtaining a better understanding
°f incinerator-gas scrubber corrosion and also of metal wastage of grates.
(27)
Field and laboratory studiesv demonstrated that the wastage
in water-wall refuse boilers can be more severe than that normally encountered
in fossil-fuel-fired boilers. The complex nature of the refuse used as the
fuel and the relatively poorer control of burning in an incinerator combine
fco increase the possibility for corrosion. The contributors to the attack
are corrosive gases and low-melting chloride and sulfur-containing salts
which exert a fluxing action on the protective films on the metal surface.
These low-melting salts primarily contain compounds such as zinc and lead
chlorides along with potassium bisulfate and potassium pyrosulfate. The
data developed reveal that the gases S02, S03, HC1, and C12 are also playing
a major role in the wastage processes.
Analyses of tube deposits and furnace gases confirm the belief
that sufficient quantities of the deleterious salts and gases are present
in all municipal incinerators to warrant careful consideration from a
corrosion standpoint. Particulate emissions should be substantial, more
controlable in water-walled incinerators.
^•1.4 Pyrolysis Processes
Pyrolysis is a process by which carbonaceous materials break down
into simpler compounds and elements by means of heating in the absence of oxygen.
-------
36
A breadown of the organic portion of the refuse into oils, gases and chars
occurs. The Bureau of Mines has a pyrolysis unit which consists of a furnace
heated with nickel-chromium resistors and a recovery train to trap products.
Dependent on the reaction conditions, one ton of municipal refuse yields
154-230 pounds of char residue, 0.5-5 gallons of tar and pitch, 1.2-2 gallons
of light oil, and 11,000 - 17,000 cubic feet of gas, in addition to 80-133
gallons of aqueous liquor, and 18-25 pounds of ammonium sulfate. Typical
thermal values for the products are: char, 8-13,00 Btu/lb; oil, 150,000 Btu/gal;
and gas, 500 Btu/cu ft.
Enviro-Chem has developed a Landgard pyrolysis process which
emphasizes disposing of the solid waste, rather than recovery. The refuse
is shredded and reduced to a fairly homogeneous mixture with particle size
( 28}
about 3-4 inches in diameter . It is then dumped into a kiln lined with
refractory material. The kiln is slightly inclined from exit to entrance
so charred material falls by gravity into the water-quenching unit. The
residue is water-cooled before passing to a magnetic separator which removes
ferrous metals. The wet residue, reduced from the original trash volume by
94 percent, is then loaded into trucks and taken to landfill.
Gases driven off from the kiln during pyrolysis are passed through
a combustion chamber where hydrocarbons are oxidized. Complete combustion
is assured by an afterburner module. Product gases--carbon dioxide, nitrogen,
etc.--are passed through an adiabatic spray-scrubber and then released to
the atmosphere. Steam plumes, which might be objectionable from a public
relations standpoint, are elmininated by heating the stack gases. Since the
gases are clean, very low stacks can be used except where otherwise required
by local ordinance. The Landgard system accepts municipal refuse as it is
and no hand separation procedures are necessary. The company guarantees to
meet all existing air pollution standards in effect when the contract was
signed. Cooling water is diverted to sedimentation basins for recirculation.
However, the Landgard process has not yet been demonstrated commercially.
The Garrett process is designed to recover salable heating fuels,
glass and magnetic metals. The organic portion of these wastes is converted
to low sulfur oil, char and gas using a flash pyrolysis process. The process
-------
37
is designed to be expanded into an integrated series of processing stages for
the recovery of over 90 percent of the raw materials contained in municipal
refuse. Incoming solid wastes are shredded, dried, and passed through an
air classifier which separates most of the metals, glass, and other inorganic
materials. The overhead stream from the air classifier is then subjected
to a two-stage screening to improve separation of inorganics. The remaining
refuse is shredded a second time and then pyrolized, where it is broken down
into smaller molecules through the application of heat in the absence of
oxygen.
Laboratory studies of the pyrolysis process resulted in the production
of approximately one barrel of good quality oil per ton of as-received refuse.
Such refuse usually will also yield about 140 pounds of magnetic metals,
120 pounds of glass, and 160 pounds of char.
The flash pyrolysis operation which is the heart of the Garrett
process was researched for over a year in a continuous laboratory reactor.
Product yields, quality, and the initial favorable economic projections have
since been confirmed at a 4 TPD pilot plant during an 18-month period of
operations at LaVerne, California.
The quantity of solid material going to landfill from a 200 ton/day
plant is only 16 tons/day. The solid debris contains unrecovered glass,
aluminum copper, zinc, nickel, and other sterile material. GE&D is currently
investigating the economics of reclamation of nonferrous metals, and if a
system can be developed, the tonnage of solid rejects going to landfill
will be halved.
Other pyrolysis processes which could reduce particulate emissions
are described in a catalog of resource recovery processes prepared by Midwest
Research Institute for the Council on Enviornmental Quality in February, 1973^29'
4.1.5 Advanced Concepts Involving Heat Recovery
Other methods of incineration involving energy conversion are
(24}
described in the NCRR Bulletin for the Summer of 1973V '. They include:
-------
38
• Burning refuse in existing heat exchangers
• Hydrogenation, in presence of CO and steam
• Anaerobic digestion to produce methane
• Cubetting/Briquetting.
Energy recovery and the productive reutilization of materials,
combined with effective control of particulate emissions should result in
the commercial development of some of these concepts over the next five years.
4.1.6 Electron-beam Irradiation
The Japan Atomic Energy Research Institute (JAERI) and Ebara Mfg.
Co., are jointly grooming a project at jAERI's Takasaki laboratories, whereby
such treatment causes both sulfur dioxide and nitrogen oxides to drop out.
In one test, exposure to a "few megarads" of irradiation produced
90 percent SO- removal and virtually 100 percent NO removal from a 10-cu m/hr
fc X
flow of flue gas containing about 1,000 ppm of SO. and 80 to 100 ppm NO .
£* X
Full process details have not been disclosed, but the Japanese
indicate that the two impurities become captured in aerosol form by electro-
static dust collectors.
4.2 Predictions for the Year 2000
Potential incinerator particulate emissions for both uncontrolled
(furnace emissions) and for those abated by control devices (stack emissions)
(25)
were estimated as follows:
Thousands of tons per year
1968 2000
Furnace particulates 182 1064
Stack particulates 142 391
These data reflect a large increase in disposal of MSW by incineration, but also
a great improvement in the control of particulates.
-------
39
5. Iron and Steel Plants
Particulates generated during the production of steel can originate
from several sources. These sources are rust, scale or dirt on the steel
scrap used as part of the furnace charge, the degrees of size and thickness
of the individual pieces of scrap (small, thin pieces of scrap will oxidize
so rapidly that they approach an actual burning situation), fine parti-
culates associated with flux additions to the charge, refractories used for
patching, and the oxidation of metallics during the period which high-
purity oxygen is blown into the molten metal bath to accelerate the
refining reactions.
Particulates of interest in this study are those that originate
from the flux additions and the oxidized metallics formed during the
oxygen lancing periods. These particulates are associated with the three
major processes for making steel; open hearth, basic oxygen (EOF or BOP),
and electric-arc furnace processes. Considerable work has been done in
the United States and Europe in an attempt to determine the mechanism or
f *\C\ *3 1 *%. *) ^ ^ ^ ZL T ^ ^ A \
mechanisms for the formation of fume during oxygen lancing^ ' ' ' ' ' ' .
Information on the actual mechanism of formation, that could possibly lead
to process modifications and improved particulate control have not been
forthcoming. However, the factors that affect the formation of fume
have been found to be: (1) height of the oxygen lance above the molten
metal bath, (2) the carbon content of the molten metal bath, (3) molten
bath temperature, (4) concentration of oxygen in the lance gas (i.e.,
purity of oxygen), and (5) size of orifice and number of orifices in the •
(34,36)
oxygen lance
5.1 Open Hearth Furnace
The open hearth furnace is a shallow hearth furnace that can
be alternately fired from either end. Refractory brick work regenerators
are located at each end of the furnace which serve to recover heat from
the products of combustion and in turn preheat the incoming air for fuel
combustion. Briefly the process consists of charging steel and iron
scrap into the furnace, melting or partially melting the scrap, charging
-------
40
molten pig iron to the furnace, and refining the steel by blowing high
purity oxygen into the molten bath. The greatest particulate problem
occurs during the time high-purity oxygen is used to remove carbon,
manganese and silicon from the molten bath. During the oxygen lancing
period very turbulent conditions are created in the bath, with the result
that significatnt quantities of iron, manganese and silicon oxides are
formed and carried into the exhaust system of the furnace.
5.1.1 Hydrocarbon Additive to Lancing Operation
Laboratory research work started in 1959 resulted in a report
in 1963 concerned with the use of natural gas in combination with oxygen,
as a means for reducing the amount of emissions during oxygen lancing of
steel . Various factors were investigated as to their affect on fume
formation. These factors were: (1) amount of methane, (2) carbon content
of the bath, and (3) temperature of the bath. The use of methane in this
manner was shown to be beneficial in reducing the amount of fume, on
laboratory scale experiments.
5.1.1.1 State of Development - Hydrocarbon Addition
Development work was not actually undertaken since the applica-
tion of the modification requires only a source of hydrocarbons and the
valving with controls to regulate the flow of hydrocarbons to the oxygen
lance. Industrial trials of tMs process modification were reported in
(38)
1966 . This work was done with propane-oxygen combinations on 225
net ton open-hearth furnaces in Canada. Results of this work were
similar to those obtained with natural gas in the earlier laboratory
studies.
5.1.1.2 Availability to Industry
There are no apparent restrictions on the availability of this
process modification for use by the industry.
Acceptance by Industry. The iron and steel industry has not
accepted this process modification, if the lack of further industrial
work can be taken as an indication of acceptance. Shortly after the final
laboratory work reported in 1963, plant trials were held, but the results
were inconclusive and were not reported. This was followed by the work
-------
41
reported in 1966. No additional work of record has been reported. Lack
of acceptance may be attributed to the following: (1) more stringent air
pollution control requirements (hydrocarbon additions, reduced but did
not eliminate emissions), (2) possible additional safety hazards associated
with such a modification, and (3) increased cost of natural gas and similar
hydrocarbons.
5.1.1.3 Degree of Effectiveness
The use of hydrocarbons in combination with oxygen during the
lancing of open-hearth steel did reduce the grain loadings per cubit foot
of exhaust gases. No data were avaialble pertaining to the agglomerating
characteristics of the particulates.
Particulate loading for the work reported on the 225 net ton
open-hearths is shown by the statistical distribution in Figure 4. A
relationship between dust loadings with the use of oxygen-propane mixtures
and oxygen alone, is shown in Figure 5, for the same data. The work con-
ducted with the oxygen-propane mixtures did not consider many of the
operating variables that would be of interest to an open-hearth furnace
steelmaker. The report did state the desireability of further tests to
evaluate the process modification in greater detail.
5.1.1.4 Environmental Effects
There were no detrimental effects mentioned in the work reported.
While the reported results did show a reduction in the grain loadings,
the work was done without the use of any type of pollution control equip-
ment. Therefore, the reported results do not lend themselves to a con-
clusion on the overall effects of collection efficiency and improvements
in the ease of collection.
-------
-------
43
i
u
2
1
1
-ft-
01 ~
a
w ' - .:.
TO
c : " •
Sr—^r-
c
ra — 11"
c ~-~-
60 ~0— fr
o
u
•H ~
•H
•a
m
3~&:*5
u
w -
H-U
> 1
— Log
~~
. ._
' - . ^^
-
i
•
1 i ' :
IT; T
\ ,
1
1
]
g (O^+Propan
;
±-;~:
..;.—-:-
-
-T-
— —
T
s -
X^ • 1
" 1
:9.-6 0-7 0.8
0.9 1.0
Dust Loading
.
i
i .
• • •
!!'!
e)) =
. . ._ _.
•
-
j
-0.631
-- •
• -- '
32
-S'
> \
1 \ i
i
i
• i
!
; :.. .-.- ~r-*
172 + 0.3221
— ~
. -
!
— 1 \
r
" 1
>
:
L. :_
. . _ . _
. - — i
j
!
1 ! i
i i i 1
43 (Du
\ :_ :
s
\ :J
S
— ] — \ — ' — '
i j
— ; : i ;
', '
^^=E,
^-_-_-
_.
1
1 i 1
liij
• - •
t Load
1
^^-EEE
• ! . i
— ~ [—
1
1
— -^— —
r
i
'Mi
1 . '
I 1
1
I ' '
! i
r~r
ing 02)^— j+
s*^~
— • — i '-
__ — ._, ,
— , ._
— — ^
— — '. 1... . . ~~
. 1 —
— — • —
, — -
| - ^- —
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.
With Oxygen Only, grains per SCF
5. Relationship Between Dust Loadings of a 225 net ton Openhearth
Operating With Oxygen Lancing and Oxygen + Propane Lancing
Furnace
5.1.1.5 Use of Liquid Oxygen and Liquid Hydrocarbon Injection
A search of the foreign literature resulted in three items origin-
ating in the USSR with respect to the use of liquid oxygen and liquid
oxygen-liquid hydrocarbons in the lancing of open hearth steels. A con-
tinuing search is in progress to obtain any information concerning reports
on the work. The three items are quoted from abstracts in the literature
and are given for information and possible future use.
-------
44
(391
Iron and Steel Refining, USSR Patent No. 236499v . Iron and
steel refining is carried out by blowing with liquid oxygen and this reduces
the smoke formation and increases the amount of refined product. Liquid
oxygen in quantities corresponding to the gas is fed through the insulated
pipes to the 100 ton ladle containing iron. The heat is spent on bringing
the metal to the boil, for evaporation and for maintaining the temperature
of the reaction. Due to the low temperature of the reaction, the forma-
tion of smoke is reduced. The oxygen is diffused in the melt ensuring
fast oxidation of the impurities.
Treatment of M>ital. USSR Patent No. 276116 . In ferrous
metallurgy, a molten metal is treated with liquid hydrocarbons. To speed
up the process and reduce smoke formation, the liquid hydrocarbons are
fed into the molten metal in a mixture with liquid oxygen.
(41}
Liquid Oxygen Blastv '. First pilot operations with the use
of liquid oxygen blast were carried out in 600-ton furnace No. 10 at the
Kommunarsk Metallurgical Plant. During a two-month period, 32 heats
were melted producing 20,000 tons of steel. The use of liquid oxygen
reduced iron losses to 90-92 kilograms per one ton of steel, reduced the
amount of particulates in waste gases by 30 to 35 percent and increased
the steel output by 1-2 percent.
V. Pereloma revealed plans to put in operation the first
commercial installation during the second half of 1973 when special
cryogenic equipment will be available.
-------
45
5.2 EOF Furnace
The EOF (Basic Oxygen Furnace) process, or BOP (Basic Oxygen
Process), as it is sometimes referred to, utilizes a pear-shaped steel
shell lined with refractories to contain the molten iron and scrap used
in making steel. The usual charge consists of scrap, molten pig iron
and flux. After the materials are charged into the EOF furnace, a water-
cooled oxygen lance is lowered into the mouth of the vessel and high-
purity oxygen blown into the charge. The high velocity of the oxygen
stream (approaching super-sonic speeds) impinging on the surface of the
tnolten metal, produces a violent agitation and intimate mixing of the
oxygen with the molten iron. Rapid oxidation of carbon, silicon and
manganese in the iron, and reduction of sulfur content by the chemical
action of the flux produces a heat of steel. Violent agitation of the
molten bath during the flowing process results in the oxidation of very
fine particles of iron, and the ultimate formation of the fume associa-
ted with EOF steel making. The first EOF was placed in operation in the
United States in 1957. Since that time the EOF process has become the
major steel producing process in the United States, accounting for
83,260,000 net tons in 1973, for 55.3 percent of the total 150,431,000
net tons of steel produced. Research on the various factors affecting
the formation of fume are the same as those described in the section on
/OQ O 1 OO
open hearth furnaces and reported in the published literature '*'
33,34,35,36).
5.3 Q-BOP Process
Work has not been reported concerning any efforts to improve the
collection characteristics of BOF steelmaking fume. However, the "Q-BOP
Process" developed in Germany to improve production capabilities has re-
sulted in a secondary benefit of lesser fume evolution during the steel-
making process.
p, i
5-3.1 State of Development - Q-BOP Process
The modified bottom-blowing process was developed and placed
into commercial operation by the Maximillianshuette Iron and Steel Company
-------
46
(42)
in West Germany, during 1967-1968 . The process was developed primarily
to produce steels from high-phosphorus pig irons which are predominant in
Europe. It is called the OEM (Oxygen Bottom-blown Maxhuette) process and
(43 441
is covered by United States Patent Number 3,706,549 . Process
modification consists of shrouding high-purity oxygen, blown into the
vessel through tuyeres in the bottom of the furnace, with a hydrocarbon.
The vessel is shown schematically in Figure 6. In 1971, officials and
technologists of the United States Steel Company visited Maxhuette,
undertook development work of their own directed toward the production of
steel from low-phosphorus pig irons, which are predominant in the United
(42)
States . In December 1971, the U.S. Steel Corporation announced th^
(45)
construction of a Q-BOP plant at their Fairfield, Alabama works ,
which was followed by an announcement in early 1972, that the conventional
BOP shop under construction at Gary, Indiana, was to be converted to a Q-BOP
shop
(48)
A very good description of U.S. Steel Corporation's development
(49)
work and plant installations is given in the published literature
Tuyeres
Steel Shell
Refractory
Lining
Refractory
Plug Bottom
Hydrocarbons
or Shielding
Gas
Oxygen
Figure 6. OEM Process Vessel, U.S. Patent 3,706,549
-------
Availability to industry. The availability of the Q-BOP, and
a few similar processes, to the industry is somewhat clouded by the
patent situation of the several steel companies involved in the various
process developments^ . This is especially true with extensive patent
litigation which started in 1966 and is still in progress^51'52'53»54>55>56)
with the conventional EOF process.
Acceptance by Industry. The acceptance by industry is of
necessity influenced by the patent situation mentioned above. However, if
the patent situation were to become clarified, it is possible that within
5 years, eight of the remaining 21 open-hearth steelmaking shops could be
replaced by Q-BOP type steel plants. Three will be closed. The remaining
10 open-hearth shops have furnaces that are of rather recent construction
and have had considerable air pollution contol equipment installed, would
probably still be operated until replacement was required. This replacement
would probably take place over a 5 to 15-year period from now.
5.3.2 Degree of Effectiveness
No published reports have been made of the Q-BOP process that
Provides quantative information on the reduction in particulate emissions.
Reports indicate that the bath is quieter, the particle content of the
waste gases less, but coarser, and that quantities are only 1/3 to 1/5
that of the BOF furnace.
5.3.3 Environmental Effects.
More definitive information on off-gases and particulates from
the Q-BOP steelmaking process must be made available before an accurate
evaluation of the effects on the environment can be made. However, if the
reported reductions in particulate emissions bear out early indications,
the only effect on the environment can be that of an improvement.
-------
48
5.4 Electric Arc Furnace
The electric-arc furnace is a short cylindrical-shaped furnace,
having a rather shallow hearth. Three carbon, or graphite, electrodes
project through the roof into the furnace. Electric energy passing
through the electrodes and into the charge create the heat required to
melt the charge. Furnaces are constructed with fixed or moveable roofs,
with the vast amjority in the United States having moveable roofs.
Charge materials usually consist of 100 percent steel scrap,
with the exception of one major steel plant that uses molten pig iron as
part of the charge. Prepared scrap is loaded into charging buckets in
advance of the melting operation and charged as required. The charging
operation consists of opening the top of the furnace, lowering the
charging bucket part way into the furnace and dropping the scrap into
the furnace. The roof is moved back into position and the electrodes
lowered through the roof for the start of the melting operation.
The melting operation is started by turning on the electric
power to the electrodes. Arcing occurs between the electrodes and scrap
as the electric current passes into and through the scrap. When the
scrap is almost completely melted a second scrap charge is added, followed
by a third or forth scrap charge. The number of scrap charges depend on
the apparent density of the scrap as it is charged to the furnace. Re-
fining is accomplished by blowing high-purity oxygen into the molten steel
to remove carbon and silicon. This, combined with the refining action of
the slag, brings the heat of steel close to its required composition.
Ferroalloys are added to achieve the final composition, power is shut off,
and the heat poured into a ladle for transfer to the ingot pouring area
or continuous casting machine
There has been no reported work concerning the increase in
particle size or the reduction of emissions as applied to electric-arc
furnace operation. One process modification reportedly does result in
the apparent reduction in fume, that of using an oxygen-fuel burner for
preheating a more dense scrap charge to eliminate successive opening of
the furnace to charge the less-dense scrap charges.
-------
49
5.4.1 Preheating and Melting with Oxygen-Fuel Burners
Most of the work on preheating has been done with respect to
preheating scrap prior to charging it into an electric furnace. This
type of operation is considered more favorable since it reduces the
actual time required to complete a heat of steel in the furnace. Develop-
ment work has been done on the use of oxygen-fuel burners to preheat and
melt scrap in the electric furnace prior to the application of electric
power(57>58>59>60).
5.4.1.1 State of Development
The required equipment and operating practices have been developed
to the point where they can be used in the routine production of steel.
Availability to Industry. There are no apparent restrictions
on the availability of this technology, unless there may be some licensing
fees connected with use of some of the specialty burners designed for
the process.
Acceptance by Industry. One of the principle deterrents to
the use of oxygen fuel, preheating and melting technology has been the
increase in price of hydrocarbons. A second deterrent has been the high
noise level created by the burners which are designed to obtain maximum
efficiency of heat output.
5.4.1.2 Degree of Effectiveness
No reports have been published which contain data pertaining
to measurements of particulates, which would permit a good evaluation to
be made of the effects of oxygen-fuel burners on the reduction of
particulates. In the discussion of a technical paper given on oxygen-
fuel preheating, when the speaker was questioned with respect to fume
occurence, made the following reply, "The amount of fume appears to be
less than normal. There is, of course, an increase in waste gases. In
one case where the burners were used before the power was turned on,
the exhaust gas coming out of the electrode ports was almost clear.1
Two reports on the technology contained photographs to illustrate the
before and after situations ' .
-------
50
5.A.1.3 Envrionmental Effects
From the meager amount of information reported it does not
appear that there would be any particular detrimental effects to the
environment from emissions. However, as stated earlier there is a
problem in the noise created by the high-efficiency oxygen-fuel burners.
The gas and oxygen emerging from the tip of the burner are traveling at
supersonic velocity. Although no reports have been made of the problem,
the high noise level situation has been verified in conversations with
operating plant personnel in past visits to electric-arc furnace steel
plants.
5.4.2 Electric-Arc Furnace--Scrap Charge Compatibility
One of the factors tha t causes an increased amount of particulate
emissions is the number of times the furnace must be opened to charge the
amount of scrap required to meet the total weight requirements. If a
furnace could be charged with a consistent, high-density scrap charge,
the furnace would need to be charged only once. A method for producing
such high-density scrap charges for electric-arc furnaces has been
(f\"\ "4
developed in Japan
5.4.2.1 State of Development
The technology involves the use of a special press to compact
low-density scrap into a single high-density scrap charge tha t conforms
to the control of the inner volume of the furnace. Scrap charges from
5 to 60 tons have reportedly been produced.'
5.4,2.2 Availability to Industry
The special presses for providing the customized scrap charges
are being marketed in the United States by ai U.S. equipment broker.
Unless there are some specialized design concepts involved, similar
equipment could probably be designed and constructed by U.S. press builders.
-------
51
5.4.2.3 Acceptance by Industry
There should be no hinderance to the acceptance of this technology,
unless the charging cranes in the electric furnace shop have insufficient
capacity to handle a complete furnace charge. In such a situation, a
scrap crane of large capacity would have to be installed and the support-
ing crane structure would have to be strengthened. The costs for such
modification of the crane and structures could rule out the use of special-
ized scrap compacting procedures.
5.4.2.4 Degree of Effectiveness
No information has been published pertaining to the air pollution
aspects of this technology.
5.5 Metallurgical Coke Ovens
Particulates originate in coke oven operations from three sources,
(1) charging of the ovens with coal, (2) pushing incandescent coke from
the oven, and (3) the quenching operation. There are no published reports
on technology or process modification that pertain to the agglomeration
of particulate emissions from coke ovens, or a reduction in particulates.
There is only one report of record pertaining to the characterization of
particulates from coke ovens . Another report on the collection and
analysis of particulate emissions during the coke-oven charging operation,
is due for release in one and a half to two months^ . Efforts to obtain
any information from the contractor were unsuccessful.
5.5.1 Particulates from Charging Coke Ovens
Moist, pulverized coal is charged into coke ovens from larry
cars. The ovens are at an incandescent heat and the initial amounts of
coal dropped into the ovens are heated very rapidly with a resulting
evolution of steam and volatile hydrocarbons. The hot expanding gases
rise very rapidly and exit the oven through the charging hole at the
start of the charge, are suppressed during the flow of the bulk of the
coal and then may again exit through the charging port at the end of the
-------
52
charging operation, before the lid can be replaced. However, if the
coke-oven battery has double collecting mains with sufficient suction
in the steam aspirator system, the gases and particulates do not reach
the atmosphere and particulate emissions are minimized. Larry cars with
integral scrubbers and other process modifications serve to improve the
efficiency of particulate collection, but they in no way cause an
agglomeration of particulates or a reduction in particulates from the
charging operation itself.
5.5.2 Particulates from Pushing Coke
Particulates generated during the pushing operation occur
primarily from the abrasion of the coke on the refractory of the oven.
The particles are hard, angular carbon. The characteristics of these
particles do not permit any agglomeration to occur. When materials of
this nature are bonded together, such as in the manufacture of briquettes,
a petroleum base material or tar is used as the bonding agent. Some
times very fine particulates may occur during the pushing operation.
These particulates originate from coal that is insufficiently coked.
It is often referred to as "green coke". Extended coking time is a
suggested remedy for eliminating the fine particulates. The extension
of coking time is not a remedy for these emissions. Incomplete coking
is caused by insufficient heat transfer which is caused by warping of
the oven walls, in the case of older ovens, incomplete combustion in the
flues, and many other structural changes that may occur in an oven with
age, that affect the heat transfer characteristics of the ovens. Main-
tenance procedures may reduce the amount of emissions, but this is
questionable. Equipment is presently being installed on coke-oven
batteries to collect the emissions that occur during pushing. It will be
some months before definite results will be available from these install-
ations.
5.5.3 Particulate Emissions During Quenching
The emissions occuring during the quenching operation are
essentially the same type of particulates generated during the pushing
-------
53
operations. They are carried into the atmosphere by the heavy clouds of
steam generated by the quench water spraying onto the incandescent coke.
These particulates can be minimized by the use of baffles in quench
towers, which trap the particulates and carry them into sumps for
recovery from the water. Special quench cars, incorporating self-
contained water spray and emissions control devices, are in the process
of evaluation on full scale, operational equipment.
-------
54
6. Ferroalloy Furnaces
Ferroalloys are used to make additions of alloying elements to
iron and steels, in order to achieve desired mechanical properties. At
one time ferrosilicon and ferromanganese were both made in blast furnaces,
but today only ferromanganese is made in blast furnaces, and that at
only two plants in the United States. All the other ferroalloys are
produced in submerged electric-arc furnaces or by alumino-thermic
techniques.
Considerable work has been reported in the past five years with
. . - £ ,, f (66,67,68,69,70,71,72,73,
respect to emissions from ferroalloy furnaces >>>>>»>»
' ' ' ' . Although the reports contain a great deal of informa-
tion pertaining to the chemical analysis of the particulates, some data
on size, and material describing the operation of the various types of
air pollution control equipment, there is an absence of data concerning
the properties of the emissions that may lead to the evolution of fewer
particulates or possibilities of agglomeration that would make collection
easier or more efficient. One study did make a comparison between the
(73)
type of ore charged, i.e., lump ore, fine ore and pellets . Using
pelletized ore concentrates in comparison to fine ore concentrates the
amounts of particulate emissions was reduced approximately 42 percent,
while the use of lump ores reduced the amount of emission only 22
percent. The extent of pelletizing ores in the United States is not
known; however, an economic evaluation would have to be made to compare
the economics of crushing, grinding and pelletizing ores, against the
cost of air pollution control without pelletizing.
-------
55
7. Process Modifications For Particulate Control
In The Primary Nonferrous Metallurgical Industry
This technical review and discussion is limited to consideration
of the primary zinc, copper, and aluminum, industries. It is concerned
only with possible changes in present procedures that will aid in control-
ling fine particulate emissions.
7.1 Zinc Roasting, Sintering, and Distillation
Of the primary zinc plants in the United States, the only hori-
zontal retort plants presently operating (Amarillo, Texas, and Bartlesville,
Oklahoma) are doing so on variance for a limited time. Since no changes in
equipment or operation are contemplated or likely for these, they will not
be considered in this review. There are five plants remaining, then, for
consideration—three electrolytic and two pyrometallurgical.
The trend in American primary zinc plant construction has been,
and is expected to continue to be, toward the hydrometallurgical or electro-
lytic process rather than toward pyrolytic processes. Two new plants
definitely planned by American Smelting and Refining Company and by National
Zinc Company, respectively, will be electrolytic. Also announcement has
been made recently by New Jersey Zinc Company of plans to build an electro-
lytic zinc plant. This will give a preponderance of zinc production to
hydrometallurgical plants where control of fine particulates is a problem
only in roasting and handling.
7.1.1 Roasting
Particulates in roasting zinc concentrates are bourne with the
gas stream through flues and in some cases over waste heat boilers to dust
collection and scrubbing systems. In all plants under consideration the
gases go to an acid plant and are well cleaned before passing to the atmos-
phere. As a consequence, the roasting operation presents a very minor
factor in fine particulate control.
-------
56
7.1.2 Sintering
The two pyrometallurgical zinc plants sinter roasted calcine
and there is some stack loss of fine gas-bourne participates. This is
minimized by water sprays and a dust collection system. No new process
modification has been suggested or is presently being investigated to our
knowledge to reduce this small loss still further.
7.1.3 Reduction and Distillation
In reducing and distilling zinc from the two pyrometallurgical
plants there is very little production of fine particulates that can escape
to the atmosphere. Gases from the retorts carrying zinc metal vapor pass
through molten zinc or through a chamber filled with a rain of molten zinc
droplets to condense the zinc. This is followed by a scrubbing system to
collect zinc oxide that may be formed. Since there is no passage of gases
through the retorts, other than vaporizing of reduced zinc, there is
extremely small carry-over of charge; hence, about all of the particulates
that emerge from the retort-condensation system is zinc oxide which is too
valuable to lose. Present methods of controlling loss of particulates
from the pyrometallurgical reduction furnaces are effective and no better
system for improvement appears to be emerging other than greater attention
to good plant housekeeping in handling dry concentrates and calcines.
7.2 Copper Roasting,. Matte Smelting, and. Converting
There are two general methods under consideration for modifica-
tion of emitted solid particles in copper plants. One is to modify the
usual reverberatory-converter procedure and equipment; the other is to
eliminate pyrometallurgical treatment entirely. A brief review of conven-
tional U. S. primary smelter practice with data on the degree of efficiencies
realized with current control practices is given in Appendix A.
-------
57
7.2.1 Pyrometallurgical Modifications
The chief smelter modifications presently being used to reduce
emission of particulates are the result of efforts to reduce or largely
eliminate emission of sulfur oxides. In copper smelting elimination of
sulfur gases is of far greater importance than reducing stack losses of
particulates. Consequently, the trend that largely affects emission of
particulates is that of controlling sulfur gases. All copper smelters now
have electrostatic precipitators and/or baghouse units, as well as settling
flues, to recover solids, since it is to their economic benefit to avoid
such losses. However, to capture the sulfur, as by passing the gases
through an acid plant, greater attention is needed to clean the gas stream
and fewer particulates escape to the atmosphere. This has meant that
efforts to raise the S02 content of the gases evolved to permit effective
acid recovery, or combining the reverberatory-converter operations, also
aids in controlling particulate emissions.
7.2.1.1 Flash Smelting(79'8°'83)
The type of flash smelting developed at the Outokumpu smelter
in Finland (Fig. 7) or the variation developed by International Nickel
Company in Canada (Fig. 8) is a favored replacement method over the present
reverberatory. Flotation concentrates, with flux and preheated air, are
injected into a hot chamber where the flash burning of sulfides in suspension
furnishes the energy needed for smelting. This eliminates the need for
coal, gas or oil burners with attendant large volumes of products of combus-
tion. In fact, with the Outokumpu flash furnace a 14 percent SO' gas can
be produced (8 to 12 percent is more common which is 15 to 25 times more
than a reverberatory furnace using air-fuel combustion) which is excellent
for making sulfuric acid. The Inco variation is to use 95 percent oxygen
in place of air which makes a still more concentrated S02 gas, i.e., 75 to
80 percent. This is used in Canada to make liquid S02 for pulp and paper
plants in the general vicinity. 'Aside from excellent utilization of heat
and diminished gas volume, flash smelting is considered to give a higher
grade matte than reverberatory smelting. This requires minimum oxidation
-------
58
Preheated
air
Concentrate Concentrate burner
OUTOKUMPU FURNACE produces 14% SO2 otfgas, an
ideal gas grade for sulluric acid plants
FIGURE 7.
Sand Chalcopyrite
I concentrate
Constant
weight feeder
Oxygen ^- — —
Pyrrhotite, chalcopyrite
concentrates and sand
^ Oxygen
Slag
Matte
INCO FLASH FURNACE utilizes 95% oxygen for combustion of concentrates, produces 75% to 80% SO, offgas
FIGURE 8.
-------
59
in the converter. Consequently, a stationary converter can be used with
a more manageable gas containment system. Where there are outlets for
sulfuric acid or for liquid S02, flash smelting is deservedly receiving
considerable attention. A disadvantage of flash smelting is that both the
reverberatory and converter slags must be slow cooled, ground, and treated
by ore concentrations methods for recovery of copper values.
An Outokumpu type flash smelting furnace has been estimated to
cost $1,600,000 (1973) and an electric furnace for slag treatment, $1,100,000.
The cost for an entire plant with waste heat boilers, dust collection
system, site preparation, etc., would be on the order of $9,000,000.
Generally, the capital cost for a flash smelting unit is considered to be
higher than for a conventional reverberatory furnace. An Outokumpu flash
furnace is planned for the new Phelps Dodge smelter in Hidalgo County, New
Mexico.
(79,80)
7.2.1.1.1. Electric. Furnaces
Where the cost of electric power is not excessive, the use of
electric furnaces is also a recognized and well-developed method of getting
a more concentrated SO™ gas stream suitable for making sulfuric acid.
Since there are no products of combustion for gas dilution, an S02 workable
range of 2 to 4 percent can be attained. Usually this can be combined
with converter gases for the acid plant.
Electric furnaces for copper smelting were pioneered in Norway
and Sweden, and there is an excellent installation in Uganda. In the
United States the first commercial electric furnace for copper smelting
was started in early 1973 at Copperhill, Tennessee. Another, by Inspiration
Consolidated Copper Company, was started at their Arizona plant, adjacent
to their conventional reverberatory smelter, in November, 1973. This unit
is a 51,000 KVA furnace (operating on unroasted but thoroughly dried con-
centrates), five large Hoboken-type converters and a Lurgi double contact
acid plant. Construction is expected to begin in early 1974 at Anaconda,
Montana, on a fluid-bed roaster and electric furnace installation to
eventually replace the present reverberatory furnace operating on unroasted
concentrates.
-------
60
An electric furnace that is smelting roasted concentrates is
considered to have some advantages over a reverberatory in that a higher
concentration of S0_ is secured by the flash roaster--electric furnace
combination; an electric furnace operating on raw dry concentrate also has
some advantages, as a lower capital cost for the simpler construction of one
unit, but many factors affect the choice. Both methods tend to reduce
the amount of particulates evolved to the atmosphere by producing less
volume of gas flow and a gas that can be utilized in an acid plant.
1.2.1.2. Continuous Smelting
Combining the present reverberatory-converter operations into
a single unit has attracted much experimental work in the past few years.
Such simplified equipment, making only one stream of gases carrying parti-
culates has the potential of greater control over particulate losses. A
number of processes in this category have progressed to the pilot plant
stage and a few to initial commercial or semi-commercial installations.
The Noranda Process. ' Developed by Noranda Mines, Ltd. in
Canada, this process uses a single long (about 70 feet) combined smelting-
converting unit. It is a horizontal, cylindrical, furnace having a central
depressed area for copper collection and a raised hearth at one end as
shown in Fig. 9. A burner heats the smelting end where concentrates and
flux are charged, but air or an air-oxygen mix is introduced through sub-
merged tuyeres along the furnace base to oxidize the matte that is formed.
The furnace can be tilted for access to the tuyeres. Gases with accompanying
particulates may run 5 percent S0« or higher as oxygen enrichment above
25 percent is used. Thus, there is only one unit to control, without the
fume and trouble of conveying molten material from one furnace to another,
and the exit gas is of sufficient S02 concentration for use in making acid
with attendant thorough cleaning of particulates. After thorough testing
on a 100-ton per day pilot plant, Noranda has built a $19,000,000 initial
commercial plant of 800 tons/day capacity which was started in 1973.
-------
61
Burner.
Concentrate
and flux
-»
D
Side view
r.r\
1 -, C
Nitrogen and SO
1 -^ I—
OlJt
Reducing gas
Smelting
and converting
White metal ' Copper
I converting I settling
I I
-
Burner
Copper
Slag reducing
and settling
• Reducing gas
tuyeres
PROCESS produces rich off-gases to facilitate SO2 recovery-
FIGURE 9.
-------
62
Kennecot't has also announced plans to install a Noranda type continuous
(85 )
converting and smelting unit.
/Q I p £\
The WORCRA Process. * ' Developed by Howard Worner and Con-
ZincRio Tinto of Australia, this process has also undergone considerable
pilot testing over the past few years. It combines smelting, converting,
and slag cleaning in one operation, making metal directly and continuously
in a long stationary furnace. Slag is removed at one end; molten copper
at the other end. Off gases in a single stream are said to range from
9 to 12 percent SCL which can be efficiently handled in an acid plant.
By utilizing the exothermic reactions efficiently in burning the sulfides
in the charge, a minimum of energy is required. No tuyeres are used as
in the Noranda process, but air lances from the top are used for converting.
Although the process has been tested in pilot plants and semicommercial
plants of up to 72-80 tons of concentrates per day, no large commercial
plant has been constructed so far. Some disadvantages may be the rather
low-grade blister copper produced which requires considerable fire refining,
and the probability of short refractory life because of the vigorous bath
agitation.
(87^
The Q-S Process. Named after the inventors, Paul Queneau
of Dartmouth and Reinhardt Schumann of Purdue, this is also a multistage,
progressive converter operation that combines continuous smelting and
converting into one furnace. Sulfide concentrates are flash smelted by
oxygen, and oxygen is also used through submerged tuyeres to effect conver-
ting to copper. Dust production is said to be minimized by this procedure,
but as the SCL rich gas is thoroughly cleaned anyway for recovery of sulfur
in some form, this is of minor importance from the standpoint of particulate
control.
A slag scavenging operation is part of the process. This involves
bottom blowing the slag with coal, oxygen, and sulfur dioxide to recover
residual copper, presumably as a matte that can be recirculated. The
process can be applied to the recovery of nickel and lead also; in fact,
the St. Joe Minerals Corporation is said to have obtained an option for the
exclusive use of the process for replacing sintering and blast furnace
operations in smelting lead concentrates. Pilot plant investigations of
the process are planned.
-------
63
(81 82 88
The Mitsubishi Process. * ' ' ' Developed in Japan by
Mitsubishi Metals Corporation, this differs from the Noranda and WORCRA
continuous processes in that it is a continuous process but the work is not
all done in a single unit. Rather, the concentrates are smelted in one
furnace, the slag and matte flow continuously through a slag cleaning
furnace to a converting furnace having overhead air lances, as shown in
Fig. 10. Since these are interconnected, there need not be the loss of
fumes that usually exist in transferring from reverberatory to converter
and the continuous converting unit is not cooled intermittently, which
favors better refractory life than with the usual converter blows. Out-
standing features are said to be economy in capital investment (compared
to a reverberatory-converter unit) and operating costs, and exit gases of
over 10 percent SO- which indicates a concentrated or low-volume gas stream.
The dusting rate is said to be low because of liquid particle entrapment
in the furnace bath. This should further decrease loss of particulates in
the gas stream. Charge preparation in drying to 1 percent moisture and
instrumentation control are considered to be complex. The process, after
thorough testing in a pilot plant, has been further demonstrated in a
1,500-ton blister copper per month semi-commercial plant that was started
in November, 1971.
7.2.1.3. Miscellaneous Pyrometallurgical Developments
Although the trend in building new pyrometallurgical copper plants
throughout the world has been toward flash smelting, electric furnaces
and continuous smelting systems, no discussion of the subject would be
complete without at least mention of other methods which may be considered
minor or presently not definitely proven to the commercial stage.
(79)
The Momoda Blast Furnace.v J Successfully developed by Sumitomo
Metal Mining Company, the Momoda blast furnace is being used in two copper
smelters in Japan, It appears to be adaptable to, and have advantages for
-------
Returning by moving bucket system or air lifting
Revert slag
Converting furnace
Slag granulation
Schematic view of Mitsubishi's semicommercial plant.
FIGURE 10.
-------
65
small-scale operation, such as a capacity of 500 tons of charge per day
per furnace. Since the energy requirement for smelting a ton of charge
in a Momoda furnace is said to be only about 28 percent that for reverbera-
tory smelting with a wet charge, there is economy in operation, at least
under Japanese conditions. A feature of the process is that the concentrates
are charged with other copper bearing materials as a stiff plasticized mass
containing 10 to 15 percent water, rather than being sintered or briquetted.
The process has been proven successful abroad but has not been tested in
this country and would appear to have no direct benefit from the standpoint
of greater particulate control.
The U. S. Bureau of Mines. ' Autogenous smelting work has been
done on a laboratory scale and no commercial development of the process
appears to be pending, but this procedure has received considerable favorable
attention. Like the Noranda and WORCRA processes, it is a continuous
smelting method in a single unit to produce copper directly from concentrates.
Oxygen is used in place of air which gives a concentrated, low-volume, high
SO- gas. This should be favorable for minimum particulate loss since the
gas would be thoroughly cleaned before going to a plant for sulfur recovery.
The furnace combines flash smelting with converting by means of an oxygen
lance immersed through the slag into the matte from the top. A simplified
sketch of this furnace is shown in Fig. 11.
/o i n I \
The Kircet Process. ' The Kircet Process is a development
in the U.S.S.R. which has been demonstrated there on a pilot-plant scale.
It consists of charging a dry copper concentrate into a cyclone type furnace
where an air-oxygen mixture (of up to 100 percent oxygen) affects oxidation
and smelting similar to flash smelting. An electric furnace attached to the
cyclone smelting furnace receives the molten product to complete the treat-
ment. A second electric furnace receives matte from the main unit and
produces blister copper and slag. Since the process is designed to treat
combined lead-zinc-copper ores, liquid zinc is recovered by condensation
in the gas stream. Incomplete descriptions give few details of how this is
done or how lead is recovered. Although a saving is claimed in being able
to float lead-zinc-copper values from the complex ores, the complexity of
recovering all three metals in a single furnace would seem to be disadvanta-
geous. Off-gases from the system are said to run 70 to 85 percent S02
-------
66
Bureau of Mines Autogenous Smelting
Outer furnace FeS3
flue
Slag overflow
White metal Copper
Verticle.elevation section through center
Copper overflow
FIGURE 11.
-------
67
(using oxygen in place of air) and only a small amount of reducing material
is needed.
Other Process Modifications. Among general improvements of present
installations, the use of oxygen enriched air to reduce the volume of gas
emissions, and attention to hooding converters to control more efficiently
gas and dust emissions, have been common and need not be detailed.
The top-blown rotary converter as developed by International
(79 81 82)
Nickel Company has followed oxygen steel-making practice. ' ' ' Although
currently a commercial installation is being made at Copper Cliff, Ontario,
to treat a mixed nickel-copper charge, this procedure is considered to be
of potential benefit for handling copper matte alone where low-cost oxygen
is available. There are metallurgical advantages to using such a converter,
such as excellent mixing, less refractory wear, and a higher temperature
in final conversion to blister copper, but for particulate control the only
advantage is in the reduced volume of gases from using oxygen.
Improvements in removal of sulfur from the gas stream in copper
plants indirectly aids removal of particulates since most procedures require
cleaning the gas stream more thoroughly than when it merely goes up the
stack. Thus, Monsanto1s Cat-Ox process which has been tested to remove S00
from power plant gases may be adaptable to dilute SO- gases presently
evolving from copper smelters. The Wellman-Power Gas procedure uses "sodium
sulfite solution to absorb S02» forming sodium bisulfite, then reversing
the reaction to release concentrated SO '.' As with Monsanto1 s Cat-Ox process,
the gas stream must be thoroughly cleaned before reaching the absorber;
hence, particulates of all sizes would be removed. Allied Chemical Corpora-
tion has been developing another procedure which reduces a comparatively
rich S02 stream with natural gas to produce elemental sulfur. ' This
could apply to a 12 percent or over S0_ gas as produced by other methods,
including the Wellman-Power or Monsanto Cat-Ox methods, where production
of elemental sulfur was much to be desired over producing acid. Also, the
Bureau of Mines has been operating a small pilot plant, and extending their
operations on a larger scale to a lead plant and coal-fired power plant on
removal of S07 from low-concentration gas streams by absorption in a sodium
(93)
citrate solution. Here again, of course, the gas stream is thoroughly
cleansed of particulates by passing through electric precipitators or
-------
68
baghouses and being water scrubbed before reaching the absorber. American
Smelting and Refining Company has had considerable experience in recovering
liquid SC>2 from copper smelter fumes, and is building a sizeable plant to
do this.W
In place of absorbing SO- from dilute gases, limestone scrubbing
has received considerable attention. The Smelter Control Research Assoc-
iation in pilot plant tests, have found the wet limestone scrubbing process
to be unreliable and have put emphasis instead on the ammonia double-alkali
process for the removal of S02 from copper reverberatory furnace gas.
7.2.2 Hydrometallurgical Copper Recovery
The effort to avoid air contamination with S0_, and, to a lesser
extent, with dust, has led many copper companies to consider hydrometallur-
gical treatment. Many different procedures have been investigated and
some have reached the commercial plant stage. This is in addition to the
well-developed present technology of leaching oxidized ores or roasted
concentrates with sulfuric acid and recovering the copper by electrolysis
or precipitation on iron scrap. Leaching fractured ores in place, heap
leaching, and treatment of mine waters with iron scrap to recover copper
are important procedures but beyond the scope of particulate control.
Improvements in using ion exchange or solvent extraction to isolate and
concentrate a desired metal from solution has enhanced the possibilities
of useful and economic recovery of metals by hydrometallurgy but such means
have only indirect effect on control of particulates.
(95)
7.2.2.1 Stanford LCPR Processv '
A lime-concentrate-pellet-roast procedure investigated by the
Process Metallurgy Group at Stanford University is noteworthy in that it
combines a pyrometallurgical advancement with hydrometallurgical extraction
and recovery of copper. The general plan, as shown in the flowsheet of
Fig. 12, is to tie up the sulfur with lime during roasting, then leach the
calcine with spent sulfuric acid electrolyte for copper extraction and
-------
69
LIMESTONE
COPPER
CONCENTRATES
WATER
(OPTIONAL)
SPENT
ELECTROLYTE
ELECTROWINNING
COPPER
CATHODES
TAILING
TAILING
BLEED
COPPER STRIPPING
ELECTROLYSIS
OR CEMENTATION
COPPER
Flowsheet (or the lime-conccntratc-pellet-roast
process with direct electrowinning
LIME OR
LIMESTONE
NEUTRALIZATION
TAILING
FIGURE 12.
-------
70
electrowtnnlng. Copper concentrates are palletized with lime and roasted
at 400-600 C whereby at least 90 percent and in laboratory tests, about
98 percent, of the sulfur is retained as anhydrite (CaSO,). Little fuel is
required, as the process is semi-autogenous, and the capital cost and
operating cost of a plant has been calculated to be favorable. The reaction
of lime to retain sulfur in roasting is not new but the pelletizing and
roasting conditions to obtain maximum results are an advancement. However,
this method has not yet been investigated on more than a laboratory scale.
Since most of the sulfur (95+ percent claimed) is retained in
roasting and is discarded as a tailings residue after leaching, and possibly
cyaniding, there is no need to clean the gases from roasting. Using a
traveling grate roasting operation as advocated, a large amount of fine
particles, including lime and anhydrite, would be expected to accompany
the gas stream. A bag house or electrostatic precipitation unit undoubtedly
would be used. This would collect most of the dust for recycling, but the
stack loss of parttculates obviously would be greater than in a strictly
hydrometallurgical plant but probably not as much as from a reverberatory-
converter plant that does not have an acid plant.
7.2.2.2 Anaconda's "Arbiter" Ammonia Leach Process^96'
This is an entirely hydrometallurgical process that has gone
through the pilot plant stage. The first commercial plant being built at
Anaconda, Montana, is scheduled to go into production in September, 1974.
A general flow sheet of this plant is shown in Fig. 13. The only particulate
control involved is in handling concentrates and lime, and in preventing
tailings from becoming windblown. These items are expected to be under
good control by recognized operating methods.
The Arbiter process, as it is often called, is essentially a
continuous ammonia leach with oxygen under only slight pressure for control,
followed by solvent extraction to isolate and concentrate the copper, electro*
lytic recovery of copper through a closed cycle of stripping the loaded
organic solvent with spent electrolyte, and disposal of ammonium sulfate.
-------
—ANACONDA ARBITER PLANT-
BLOCK FLOW DIAGRAM
Niti
i
Coi
G
Ov
To
rogen NH3
L Y
T
C
<
1
e
ibusj
ises
k
=5— -
irhe
Atrr
OXYGEN
"PLANT
Mon
BOILERS
ad
losphere
r —
AMMONIA
VENT
SCRUBBER
02
J
Feed Slurry .
""^ ""^ ** j
LEACH
REACTORS
i
Amr
1
Steam
nonia
*
Li
^
Raff
1
me
f\
AMMONIA
RECOVERY
1
.«,. CCD p. fV-TA-riQM ,
^_ THICKENERS
Ver
»
inate
%
Conce
it Gas To £
&
Residue
Y To Disposal
CLARIFICATION
*" FILTERS
Strong Solution
1 ,
.LIQUID ION ^ prTRnwiMWiMr 1
^ LLLVy J KUWmNINo
EXCHANGE
* i *
Gypsum Slurry
To Disposal
Spent Electrolyte
IOO-TPD
Copper Cathode
FIGURE 13.
-------
72
In the commercial plant provision is made for a "lime boil" or hot treat-
ment of the ammonium sulfate solution after copper removal by addition of
lime with subsequent regeneration of anmonia and production of a calcium
sulfate residue. This can be effected also by adding lime directly to the
leaching tanks to regenerate amnonia directly and pass the calcium sulfate
into the concentrate tailings. Alternate methods of ammonium sulfate
disposal *re to dry the solution and sell the crystals, to thermally reduce
the arrmonium sulfate to sulfur and recover ammonia, or to obtain bacterial
decomposition of the pmmonium sulfate solutions to produce ammonia and
elemental sulfur. The process is flexible in that comparatively low capa-
city plants can still be efficient and variations in operation can be made
readily to accommodate different concentrates or objectives. Ammonia
leaching has definite advantages, such as no solution of pyrite or iron
oxides and inexpensive materials of construction.
7.2.2.3 Sherritt:-Gordon Process
From its long and successful experience in recovering nickel
end cobalt by pressure-leaching with amnonia, the entrance of Sherritt
Gordon technology into the copper field would be expected. However, the
direct application to copper concentrates of ammonia leaching as practiced
at the Fort Saskatchewan, Alberta, plant has not been considered to be
economically attractive. Likewise, their procedure for precipitating metal
powders from solution by hydrogen under pressure has not been entirely
successful commercially as applied to copper.
Announcement was made in 1971 that Sherritt Gordon and Cominco
would jointly explore and develop a hydrometallurgical process for recovery
of copper and other metals, and elemental sulfur, from sulfide concentrates.
A procedure mentioned, one that has been patented by Sherritt Gordon, is
pressure leaching with sulfuric acid. This would appear to have an advantage
over the Arbiter process in recovering elemental sulfur directly, but pres-
sure leaching with acid would seem to be less desirable than near-atmosphere
leaching with ammonia.
-------
7V2,2.4 Cymet Process
73
(81,97,98)
The Cyprus Metallurgical Processes Corporation has developed a
ferric chloride leaching process for sulfide copper ores which is considered
to be essentially pollution free. The only step that may involve loss of
particulates, other than general handling of concentrates, is the possible
need for grinding the concentrates to facilitate leaching; otherwise, all
operations are hydrometallurgical. A flow sheet of the general process is
shown in Fig. 14.
This process is rather complicated as it involves leaching with
ferric chloride anolyte recycled from diaphragm iron-chloride cells,
production of electrolytic iron, recovery of elemental sulfur, and recovery
of copper electrolytically as a powder which is subsequently electrorefined.
The many interdependent operations which add complexity make development
difficult, but a large scale $9,000,000 pilot plant for demonstration of
practicality and cost has been built. Recovery of elemental sulfur and
of high grade iron are features of the process, and their value is expected
to aid substantially in reducing operating costs.
7.2.2.5 Duval Corporation and Other Processes
Typical of other entirely hydrometallurgical processes for copper
recovery, the Duval Corporation announced several years ago the development
(99)
of a pollution-free process for treating copper sulfide concentrates.
This also involves a chloride leaching, closed-cycle system with recovery
of elemental sulfur and electrolytic grade copper, but recovery of iron
oxide in place of electrolytic iron as done by Cymet. A pilot plant was
planned and presumably built but details of results have not been noted.
Other suggested procedures in recovering coppy hydrometallurgically
have been mentioned or briefly discussed in the technical literature in the
past few years. These have largely been variations of the processes out-
lined. °* In all cases, since they depend entirely on hydrometallurgy,
their effect has been the same--little or no air pollution from particulates.
-------
74
"
trVty^l | Q- ET-0
Schematic Ffowsheet"
Cymef Process
C>prus McUllurgical Process CorpO'at°
Los Angeles, California
FIGURE 14.
-------
75
7.3 Aluminum Reduction Cells
At present there are 14 U.S. producers operating 32 reduction
plants which, in 1970, produced nearly 4 million tons of aluminum,
46.6 percent of the world total' .
The processing of primary aluminum is essentially one of
materials handling, the difference in the process modifications is in the
type of reduction cells used. There are three types, prebake (PB) cells
which use prebaked carbon anodes, and two types of Soderberg cells which
use single, baked-in-place anodes. One type of Soderberg anode is held
in place with horizontal studs (HSS) and the other with vertical studs
(VSS). Schematic drawings of each type are shown in Figures 15, 16 and 17.
In plants using prebaked (PB) cells, the anode baking furnace
used in manufacture the anode can be a source of particulate and gaseous
emissions. However, these may be controlled by electrostatic precipitators
or more commonly by wet scrubbers.
In operation, the prebake anodes in the reduction cell produce
less volitization of pitch and fouling of collection system scrubbers
than do the Soderberg types. Another advantage of the prebaked pot is that
it is more readily enclosed without interferring with process operations.
Prebaked (PB) potlines accounted for 59 percent of U.S. aluminum production.
In the course of operation with horizontal stud Soderberg cells
(HSS), the necessity for frequently changing the studs in turn makes
necessary the frequent removal of parts of the hoods with the consequent
escape of emissions. There is a further disadvantage in that the hooding
system does not allow for the burning of organics or condensed tars;
these unburned organics can cause fouling of the emission control system.
HSS Soderberg units account for 25 percent of U.S. production.
In the vertical Soderberg stud (VSS) pot, the anode compartment
is stationary, so it permits the installation of a "skirt" around the
base of the anode, and since the volume of gas is small, carbon monoxide
and hydrocarbons can be burned in integral gas burners as shown in Figure
1? • Since the skirt does not completely enclose the pot, additional
hooding is necessary to minimize fluoride emissions into the pot room.
v-S.s. Soderberg production units account for 13 percent of U.S. aluminum
Production.
-------
Alumina Hopper
Molten Crvolite
Segmented Doors
Handle
Aluraina
rj
rS
\\\ • • -v '^
\\V\\v
•
\A\\\\\
Primary
Control Svste:
Carbon Anode
ancle
.:•
Molten Aluminum
FIGURE 15. SCHEMATIC DRAWING OF A PREBAKED ANODE CELL
-------
Alumina Hopper
Carbon Anode
Alumina
V-^L-V^-''- ' •y//' /.•/.'
To Primary
Control System
Kood Door
Anode Studs
Molten Aluminum
Molten Cryolite
FIGURE 16. SCHEMATIC DRAWING OF A HORIZONTAL STUD SODERBERG ALUMINUM REDUCTION CELL
-------
Carbon Anode
Skirt
Exposed
Cell Surface
Molten Cryolite
Molten Aluminum
Anode Studs
Control Systen
Gas and
Tar Burning
FIGURE 17. SCHEMATIC DRAWING OF A VERTICAL STUD SODERBERG ALUMINUM REDUCTION CELI
-------
79
There is a continuous evolution of gaseous reaction products
and fume from the reduction cells consisting chiefly of carbon dioxide,
carbon monoxide, sulfur oxides, gaseous and particulate fluorides,
alumina and carbonaceous materials. Hydrocarbon emissions from hot
HSS and VSS anodes can also result in a visible haze problem. Table 1
shows total industry-wide emissions in 1970, Table 2, the emissions
from a typical uncontrolled pre-bake potline, and Table 3, the removal
efficiency of selected primary and secondary control systems.
Primary control systems involve the hooding and ducting for the indivi-
dual cells; secondary control systems involve pot room collection
systems located in the roof of the cell room to remove fluorides which
escape the primary system.
In 1970, in the U.S., 75 percent of the plants had primary
controls only, 15 percent had secondary controls only, 7 percent had
both, and 3 percent had none at all( '.
Collection efficiencies approaching 100 percent are not
possible under the present state of the technology (see Table 3).
The best achievable appears to be 95 percent delivered to the primary
system, with 5 percent going to the secondary roof monitors. '
7.3.1 New Aluminum Reduction Processes
There are a number of new aluminum reduction processes under-
going development. Among these are Alcoa's chloride electrolysis
process (the ASP process), Alcan1s sub-halide process, and the Applied
Aluminum Research Corporation^ (manganese reduction of aluminum chloride)
process^ ' . Furthest along is Alcoa's chloride electrolysis
process. . In this, alumina is reacted with chlorine in a reactor
to produce aluminum chloride which in turn is electrolyzed in a closed
cell separating the compound into aluminum and chlorine. The separated
chlorine is recycled continuously to the reactor in a closed loop
system. The process has the advantages of up to 30 percent less energy
consumption, and presumably less dusting since there are no fluorides,
the chief effluent in the conventional Hall process which causes the
-------
80
TABLE 1. EMISSIONS FROM PRIMARY
ALUMINUM INDUSTRY
1970 (1,2)
Pollutant
Total Huorino
Gaseous fluorides
Fluorine in particulars
Total solids
Emissions (Ions)
Pot- Bake
rooms plants
Total
?3,200
10.200
13.000
52.800
650 23,850
600 10,800
50 13,050
4,200 57,000
TABLE 2. EMISSIONS FROM AN UNCONTROLLED
PREBAKE POTLINE (1,2)
Pollutant
S0»
"F" as gaseous fluorides
F as solid fluoride's
Total F
Total solids
emissions.
Ibslton Al
60
28
18
46
92
TABLE 3. FLUORIDE REMOVAL EFFICIENCIES
OF SELECTED PRIMARY AND SECONDARY
CONTROLLED SYSTEMS (1,2)
Control System
Coated filter dry scrubber
Fluid bed dry scrubber
Inieetod alumina dry
scrubber
Wet scrubber • wet ESP"
Dry ESP • wet scrubber
Floating bed
Spray screen
Vcnlurii
Bubbler scrubber r
wel ESP
Fluoride removal
Dlliciencles. %
IIP
90
99
98
99
93
98
93-95
99
Panic-
ulate
98
98
98
99
90-95
87
45-85
96
Total
F
94
99
98
99 >
94-96
95
62-77
98
99
98
99
1 Electrostatic precipit.itor.
-------
81
industries a most serious air pollution problem, and no working of
finely powdered Al-0 into the bath. Also, the requirement for cryolite
(Na.Al F,) is circumvented at a time when this mineral has become
3 o
scarce and costly. There have been no claims as to the degree of re-
duction in dusting except for that implied in the Alcoa statement that
the shift from the Hall process to the ASP process, eliminates both
the need for cryolite and the expense of containing cryolite emissions"
A number of the other new processes have as their goal, the
utilization of alternative mineral sources of aluminum such as laterite,
kaolin clays, and alunite . The implementation of these processes
does not affect materially the amount and type of emissions from the reduction
cells.
-------
82
REFERENCES
(1) "Particulate Polycyclic Organic Matter, National Academy of Sciences,
Committee on Biologic Effcts of Atmospheric Pollutants (1972).
(2) "Abatement of Particulate Emissions From Stationary Sources",
National Research Council - National Academy of Engineering, COPAC-5,
Washington, D. C. (1972).
(3) Office of Coal Research, Press Release, July 3, 1973.
(4) "Evaluation of Coal-Gasification Technology: Part I--Pipeline-
Quality Gas", National Academy of Sciences, National Adademy of
Engineering (December, 1972).
(5) "Evaluation of Coal-Gasification Technology: Part II--Low-BTU
Fuel Gas", National Academy of Science, National Academy of Engineer-
ing (to be published in 1973).
(6) "Clean Energy from Coal—A National Priority", Amnual Report for
Calendar Year 1972, Office of Coal Research, U.S. Department of
Interior.
(7) Karnavas, J. A., LaRosa, P. J., and Pelczarski, E. A., "Ttoo Stage
Coal Combustion Process", Chemical Engineering Progress, Vol. 69,
No. 3, p 54 (March, 1973).
(8) Chementator, Chemical Engineering, August 6, 1973, page 34.
(9) Personal communication with representatives of Atomics International
Division,North American Rockwell.
(10) Cover, A. E., Schreiner, W. C., and Skaperdas, G. T., "Kellogg1s Coal
Gasification Process", Chemical Engineering Progress, Vol. 69, No. 3,
p 31-36 (March, 1973).
(11) Oldenkamp, R. D.,'and McKenzie, D. E., "The Molten Carbonate Process
for Control of Sulfur Oxide Emissions", APCA paper 68-183 (1968).
(12) Reid, W. T., "Superslagging Combustor", Battelle-Columbus internal
memorandum to D. W. Locklin confirming disclosure of concept
(April 2, 1973).
(13) Squires, A. M., "Clean Power from Coal", Science, 169 (3948)
(August 28, 1970), p 821.
(14) Jimeson, R. M., and Shaver, R. G., "Credits Applicable to Solvent
Refined Coal for Pollution Control Evaluations", p 25C, presented
at the Symposium on Synthetic Hydrocarbon Fuels from Western Coals,
AIChE Meeting, Denver, Colorado, August 30, 1970.
(15) Martin, G. B., Pershing, D. W., and Berkau, E. E., "Effects of Fuel
Additives on Air Pollutant Emissions from Distillate-Oil-Fired
Furnaces", U. S. Environmental Protection Agency (June, 1971).
-------
83
(16) Andrews, R. L., Siegmund, C. W., and Levine, D. G., "Effect of Flue
Gas Recirculatlon on Emissions from Heating Oil Combustion",
presented at 61st Annual Meeting, APCA, St. Paul (June 23-27, 1968),
41 pp.
(17) Bienstock, D., Amsler, R. L., and Bauer, E. R., Jr., "Formation of
Oxides of Nitrogen in Pulverized Coal Combustion", Jour. APCA, 16^
(8), (August, 1966), p 442-445.
(18) Burroughs, L. C., "Air Pollution by Oil Burners Measurable but
Insignificant", Fuel Oil and Oil Heat, 22 (6), (June, 1963), p 43-46.
(19) "HEW Issues Guidelines to Control Pollution; Tests Show Distillate
Within Limits", Fuel Oil and Oil Heat, 28 (3), (March, 1969), p 57.
(20) Kroshel, C. F., "Improving Performance of Domestic Heating Oil Equip-
ment Througti Installation and Servicing Procedures", presented
at the 40th Annual Convention, NOFI, Chicago, p 26, (April 12,
1962).
(21) Ornig, A. A., Schwartz, C. H., and Smith, J. F., "A Study of the
Minor Products of Coal Combustion", Paper No, 64-PWR-4, presented,at
National Power Conference, Tulsa, Oklahoma (September 27-October 1,
1964), 8 pp.
(22) Ornig, A. A., Smith, J. F., and Schwartz, C. H., "Minor Products of
Combustion in Large Coal-Fired Steam Generators", Paper No. 64-WA/FU-2,
presented at Winter Annual Meeting, ASME, New York (November 29-
December 4, 1964), .12 pp.
(23) Diehl, E. K., DuBreil, F., and Glenn, R. A., "Polynuclear Hydrocarbon
Emission from Coal-Fired Installations", Trans. ASME, Series A,
Jour. Engineering for Power, 89 (2), (April, 1967), p 276-282.
(24) NCRR Bulletin; (Summer, 1973), Vol. Ill, No. 3, National Center for
Resource Recovery, Inc., 1211 Connecticut Avenue, N.W., Washington,
D. C., 20036.
(25) Niessen, W. R., "Systems Study of Air Pollution from Municipal Incin-
erators, Vol. 1, Report prepared by Arthur D. Little, Inc. (March,
1970), under Contract CPA-22-69-23.
(26) Hazard, H. R., Trip report covering visit to Combustion Power Company,
Inc., on February 6, 1974.
(27) Miller, P. D., et al., "Corrosion Studies in Municipal Incinerators",
SW-72-3-3, EPA Solid Waste Research Laboratory, National Environmental
Research Center (1972).
(28) Environmental Science and Technology (April, 1971), Pyrolysis of
Refuse Gains Ground, p 310-312.
(29) Franklin, W. E., Bendersky, D., Shannon, L. J., and Park, W. R.,
"Resources Recovery, Catalog of Processes," Final Report, Project
No. 3634-D, Prepared by Midwest Research Institute for the Council
on Environmental Quality (February, 1973).
Two Battelle reports that served as general references for the stationary
combustion and municipal incineration sections of this report are:
-------
84
A. The Federal R&D Plan for Air-Pollution Control by Combustion-
Process Modification, PB-198-066, prepared for EPA (January 11,
1971).
B. Working Paper on Factors Affecting the Future of the Coal Industry
in the United States (April 1, 1973). Not for Distribution
Outside Battelle.
(30) Rengstorff, G. W. P., "Formation and Suppression of Emissions From
Steelmaking Processes, " AIME Open Hearth Proceedings, 44. 120-143 (1961).
(31) Rengstorff, G. W. P., "Factors Controlling Emissions From Steelmaking
Processes," AIME Open Hearth Proceedings, £5, 204-219 (1962).
(32) Turkdogan, E. T., Grieveson, P., and Darken, L. S., "Mechanism of the
Formation of Iron Oxide Fumes", AIME Open Hearth Proceedings, 45.
470-490 (1962).
(33) Bates, R. E., "Fume Formation", Journal of the Iron and Steel Institute
201 (9), 747-751, (September, 1963).
(34) Morris, J. P., Riott, J. P., and Illig, E. G., "A New Look at the
Cause of Fuming", Journal of Metals, J.8 (7), 803-810 (July, 1966).
(35) Rossi, G., and Perin, A., "Some Notes on Brown Fume Powders",
Journal of the Iron and Steel Institute, 207 (10), 1365-1368, (October,
1969).
(36) Ellis, A. F., and Glover, J., "Mechanism of Fume Formation in Oxygen
Steelmaking", Journal of the Iron and Steel Institute, 209 (8),
593-599, (August, 1971).
(37) Rengstorff, G. W. P., "Role of Methane and Other Factors in Controlling
Emissions from Steelmaking Processes", AIME Open Hearth Proceedings,
4j6, 438-452, (1963).
(38) Savard, G., and Campbell, J. C., "Suppression of Iron Oxide Fumes
in the Open Hearth, Furnace", 33/The Magazine of Metals Producing, J>
(2), 61-70, (February, 1967).
(39) Kocho, et. al., "Iron and Steel Refining—USSR Patent No. 236499",
Soviet Bulletin of Discoveries, Inventions, Models and Trade Marks
(March 2, 1969).
(40) Naydek, V. L., et. al., "Treatment of Metal—USSR Patent No.
276116", Soviet Bulletin of Discoveries, Inventions, Models and Trade
Marks, (July 14, 1970).
(41) Kiyanskiy, D., "Liquid Oxygen Blast", Rabochaya Gazeta, 45, (4912),
3 (February 22, 1973).
(42) "Q-BOP; From Blow to Go in 90 Days", Journal of Metals, 2A (3),
31-37, (March, 1972).
(43) Chatterjee, A., "The New Oxygen Steelmaking Process," Iron and Steel
International, .46 (5), 440-448 (October, 1973).
-------
85
(44) Knuppel, H., et. al., "Method of Refining Pig-iron Into Steel",
U.S. Patent Number 3,706,549, Patent Gazette, 905. 567 (1972).
(45) "U.S. Steel to Substitute Q-BOP for Open Hearths at Fairfield",
Metalworking News, 8 (December 20, 1971).
(46) "Q-BOPs for Gary", Journal of Metals, 2A (4), 8 (April, 1972).
(47) "U.S. Steel Managers Rave Over Q-BOP", American Metal Market, 1
(May 25, 1973).
(48) "Development in the Iron and Steel Industry During 1973," Iron
and Steel Engineer,..51 (1), D24 (January, 1974).
(49) Hubbard, H. N., Jr., and Lankford, W. T., Jr., "Development and
Operation of the Q-BOP Process in the U. S. Steel Corporation",
Iron and Steel Engineer, j>0 (10), 37-43 (October, 1973).
(50) "Bottom-blown Steel Processes Now Number Three: Q-BOP, LWS, and
SIP", 33 Magazine, 12 (9), 34-38, (September, 1972).
(51) "Kaiser Suing J&L Over L-D Rights", Metalworking News, 12
(November 27, 1967).
(52) "Oxygen Furnace Patent Fight to Continue", American Metal Market,
2 (March 10, 1969).
(53) "High Court Bars Oxygen Steel Patent", Metalworking News, 16
(March 10, 1969).
(54) Butler, P., "EOF Patent Fight Flames in Court Today", Metalworking
News, 11 (September 27, 1971).
(55) "J&L to Appeal Court Decision Charging Patent Infringement",
American Metal Market, 81 (23), 4 (February 1, 1974).
(56) "Q-BOP: Year II", Journal of Metals, 25 (3), 37 (March, 1973).
(57) Hinds, G. W., and Hodge, G. W., "Use of Oxygen-Fuel Gas Burners for
Scrap Meltdown in Electric Furnace", AIME Electric Furnace Proceedings,
J.7 290-298 (1959).
(58) Howard, V. J., "Auxiliary Meltdown Toarch", AIME Electric Furnace
Proceedings, .17, *19 (1959).
(59) Howard, V. J., "Use of the Oxygen Gas Burner for Scrap Meltdown in
the Small Arc Furnaces", AIME Electric Furnace Proceedings, 18. 398-405
(1960).
(60) Spenceley, G. D., and Williams, D. I. T., "Fumeless" Refining with
Oxy-fuel Burners", Steel Times, 150-158, (July, 1966).
(61) Grobel, E. A., and Maselle, A. J., "Blowing with Oxygen-Natural Gas
for Decarburization and Fume Suppression", AIME Electric Furnace
Conference, .27, 67-70 (1969).
(62) Discussion to: Hinds, G. W., and Hodge, G. W., "Use of Oxygen-Fuel
Gas Burners for Scrap Meltdown in Electric Furnace", AIME Electric
Furnace Proceedings, JJ, 300 (1959).
-------
86
(63) "New Scrap Press Reduces Electric Furnace Operating Cost", Iron and
Steel Engineer, 43 (8), 146, (August, 1966).
(64) Herrick, R. A., and Benedict, L. G., "A Microscopic Gassification of
Settled Particulates Found in the Vicinity of a Coke-Making Operation",
Journal of the Air Pollution Control Association, .19 (5), 325-328
(May, 1969).
(65) Telephone Conversation with Mr. Robert Bee, Mitre Corporation in
reference to EPA Contract Number 68-02-0290.
(66) Ferrari, R., "Experiences in Developing an Effective Pollution Control
Systems for a Submerged Arc Ferroalloy Furnace Operation", Journal
of Metals, 20, 95-104, (April, 1968).
(67) Scott, J. W., "Design of a 35,000 KW High Carbon Ferrochrome Furnace
Equipped with an Electrostatic Precipitator", AIME Electric Furnace
Proceedings, .29, 80-82, (1971).
(68) Young, J. H., and Singer, D. H., "Manufacture of Low Carbon Ferro-
chromium at the Steubenville Plant, Foote Mineral Company", AIME
Electric Furnace Proceedings, .29, 83-87 (1971).
(69) "Ferroalloy Fume Collection From Submerged Arc Electric Furnace
Operation", Industrial Heating, 38. (1), 86 (January, 1971).
(70) Fegan, G. J., "Cleaning Ferroalloy Furnace Fume with High Energy
Scrubbers", AIME Electric Furnace Proceedings, ,30, 65-68 (1972).
(71) Meredith, W. R., "Operation of a Baghouse Collecting Silica Fume",
AIME Electric Furnace Proceedings, .30, 69-71 (1972).
(72) Sherman, P. R., and Springman, E. R., "Operating Problems with High
Energy Wet Scrubbers on Submerged Arc Furnaces", AIME Electric Furnace
Proceedings .30, 72-76 (1972).
(73) Rentz, 0., Siebert, G., and Stracke, R., "Reducing Fume Emissions by
Improving Furnace Operation, by Feed Pretreatment", AIME Electric
Furnace Proceedings, 30, 77-83 (1972).
(74) Lopuszynski, T. W., Trunzo, J. P., and Wllbern, W. L., "Design and
Operation of a 45 MW 50 Percent Ferrosilicon Furnace", AIME Electric
Furnace Proceedings, 30, 89-93 (1972).
(75) "Development Document for Proposed Effluent Limitations Guidelines
and New Source Performance Standards for the Smelting and Slag
Processing Segment of the Ferroalloy Industry", U. S. Environmental
Protection Agency, Contract No. EPA 440/1-73/008 (August, 1973).
(76) Jenkins, R. D., "Potential Utilization and Disposal of Particulate
Materials Captured From a Silicon Metal Furance", Preprint AIME
Electric Furnace Conference, 10 pp (December, 1973).
(77) Killin, A. M., "Progress Report Air Pollution Control Study of the
Ferroalloy Industry", Preprint1 AIME Electric Furnace Conference,
9 pp (December, 1973).
(78) McClure, R. N., "Disposal of Submerged Arc Furnace Fume", Preprint
AIME Electric Furnace Conference, 6 pp (December, 1973).
-------
87
(79) "Copper Smelting Today: The State of the Art", Chemical Engineering,
Engineering and Mining J., Special Section, pp. p-z, (March, 1973).
(80) Lane White, "The Newer Technology: Where it is Used and Why",
Chemical Engineering, E&MJ, Special Section pp AA-CC, (March, 1973).
(81) Price, F.- C.,"Copper Technology on the Move", Chemical Engineering,
E&MJ, Special Section, pp RR-DDD, (March, 1973).
(82) Shoemaker, R. S,,"Minerals Processing in 1973", Mining Congress J.,
pp 24-29, (February, 1974).
(83) Holderreed, F. L., "Copper Smelting", Mining Engineering, p 45,
(September, 1971).
(84) Themelis, N. J., McKerrow, G. C., Tarassoff, P., and Hallett, G. D.,
"The Noranda Process", J. of Metals, pp 25-32, (April, 1972).
(85) American Metal Market, p. 19, (March 26, 1974).
(86) "What's Happening in Copper Metallurgy", E&MJ, pp 75-79, (February,
1972).
(87) "Form Consortium to Exploit New QS Process", J. of Metals, p. 12,
(March, 1974).
(88) "Mitsubishi's Continuous Copper Smelting Process Goes on Stream",
E&MJ, pp 66-68, (August, 1972).
(89) Suzuki, T., and Nagano, T., "Development of New Continuous Copper
Smelting Process", Paper given at Tokyo Meeting of AIME, May 27,
1972.
(90) Worthington, R. B., "Autogenous Smelting of Copper Sulfide Concen-
trate", U. S. Bureau of Mines, Report of Investigation 7705 (1973).
(91) Quarm, T. A. A., "Copper Smelting with a Cyclone Furnace", E&MJ,
pp 80-82, (October, 1969).
(92) Hunter, William D., and Michener, A. W., "New Elemental Sulfur
Recovery System Establishes Ability to Handle Roaster Gases",
E&MJ, pp 117-120 (June, 1973).
(93) George, D. R., Crocker, L., and Rosenbaum, J. B., "The Recovery of
Elemental Sulfur from Base Metal Smelters", Mining Engineering,
PP 75-77 (January, 1970).
(94) "ASARCO Building $16 Million Liquid S02 Plant", E&MJ, p 26,
(September, 1972).
(95) Bartlett, R. W., and Haung, H. H., "The Lime-Concentrate-Pellet
Roast Process for Treating Copper Sulfide Concentrates", J. of
Metals, pp 28-34, (December, 1973).
(96) Arbiter, N., "Anaconda's Ammonia Leach Process", Paper presented
at the Dallas Meeting of AIME, February, 1974.
-------
88
(97) Krusi, P. R., Allen, E. S., and Lake, J. L., "Cymet Process--
Hydrometallurgical Conversion of Base Metal Sulfides to Pure
Metals", CIM Transactions, Volume 76, pp 93-99 (1973).
(98) Kruesi, P. R., Allen, E. S., and Lake, J. L., "Inventor,
Developers Explain How New Cymet Process Works", Pay Dirt, pp 4-10,
(October 23, 1972).
(99) "Duval Claims Development of 'Polution Free1 Hydrometallurgical
Copper Refining Process", E&MJ, p 171, (September, 1970).
(100) Malouf, E. E., "Current Copper Leaching Practices", Mining
Engineering, pp 58-60 (August, 1972).
(101) Beal, John, "Copper in the U.S.--A Position Survey", Mining Engineering,
pp 35-47, (April, 1973).
(102) Iversen, R, E.,"Air Pollution in the Aluminum Industry, Journal of
Metals, V. 25, No. 1, pp 19-23 (January, 1973).
(103) Gerard, Gary, Editor; "Extractive Metallurgy of Aluminum, Volume 2,
Aluminum", Interscience Publishers, New York, p 572, (1963).
(104) Anon., "New Smelting Process, Alcoa Prepares the Site for a Pilot
Plant", American Metal Market, p. 18, (September 18, 1973).
(105) Piccolo, L., Chirga, M., and Calcagno, B., "Gas-solid Reactor:
Effects of Chemical and Fluodynamic Parameters on Alumina Chlorination
Process", Chimie ET Industrie-Genie Chimique, Vol. 101, No. 19,
pp 2485-2489, (November, 1971).
(106) Bureau of Mines, U. S. Department of Interior, "Aluminum from
Domestic Sources, a Miniplant to Evaluate Alumina Recovery Processes",
pp. 85.
(107) Hardesty, D. R., and Weinberg, F. J., "Electrical Control of Particulate
Pollutants from Flames" 14th International Symposium on Combustion,
Combustion Institute, p 907, (1973)
-------
APPENDIX A
-------
APPENDIX A
CONTROL OF FINE PARTICULATE EMISSIONS IN CONVENTIONAL
COPPER SMELTING PRACTICE
INTRODUCTION
U.S. copper ore reserves are predominately sulfides rather
than oxides with a low overall copper content (1% or less). Concentra-
tion and separation from other mineral values and gangue by grinding and
flotation yields concentrates containing 15 to 35 percent copper. Con-
centrates with a high sulfur content, those that are relatively high in
iron sulfide, and those containing certain volatile impurities such as
arsenic, antimony, and selenium require a preliminary roasting step.
Where roasters can be by-passed, air drying, kiln drying, or drying in
a roaster may be practiced. In other respects, smelting practice at the
15 U.S. primary copper smelters (Table A-l) until very recently, has
followed the universal technology of matte smelting, converting, and
refining shown in Figure A-l. By the way of introducing new modifications
in pyrometallurgical practice and new process schemes which by-pass the
smelting step entirely, we shall describe briefly what has been, up to
the present, conventional smelting practice.
Conventional U.S. Primary Copper-Smelting Practice
The three principal processing steps in conventional U.S.
smelting practice are roasting, matte-smelting, and converting. We
shall describe each in turn.
Roasting
As explained above, the need for roasting depends on the sulfur
content of the copper concentrates. The conventional oxidizing roast
oxidizes some of the sulfur content of the concentrate and converts part
of the iron sulfide to iron and sulfur oxides.
-------
A-2
TABLE A-l. ANNUAL COPPER PRODUCTION OF SMELTERS AND REFINERIES
IN THE U.S.
State and Location
Arizona
Hayden
Hayden
Miami
San Manuel
Morenci
AJo
Douglas
Maryland
Baltimore
Baltimore
Michigan
White Pine
Montana
Anaconda
Great Falls
Nevada
McClll
New Jersey
Perth Amboy
Perth Amboy
New Mexico
Hurley
New York
Maapsth(NYC)
Tennessee
Copperhill
Texas
El Paso
Utah
Carfield
Washington
Tacoma
Company
Asarco
Kennecott
Ins piration
Magma
Fhelps Dodge
Phelpi Dodge
Phelps Dodge
--
White Pine
Anaconda
Kennecott
__
Kennecott
«
Cities Service
Asarco
Kennecott
Asarco
Smetteri
1972 Annual
103 kkg
163
58
107
166
165
SO
115
-
65
183
39
— —
83
..
19
91
236
91
Refineries
Production
1C3 Short
Tont>
ISo(c)
6
118
183
ISZ(e)
55
127
--
72
202
43
—
92
«
2,«>
100;
260
100
Company
Inspiration
Magma
Kennecott
Asarco
White Pine(0
Anaconda
-.
Asarco
Anaconda
Kennecottlfl
Pbelps Dodge
..
1 Phelps Dodge
I Phelps Dodgelfl
Kennecott
Asarco
Annual
103 kkg
66
52
161
131
63
159
„
US
103
•0
65
„
311
23
«*
109
Production
10* Short
Tons(b)
TS
ST
ITS
144
TO
ITS
••
160
111
M
72
«•
420<*>
25(a>
260
120
(a) Data are (or 1971 from Yearbook of the American Bureau of Metal Statistics published in 1972.
(b) Data supplied by company.
(c) Capacity of production.
(d) Reference (IV.I).
(e) Reported as fire refined production.
(f) Fire refinery.
-------
A-3
Copper bearing material
(Blended concentrate* + direct •melting ore, etc. )
Flux
(Sand, gravel, low grade siliceous ore)
equal*
Furaace charge
_ I
Multiple hearth roaster.
(partial roast)
Fuel
It air'
Gaa
(3 to 5% SO2)
-i «- i • I Flue i Cold slaa
If tC*lc""« Tdust n matte Gas
T • 1 rift
Reverberate ry
furnace
Air-
. | Cottrell
I precipltator.
(1%SO2) J — — — — -.- ;
I .J Wa.te heat 1
~H boil*"
W Slag to dump
(37% SiO2 0. 5% Cu) r —-—•*•"
Matte J Cottrell J_
(30% copper) j_^precipitator. t
^^-X
Cold
iwoia .
'lV« V
Flux
Pierce-Smith
converter
Converter .lag
26% Si02
2%Cu
Blister
copper
J "i-__*J "H
c
::i:-;
Scrubber j
~T.:^
——-.
Hot I
* cottrelU I
""1
Acid
. Plant
i
Fire refining
furnace
H2SO4 I Wet
I cottrell
Anode, to electrolytic refinery
(gold, silver, selenium, tellurium recovery)
FIGURE A-1. GENERALIZED DIAGRAM OF CONVENTIONAL SMELTING FLOWSHEET
-------
A-4
In current conventional practice roasting may be carried out
either in multiple-hearth or fluid-bed roasters.
The multiple-hearth roaster is a cylindrical, brick-lined
vessel divided from top to bottom by horizontal brick hearths. Each
hearth has either one or several drop holes located alternately on the
inner and outer peripheries of successive hearths. A central, rotating,
brick-lined steel column extends vertically through the center of the
roaster. On each hearth, arms equipped with rabbles are fixed to the
rotating central shaft. Feed is dropped onto the top drying hearth near
the central shaft and rabbled to the outside of the hearth, where it
falls through the drop holes to the hearth below. The rabbles on this
hearth push the feed toward the central column, the feed drops to the
next hearth, and so forth until feed exists at the bottom hearth.
Air introduced at the bottom of the roaster passes up through
the heated chambers, and the oxygen in the air stream reacts with the
iron and sulfur in the feed to liberate heat, which sustains the roaster's
hearth temperature. The gases leaving the top of the roaster contain from
2 percent to 6 percent SO --rather low for sulfuric acid manufacture—and
carry away about 6 percent of the roaster calcine product. The coarser
calcine product rabbled from the bottom hearth is comprised of roasted
copper concentrates and flux and totals 94 percent of total calcine
produced. Sulfur elimination in the roaster is governed by regulation of
air flows and charge retention time on the hearths.
The fluid-bed roasting process is characterized by a gas-solid
reaction in a dense suspension of solids maintained in a turbulent mass
by the upward flow of gases that affect the reaction. The roaster is
essentially a cylindrical refractory-lined steel shell used to contain
the suspended solids.
Air is forced into the roaster through tuyeres in a refractory-
lined steel distribution plate that is placed at the bottom of the shell.
The two best-known types of fluid-bed roasters, the Lurgi and the Dorr-
Oliver, are characterized by different tuyere design.
-------
A-5
The use of fluid-bed roasting installations is not yet as
common in copper smelting as it is in zinc, nickel, iron calcine, and
sulfuric acid production units. This is partly because the copper
industry is reluctant to wipe out the large investment already made
in existing multiple-hearth roasters and partly because of difficulties
experienced in fluid-bed operation, shut downs caused by sintering of the
roaster bed, and heavy sulfate carryover to the electrostatic precipi-
tators.
Another possible drawback is the excess calcine carryover of
80 percent in outlet gases, as compared with 6 percent for multiple-
hearth roasters. A more elaborate dust-handling system is therefore
required with fluid-bed units.
There are also some advantageous features to fluid-bed roasting.
There are no moving parts in the combustion chamber, and maintenance is
simplified accordingly. The vigorous gas currents existing in the fluid
bed maintain very uniform bed temperature and composition. And the
roasting action is so vigorous that little excess air is required,
permitting S02 contents of 12 percent to 14 percent to be obtained.
Matte Smelting in Conventional
Reverberating Furnaces
Matte smelting is done in a reverberatory furnace (Figure A-2).
These are large shallow-hearthed structures of up to 40-meters (130 feet)
long, 12-meters (38 feet) wide, with the capability of treating up to
1450-metric tons (1600-short tons) of charge per day. The objective of
this treatment is to collect virtually all of the copper in a molten
copper-iron-sulfide "matte" layer which is tapped from the furnace for
subsequent treatment in converters.
The reverberatory furnace has remained the workhorse of the
copper smelting industry for 80 to 90 years and is only now being
challenged by newer furnaces that emit a more concentrated exhaust gas.
In conventional operation, calcine from the roaster hoppers,
flues, and electrostatic precipitators is gravity charged into hoppers
staggered along each side of the furnace and located above drop holes
in the roof (Figure A-2). Heat is supplied by coal or gas burners
located at one end of the furnace with about (4,000,000 Btu) required per
-------
i" " .i-• »»«•"••» -. Fettling drag "• Fettling pipet .'™t,':~ •,'
•: '..'I" --'V-' * conv-yor v -•-X-':..^,. ; . , , lrf~*:^^....:'>^
\-**Hfe • ——^y***
Fuel.
Converter.
Jl89-
Air and
oxygen
'Nrj^^SS ^-.c, N-—-rr?^
/ jT*f^\ Slag''
„ . ii^v.-.'^i , . v ••;-*"•??" ' ^ ••••-•,
BumenX ^Matte
::.J5^i^«;^iaft^^»:rv ^•-i:f^?. ^- v.1. ;«_,> " ' :f^Slii8^::''vl•'•
Off^u
' $jji
-«'•".->•.
.:'.-W
Matte
,*l8fl
P^t-
' -•'•
FIGURE A-2. CUTAWAY VIEW SHOWING KEY FEATURES OF A CONVENTIONAL
REVERBERATORY FURNACE.
-------
A-7
ton of charge smelted when the charge is hot roasted material. Thermal
efficiency of the reverberatory furnace is low; however, furnace
installations are usually equipped with waste-heat boilers, which
recover much of the heat of the combustion gases in the form of super-
heated steam.
Anywhere from 11 percent to 16 percent of the sulfur content
in a calcine charge, and up to 20 percent of the sulfur content in an
unroasted charge, is liberated during smelting in a reverberatory
furnace and is mixed with the products of combustion to give an outlet
gas usually containing between 0.5 percent and 1.0 percent SO . Gases
pass through waste-heat boilers and electrostatic precipitators and are
then vented to the atmosphere. Considerable experimental work is being
done using limestone scrubbers to remove small percentages of S0« from
stack gases, but the problems involved in treating upward of 3-million cubic
meters (100-million cubic feet) of gas per furnace per day are formidable.
Generally, slag is side-tapped near the exhaust end of the
furnace into locomotive-drawn slag pots of about 6-cubic meters (200-cubic
feet) capacity. Slag may be either granulated or dumped molten.
Matte is tapped into ladles, each containing about 16 metric
tons (18 short tons), which are drawn down the matte tunnel to the
converter aisle by winch and then charged by cranes into one of the
waiting Peirce-Smith Converters.
Converting;
Converting is a batch process and is done in large, horizontal
cylindrical Peirce-Smith side-blown converter furnaces up to 9 meters
•
(30 feet) in length and about 4 meters (13 feet) in diameter, with a
centrally located aperture for charging, unloading and,.exit of gases.
They are also fitted with numerous tuyeres along their length near the
bottom through which the air necessary for converting is blown. The
converters are mounted on trunnions so that they may be tilted. During
the converting operation, the furnace is run in an upright position with
the aperture directly beneath a hood through which the converting gases
are exhausted. The furnace is tilted for charging, discharging, and
inspection.
-------
A-8
In operation, the converter converts matte from the reverberatory
furnace to an impure form of copper called "blister copper" (so called
from the appearance of its cast surface which shows evidences of gas
expulsion).
It is a two-stage oxidation process involving high through-puts
of air, 850 cubic meters (30,000 cubic ft) per minute, through the molten
matte.
In the first stage, the iron sulfide component of the matte
is oxidized to sulfur dioxide and iron oxide, leaving nearly all of the
copper as molten copper sulfide or "white metal". Sulfur dioxide
leaves the converter as a gas and may be processed to sulfuric acid.
Fumes and dusts which may contain lead, bismuth, arsenic, etc.,
produced by converting are collected and sent to lead smelters. The
iron oxide formed during converting reacts with silica, added as a flux
to the converting operation, to form a molten iron silicate slag which is
removed and returned to the reverberatory furnace for reprocessing since
it contains significant concentrations of copper.
In the second stage, after removal of the slag, blowing is
continued until virtually all of the remaining sulfur is oxidized and
removed as sulfur dioxide. The converted copper is normally transferred
while still molten to the refining furnace. Blister copper at this stage
is still relatively impure; it contains varying amounts of heavy metals,
arsenic, some sulfur, and all of the precious metal constituents originally
present in the concentrates.
A feature of the Peirce-Smith converters is that they can be
modified into a kind of smelting furnace; oxygen is mixed with blowing
air to increase the heat available from exothermic converting reactions.
The chief disadvantage of the Peirce-Smith converter is the
relatively low concentration, 2 percent to 6 percent SO outlet gas
produced when excessive air is allowed to infiltrate into the off-take
hood over the converter mouth.
In conventional primary copper smelter practice, converters
produce more than half of the SO. emissions, the converter can produce
-------
A-9
S02 gas emissions containing in excess of 6 percent SCK. Gas concentra-
tions are not steady, since the conversion of matte to blister copper
is a batch process. However, with several converters, these irregular
streams can be averaged out for sulfuric acid manufacture by regulating
blowing schedules, with good results. Most of the sulfuric acid from
smelters is made from converter gas.
Current Emission Control Practice in Conventional
U.S. Primary Copper Roasting, Matte-Smelting and Converting Operations
The type of emission control devices used in U.S.
primary copper smelter operations, with their control efficiencies, are
shown in Table A-2.
Seven of the 15 U.S. smelters have roasters; all are equipped
with emission control devices. Smelters equipped with fluid-bed roasters
are controlled with acid plants, those with multihearth roasters are
controlled with electrostatic precipitators (ESP's). Control efficiencies
of the acid plants were 99.5 percent, the ESP's handling emissions from
the multihearth roasters had efficiencies of 90-90.5 percent. Particulate
emissions from the roaster are in the form of fine dust and fume, in
many cases, with a particle size of less than 1 micron. Roaster particu-
late emissions will contain, in addition to 10 percent copper and a
large amount of silica and alumina from the gangue, quantities of the
more volatile concentrate constituents such as zinc, cadmium, antimony,
and arsenic in about the same amounts as that in the concentrate.
Most of the reverberatory furnaces are equipped with either
ESP's or cyclones to clean the reverberatory smelter gases. As of 1972,
four were not equipped with emission control devices of any kind. Particle
sizes of the dust and fume emissions from the reverberatory furnace varied
from less than 0.1 micron to greater than 5 microns. "In an ESP controlled
reverberatory, 70 percent of the stack emissions were less than 0,5 micron,
with 34 percent between 0.1 and 0.3 micron. Compositions of typical
emissions from reverberatory stacks are given in Table A-3.
All except three of the converters listed in Table A-2 have
some type of particulate emission control, these may be ESP, acid plant,
scrubber or multicyclone, alone or in combination. Raw emission factors
-------
A-10
TABLE A-2 CONTROL PRACTICE IN COPPER SMELTING (1972)
State
Nevada
JJew Mexico
(Utah
Arizona
Size of
Feed, T/da
170
750
(19% Cu)
2200
1270
(capacity)
Operation
Copper
Produced,
T/da
-30
200-250
800
Operation
Reverberatory
smelter
Converter
Reverberatory
smelter
Converter
Reverberatory
smelter
Converter
Roaster
Efficiency
Control Device 7.
Balloon flues-long
brick flue-multi-
clone-ESP*-tall
stack " 70-85
Cyclone —
ESP-stack 95
Multicyclone-stack 85
Balloon flues-ESP-'
stack*** 45
ESP-wet scrubber-
ESP-drying tower
acid plant —
Cyclones-gas cooler-
Arizona
Arizona
550
Arizona
(a)
2000
(capacity)
Reverberatory
smelter
Converter
300 Reverberatory
Converter
Roasting
(F-B, D-0)
Reverberatory
furnace
Converter
Refining
450 Roasting (Mh)
Reverberatory
furnace
Converter
ESP-acid plant
Gas cooling-balloon
flue-ESP-stack
Balloon flue-ESP-
acid plant —
None 0
None 0
Brink's mist eliminator
Acid plant
ESP
ESP
None
ESP-stack • 98
ESP-stack 98
Cyclone-ESP-atmos-
phere or acid plant
-------
A-11
TABLE A-2. CONTROL PRACTICES IN COPPER SMELTING
(Continued )
Size of Operation
; Copper
Concentrate Produced,
,state Feed, T/da T/da
^ington(h) 300
*
Operation
Roasting
Reverberatory
Converter
Control Device
ESP
ESP
Cyclone-water spray-
ESP-scrubber-ESP-
acid plant
Efficiency
•,,_
—
-_
Refining (electro-
ill«essee(b) 300 -57
'chigan & 220 (c)
[
2°na 60,000 T
(ore) -450
*°na 350
!**s(e) 350
(T Cu)
v_
lytic)
Roasters
(fluid bed)
Reverberatory
smelting
Converter
Reverberatory
furnace
Converter
Reverberatory
_ id )
furnace ...
(d }
Converter
Fire-refining
Electrolytic
refining
Roasters
Reverberatory
furnace
Converter
Roasters
Reverberatory
furnace
Converter
Refining
Scrubber-acid plant
Settling chamber
Scrubber-acid plant
ESP or cyclone
None
ESP
ESP
7
?
ESP
None
ESP
ESP
ESP
ESP
None
-10
—
-95
0
—
—
—
— —
-90
0
-95
—
-98
—
0
-------
A-12
TABLE A-2 CONTROL PRACTICES IN COPPER SMELTING
(Continued)
State
Arizona
Montana
Size of Operation
Copper
Concentrate Produced,
Feed, T/da T/da Operation
670 -200 Reverberatory
furnace
Converter
Oxidizing
furnace
Refining and
anode furnace
2800 Roasters
Reverberatory
furnace
Converter
Control Device
ESP
ESP
ESP
ESP
None _
None
None
Efficiency
j ,
—
—
—
— —
-.0
0
0
(a) Estimated (EPA) particulate emissions/day of 1.2 T/da; 2000 T (concentrate) capacity
(29-31% Cu).
(b) New copper system planned.
(c) Stack losses "1.5 T/day of particulate dust; single stack for plant.
(d) Particulate emissions 14.8 T/day at 35 percent Cu.
(e) Plan to build acid plant to treat converter gases.
(f) 0.7-T dust emitted/day; additional emission controls being constructed.
(g) Ore mined at a rate of 230,000 T/day.
(h) Total stack emissions—253 Ibs/hr.
* ESP - Electrostatic precipitator.
*** New system planned to increase efficiency to 95+%. ESP-wet scrubber-packed tower
system has given 99.5% efficiency.
-------
A-13
TABLE A-3. COMPOSITION OF TYPICAL EMISSIONS FROM
REVERBERATORY STACKS
Element, %
Arsenic
Cadmium
Copper
Selenium
Zinc
Chromium
Manganese
Nickel
Vanadium
Boron
Barium
Reverberatorv Stack
(1)
._ 0.0102
<0.0001
16.25
0.0006
0.035
0.0049
0.0085
0.0185
0.0101
nil
nil
(2)
N.D.
0.13
11.0
0.0009
2.9
0.045
0.063
0.064
trace
trace
trace
(3)
0.21
0.03
6.3
0.19
0.89
0.009
0.023
0.012
0.01
0.14
0.065
(1) New Mexico.
(2) Arizona.
(3) Nevada.
-------
A-14
for the uncontrolled converters have been estimated to be between 22-24 kg
per metric ton (45-90 pounds/short ton). Typical reported compositions
of emissions from converter operations are given in Table A-4. Character-
istics of the emissions from the converter stack of the second smelter in
the listing in Table A-2 are given in Table A-5. This converter was
controlled with a multicyclone.
In addition to the emissions noted above, there is an additional
problem involved in transfer operations in the batch operations of a
conventional smelter. The new continuous process described in the
accompanying report would eliminate this.
-------
A-15
TABLE A-4. COMPOSITIONS OF ATMOSPHERIC EMISSIONS
FROM CONVERTER OPERATIONS IN PRIMARY
COPPER INDUSTRY
Composition of Converter Emissions,
Percent, Metal Basis
Element
Arsenic
Cadmium
Copper
Selenium
Zinc
Chromium
Manganese
Nickel
Vanadium
0.
0.
1.
-------
A-16
TABLE A-5. ANALYSES OF CONVERTER STACK EMISSIONS
FROM A WESTERN U.S. SMELTER CONTROLLED
WITH A MULTICYCLONE COLLECTOR IN THE
FLUE AHEAD OF THE STACK
Gas stream characteristics
--volume flow rate: 160, 000 SCFM @60°F and 750 mm Hg
(upper limit) and 120, 000 SCFM @60°F and 750 mm Hg
(lower limit)
Condition assumes three converters in operation, two in
the stack and one out.
--temperature: 550°F
--pressure drop in system: 1. 75" w. c.
--gas color: pale white--transparent
--odor: pungent, irritating
--composition volumetric:
S02 : 3. 0%
H20 : 3. 0%
CO2 : 0. 5%
02 : 19,4%
N2 : 74. 1%
Particle characteristics
--grain loading: 0. 7 grains/ft
--size analysis of dust:
Size microns^ Percent
0-5 7
5-10 5
10-15 3
15-20 2
20-25 1
25-30 1
30-- 81
-------
TECHNICAL REPORT DATA
(Please read JnXructions on the reverse before completing)
1. REPORT NO,
EPA-650/2-74-100
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE AND SUBTITLE
Process Modifications for Control of Particulate
Emissions from Stationary Combustion,
Incineration. and Metals
5. REPORT DATE
October 1974
6. PERFORMING ORGANIZATION CODE
rfril *-* -*~J"rf A MiM AWAJI • MiiiM JHW ilfciilfcrf ^^^_^^_
7.AUTHOR(S)R> Nekervis, J. Pilcher, J. Varga Jr. ,
B. Gonser, and J. Hallowell
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Battelle, Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
10. PROGRAM ELEMENT NO.
1ABQ12; ROAP 21ADK-017
11. CONTRACT/GRANT NO.
68-02-1323 (Task 9)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 3/74-7/74
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
rep0rf summarizes the state of process modifications relative to the
control of fine particulate emissions from stationary combustion sources (electric
utilities and industrial processes); municipal incinerators; iron and steel plants;
ferro-alloy plants; and non-ferrous metal smelters (zinc plants, copper smelters,
and aluminum reduction cells). This study is to uncover modifications to conventional
or new unconventional practices which appear to improve the control of fine particu-
late emissions in these five areas. Modifications to conventional stationary combus-
tion sources considered include ash-fluxing, SOS addition to flue gas, staged combus-
tion, use of fuel additives, fly ash agglomeration, solvent refining, and flue gas
recirculation. Unconventional systems studied include fluid-bed, coal gasification,
and submerged combustion. For incinerators, combined fuel/refuse firing, gas cool-
ing, and pyrolysis methods are considered. For iron and steel plants, emphasis is
given to the bottom-blowing oxygen process (Q-BOP). Modification of the conventional
reverberatory smelting procedure and the introduction of hydrometallurgical methods
are discussed for copper; the AI-CI electrolytic (ASP) process is considered for alu-
minum. The stage of development, availability or acceptability by industry, emis-
sion reduction efficiency, and environmental impact of each process is considered.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Air Pollution
Combustion
Electric Utilities
Industrial Processes
Incinerators
Iron and Steel Industry
Ferroalloys
Smelting
Zinc Industry
Copper Converters
Aluminum Industry
Air Pollution Control
Stationary Sources
Fine Particulate
Process Modifications
13B
21B
13H
11F
8. DISTRIBUTION STATEMENT
19. SECURITY CLAS
Unclassified
CLASS (This Report)
21. NO. OF PAGES
Unlimited.
116
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73).
------- |