U.S. Environmental Protection Agency Industrial Environmental Research CDA f\C\C\ f~7 7A
Office of Research and Development Laboratory
Cincinnati. Ohio 45268 December 1976
ENVIRONMENTAL
CONSIDERATIONS OF
SELECTED ENERGY
CONSERVING MANUFACTURING
PROCESS OPTIONS:
Vol. XIV. Primary Copper
Industry Report
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentallycompatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-76-034n
December 1976
ENVIRONMENTAL CONSIDERATIONS OF SELECTED
ENERGY CONSERVING MANUFACTURING PROCESS OPTIONS
Volume XIV
PRIMARY COPPER INDUSTRY
EPA Contract No. 68-03-2198
Project Officer
Herbert S. Skovronek
Industrial Pollution Control Division
Industrial Environmental Research Laboratory - Cincinnati
Edison, New Jersey 08817
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45628
For sale by the Superintendent of Documents, U.S. Government Printing Offico. Washington, O.C. 2M02
-------
DISCLAIMER
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Approval does not signify that the contents necessarily reflect the
views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
ii
-------
FOREWORD
When energy and material resources are extracted, processed, converted,
and used, the related pollutional impacts on our environment and even on our
health often require that new and increasingly more efficient pollution con-
trol methods be used. The Industrial Environmental Research Laboratory -
Cincinnati (lERL-Ci) assists in developing and demonstrating new and im-
proved methodologies that will meet these needs both efficiently and
economically.
This study, consisting of 15 reports, identifies promising industrial
processes and practices in 13 energy-intensive industries which, if imple-
mented over the coming 10 to 15 years, could result in more effective uti-
lization of energy resources. The study was carried out to assess the po-
tential environmental/energy impacts of such changes and the adequacy of
existing control technology in order to identify potential conflicts with
environmental regulations and to alert the Agency to areas where its activi-
ties and policies could influence the future choice of alternatives. The
results will be used by the EPA's Office of Research and Development to de-
fine those areas where existing pollution control technology suffices, where
current and anticipated programs adequately address the areas identified by
the contractor, and where selected program reorientation seems necessary.
Specific data will also be of considerable value to individual researchers
as industry background and in decision-making concerning project selection
and direction. The Power Technology and Conservation Branch of the Energy
Systems-Environmental Control Division should be 'contacted for additional
information on the program.
David G. Stephan
Director
Industrial Environmental Research Laboratory
t Cincinnati
iii
-------
EXECUTIVE SUMMARY
The primary copper industry is one of the most energy intensive industries
in the United States, ranking twelfth in terms of energy purchased in the
industrial sector. The industry is divided into four segments: mining, bene-
ficiation, smelting, and refining. The U. S. primary refined copper production
of over 2 x 10" ton/yr is derived from about 25 major mines, 18 smelters, and
16 refineries. About 20% of the refined copper is derived from copper scrap.
All energy forms are used in the copper industry. Melting of the charge
materials is accomplished with fossil fuel (natural gas, oil, or pulverized
coal firing) or with electric-resistance heating. Electricity is used for
electrolysis for recovering and purifying copper by plating it from a solution.
Electricity is used for gas, liquid and solid materials handling. Air pollu-
tion control has increased gas handling requirements and electricity consump-
tion. The industry traditionally has practiced waste heat recovery from the
hot exit gases in order to reduce overall energy requirements.
This report assesses the pollutional consequences of selected new copper
process options which may be implemented in the smelting and refining segments
of the industry. We have examined six process options in depth. These are
Outokumpu flash smelting, the Noranda process, the Mitsubishi process, oxygen
use in smelting, metal recovery from slags (flotation or electric furnace),
and the Arbiter process. Tables ES-1 and ES-2 summarize our findings with
regard to the economic, environmental and energy implications of these six
options.
The first three processes (Outokumpu, Noranda, and Mitsubishi) can be
considered together. All three are pyrometallurgical processes which use
exothermic reactions occurring during smelting to reduce net energy require-
ments. Use of any one of these new processes could reduce energy consumption
by 30-50% (the latter when using oxygen enrichment). The processes are flex-
ible in that they can be fueled by gas, oil, or coal. Compared to the conven-
tional process (which produces a large, dilute, S02~containing stream) the new
processes produce only concentrated SC^ streams which can be treated effectively
and economically for SC>2 recovery by conversion to sulfuric acid. Therefore,
S02 emissions are also significantly lower for these processes and a sulfur
capture of over 90% is achievable, compared to 50-70% for conventional smelting.
At present, it appears that the sulfuric acid can generally be sold only at a
price below the full production cost. It is used for leaching, neutralization,
or phosphoric acid manufacture. The major hindrance to acceptance of these
processes for new installations today is that their applicability to impure
concentrates is unproven. Also, because the growth in copper demand is low,
few new smelters will be built in the near future. On the other hand, because
of EPA regulations on modification of existing smelters, much of the future
smelting capacity will result from the construction of new smelters.
iv
-------
TABLE ES-1
QUALITATIVE SUMMARY OF COSTS/ENERGY/ENVIRONMENTAL CONSEQUENCES OF PROCESS OPTIONS IN THE COPPER INDUSTRY
Base Line Process: Conventional Smelting and Refining of Concentrates
PROCESS OPTIONS
COSTS
FMPHfV
CHEiKLrl
ENVIRONMENTAL
Outokumpu
Smelting
Slightly higher
capital cost.
Lower operating
cost because of
energy savings.
Slightly lower
air pollution
control cost be-
cause more con-
centrated so2
streams are
handled ,
Lower energy
cost use.
Overall saving
of 30-50%.
Higher levels of
sulfur recovery
are possible
(over 90%).
Minimal or no
water pollution.
Solid waste prob-
lem unchanged.
No change in feed-
stock but possible
changes in impurity
distributions which
might require
different product
treatment.
Noranda
Smelting
Same as
Outokumpu
Same as
Outokumpu
Same as
Outokumpu
Mitsubishi
Smelting
Same as
Outokumpu
Same as
Outokumpu
Same as
Outokumpu
Oxygen Use
in Smelting
Slightly lower
capital cost.
Slightly higher
maintenance re-
quirements .
Lower operating
costs.
Lower energy
use even after
energy used in
oxygen manu-
facture is taken
into account.
No major change.
Possible change
in impurity dis-
tributions which
might require
different product
treatment.
Metal Recovery from Slags
Electric Furnace
Lower capital
cost. Lower
operating cost,
but lower copper
recovery.
About the same
energy require-
ments .
Solid waste
problem un-
changed . No
water pollution.
Flotation
Higher capital
cost. Higher
operating cost,
but higher copper
recovery .
About the same
energy require-
ments.
Solid waste of
finer particle
size. Potential
water pollution
problem.
Arbiter Process
Lower capital coet
at smaller plant
size. Higher op-
erating cost.
Higher energy re-
quirements. Most
of the energy re-
quired is in the
form of electricity.
Minimal air pollution.
Change In type of
solid wastes generated.
Some aqueous effluents.
-------
TABLE ES-2
QUANTITATIVE COMPARISON OF BASE LINE AND ALTERNATIVE PROCESSES IN THE COPPER INDUSTRY
Environmental
Pollution Control
Costs ($/ton of
Copper)
Energy Consumption
106 Btu/ton
of Copper3
Process Economics
Investment ($/
annual ton)
Outokumpu
Reverb Electro- Smelting Slag Cleaning
Smelting Refining No Noranda 'Mitsubishi Electric
(Base Case) (Base Case) Oxygen Oxygen Smelting Smelting Furnace Flotation Arbiter Process
54-64
23-26
650-750
Pollution Control
and Operating Cost
($/ton)b 340-370
4.6
450
59 59
15 13.2 12.5
750 500
750
230 336 259 336
59 15 1.90
12 1.2V 0.8
750
333 10.47c'd 12.40c'd
61
1000
636
alncludes electricity at 10,500 Btu/kWh; oxygen at 360fcWh/ton.
Includes pre-tax return on investment - excludes concentrate cost.
c$/ton of slag.
Equivalent to about $200-360/ton of recovered copper.
-------
The use of oxygen in smelting decreases fuel requirements and is energy
efficient overall since the decrease in fuel requirements is usually larger
than the energy required for oxygen separation. Other than higher operating
temperatures and the resulting increased maintenance, we can see no real
disadvantages to this option.
The two processes for recovering copper from slag remove a major con-
straint in conventional smelting: the need to maintain a low matte grade in
the primary smelting unit to minimize copper losses in the slag. These
processes are, therefore, important adjuncts to the new smelting processes
which operate at high matte grades as well as being potentially usable in
existing smelters. Both flotation and electric furnace can decrease flux
requirements in smelting, increase the overall recovery of copper in some
cases, and improve smelting unit operations. Operating costs for electric
furnace cleaning are lower, but the higher costs of slag flotation are com-
pensated for by higher copper recovery. Energy requirements for both processes
are about equal. Both processes produce solid waste. However, that from
flotation has a much smaller particle size and requires different disposal/
storage techniques.
The Arbiter process (a hydrometallurgical process) is significantly more
expensive for treating chalcopyrite concentrates than conventional smelting
and refining. The process is less polluting but uses more energy than the
conventional technology. The major advantage of the process is its ability
to operate on a small scale (30,000-40,000 ton/yr copper) compared to smelting
which requires a minimum plant size of over 100,000 ton/yr of copper.
This report was submitted in partial fulfillment of contract 68-03-2198
by Arthur D. Little, Inc. under sponsorship of the U.S. Environmental Protec-
tion Agency. This report covers a period from June 9, 1975 to December 1, 1975.
vii
-------
TABLE OF CONTENTS
FOREWORD 111
EXECUTIVE SUMMARY iv
List of Figures xii
List of Tables xiv
Acknowledgments xvii
Conversion Table six
I. INTRODUCTION 1
A. BACKGROUND ' 1
B. CRITERIA FOR INDUSTRY SELECTION 1
C. CRITERIA FOR PROCESS SELECTION 3
D. SELECTION OF PRIMARY COPPER INDUSTRY PROCESS OPTIONS 3
II. FINDINGS, CONCLUSIONS, RECOMMENDATIONS 8
A. PYROMETALLURGICAL PROCESS 8
B. OXYGEN USE IN SMELTING 11
C. RECOVERY OF METALS FROM SLAG 11
D. THE ARBITER PROCESS 12
E. RESEARCH AREAS 12
1. Impurity Distributions 12
2. Impurity Removal 12
3. Feedstocks 13
4. Metal Recovery/Separation 13
5. Miscellaneous 13
III. INDUSTRY OVERVIEW 14
IV. COMPARISON OF CURRENT AND!ALTERNATIVE PROCESSES 21
A. REASONS FOR CHOOSING OPTIONS TO BE ANALYZED IN DEPTH 21
1. Problems of Current Technology 21
2. Reasons for Choosing Options for Detailed Analysis 22
ix
-------
TABLE OF CONTENTS (Cont.)
B. BASE LINE: CONVENTIONAL SMELTING AND REFINING
1. Introduction
2. Effluents
3. Economic Factors
C. OUTOKUMPU FLASH SMELTING 41
1. Flash Smelting - Concept and Operations 41
2. Current Status of Flash Smelting 43
3. Effluents from Flash Smelting 44
4. Technical Considerations 49
5. Economic Factors 51
D. THE NORANDA PROCESS 55
1. Concept and Operation 55
2. Current Status 58
3. Effluents 58
4. Technical Considerations 64
5. Economic Factors 66
E. THE MITSUBISHI PROCESS 71
1. Concept and Operations 71
2. Current Status 76
3. Effluent Control 76
4. Technical Considerations 79
5. Economic Factors 79
F. THE USE OF OXYGEN IN SMELTING 82
1. Concept and Operations 82
2. Current Status 83
3. Effluents 83
4. Technical Considerations 84
5. Economic Factors 85
G. METAL RECOVERY FROM SLAG 85
1. Background 85
2. Slag Cleaning via Flotation 89
3. Slag Cleaning in Electric Furnaces 93
4. Economic Factors 95^
x
-------
TABLE OF CONTENTS (Cont.)
H. THE ARBITER PROCESS 95
1. Concept and Operations 95
2. Current Status 98
3. Effluents 99
4. Technical Considerations 101
5. Economic Factors 102
V. IMPLICATIONS OF POTENTIAL PROCESS CHANGES 107
A. INTRODUCTION 107
B. PYROMETALLURGICAL PROCESSES 107
C. OXYGEN USE IN SMELTING 109
D. RECOVERY OF METALS FROM SLAG 109
E. THE ARBITER PROCESS 110
F. SUMMARY 110
REFERENCES 111
APPENDIX A - PRESENT COPPER TECHNOLOGY 114
APPENDIX B - GLOSSARY 123
xi
-------
LIST OF FIGURES
Number Page
III-l Generalized Flowsheet for Copper Extraction From
Sulfide Ores 16
III-2 Where the Smelters and Refineries are Located 17
III-3 Trends in Net U.S. Imports and Production of Refined Copper 20
IV-1 Sources of Emissions in Conventional Copper Smelting
and Refining 25
IV-2 Diagram of Impurity Flow Within a Copper Smelter 31
IV-3 Flowsheet of the Harjavalta Outokumpu Smelter 42
1V-4 Sources of Emissions in the Outokumpu Flash Smelting Process 45
IV-5 Diagram of Impurity Flow Within a Smelter Using Outokumpu
Flash Furnaces 48
IV-6 Schematic of -the Noranda Process Reactor 56
IV-7 Material Flowsheet (per hour) for Noranda Process Plant
(800 ton concentrate/day) 56
IV-8 Sources of Emissions in the Noranda Process 61
IV-9 Smelting Rate Versus Tonnage Oxygen Added 67
IV-10 Fuel Ratio Versus Tonnage Oxygen Added 68
IV-11 Schematic View of Mitsubishi's Semi-Commercial Plant 72
IV-12 Copper Losses in Smelting Furnace Slag 75
IV-13 Emission Sources in the Mitsubishi Continuous Copper
Smelting Process 77
IV-14 Diagram of Impurity Flow Within a Smelter Using the
Mitsubishi Process 80
xii
-------
LIST OF FIGURES (Cont.)
Number Page
IV-15 Slag Composition and Impurity Levels 80
IV-16 Off-Gas Volume and Energy Consumption at the Harjavalta
Capper Smelter 84
IV-17 Copper Content of Reverberatory Slags Plotted Against
Matte Grade 87
IV-18 Solubility of Copper in Silica-Saturated Slag as a Function
of Oxygen and Copper Content of Matte 88
IV-19 Effect of Oxygen Pressure on Cu 0 Content of Slag 88
IV-20 Typical Slag Milling Flowsheet 91
IV-21 Anaconda Arbiter Plant - Block Flow Diagram 97
IV-22 Sources of Emissions in the Arbiter Process 99
A-l Typical Flowsheet - Sulfide Copper Ore Flotation 118
xiii
-------
LIST OF TABLES
Number Page
1-1 Summary of 1971 Energy Purchased in Selected Industry
Sectors 2
1-2 Copper Extraction from Sulfides: Conventional Processing
and Process Options 5
II-l Qualitative Summary of Costs/Energy/Environmental Con-
sequences of Process Options in the Copper Industry 9
II-2 Quantitative Comparison of Base Line and Alternative
Processes in the Copper Industry 10
III-l Leading World Producers of Refined Copper 14
III-2 U.S. Mine Production of Copper 15
IV-1 Emissions from Conventional Smelting 26
IV-2 Costs of Operating an Acid Plant: Conventional Smelting 29
IV-3 Summary of the Weight Distribution of Minor Elements in
the Conventional Smelting Process 32
IV-4 Raw Waste Characteristics of Certain Streams in Copper
Smelting 35
IV-5 Estimated Water Pollution Costs - Copper Smelting 36
IV-6 Operating Costs: Conventional Smelting 37
IV-7 Cost of Pollution Controls in Conventional Smelting 39
IV-8 Operating Costs: Conventional Electrorefining 40
IV-9 Emissions from Outokumpu Smelting . 46
IV-10 Sulfur Distribution in Flash Furnace and Converter Gases 46
IV-11 Sulfur Loss in Various Processing Steps in Flash Smelting 46
IV-12 Operating Costs for an Acid Plant: Outokumpu Smelting 47'
IV-13 Estimated Distribution of Minor Elements among the Various
Streams Identified in Figure IV-5 49
xiv
-------
LIST OF TABLES (Cont.)
Number Page
IV-14 Operating Costs: Outokurapu Flash Smelting 52
IV-15 Energy Balance for Outokumpu Flash Smelting 54
IV-16 Cost of Pollution Control in the Outokumpu Flash
Smelting Process 55
IV-17 Comparison of the Noranda Process vs. Conventional
Reverberatory-Converter Works 59
IV-18 Typical Commercial Operation with Air 60
IV-19 Emissions from Noranda Smelting 61
IV-20 Noranda Process - Distribution of Minor Elements when
making 70% Grade Matte 63
IV-21 Recovery of Minor Elements in Slag Milling 65
IV-22 Observed Distribution of Minor Elements in Noranda
Process Making Copper 66
IV-23 Typical Commercial Operation with Oxygen 67
IV-24 Operating Costs: Noranda (Matte) Process 69
IV-25 Cost of Pollution Control in the Noranda Smelting Process 70
IV-26 Blowing Conditions 72
IV-27 Average Compositions and Throughput of Concentrates 73
IV-28 Smelting Rate and Fuel Consumption of Various Smelters 73
IV-29 Copper Losses in Slag 75
IV-30 Typical Analyses of Blister Copper Produced 76
I
IV-31 Emissions from the Mitsubishi Process 78
IV-32 Operating Costs: The Mitsubishi Process 81
IV-33 Cost of Pollution Control in the Mitsubishi Continuous
Copper Smelting Process 82
IV-34 Operating Costs: Outokumpu Flash Smelting using Air and
35% Oxygen 86
-------
LIST OF TABLES (Cont.)
Number Page
IV-35 Efficiency of Metals Recovery from Slags by Flotation 91
IV-36 Slag Flotation Operations 92
IV-37 Operating Costs: Slag Cleaning Process 96
IV-38 Emissions from the Arbiter Process 100
IV-39 Operating Costs: The Arbiter Process 104
IV-40 Arbiter Process - Solid Waste Disposal 106
A-l Identified and Hypothetical Copper Resources 116
xvi
-------
ACKNOWLEDGMENTS
This study could not have been accomplished without the support of a
great number of people in government agencies, industry, trade associations
and universities. Although it would be impossible to mention each individual
by name, we would like to take this opportunity to acknowledge the particular
support of a few such people.
Dr. Herbert S. Skovronek, Project Officer, was a valuable resource to us
throughout the study. He not only supplied us with information on work
presently being done in other branches of EPA and other government agencies,
but served as an indefatigable guide and critic as the study progressed. His
advisors within EPA, FEA, DOC, and NBS also provided us with insights and
perspectives valuable for the shaping of the study.
During the course of the study we also had occasion to contact many
individuals within industry and trade associations. Where appropriate we
have made reference to these contacts within the various reports. Frequently,
however, because of the study's emphasis on future developments with compara-
tive assessments of new technology, information given to us was of a confiden-
tial nature or was supplied to us with the understanding that it was not to be
credited. Therefore, we extend a general thanks to all those whose comments
were valuable to us for their interest in and contribution to this study.
Finally, because of the broad range of industries covered in this study,
we are indebted to many people within Arthur D. Little, Inc. for their parti-
cipation. Responsible for the guidance and completion of the overall study were
Mr. Henry E. Haley, Project Manager; Dr. Charles L. Kusik, Technical Director;
Mr. James I. Stevens, Environmental Coordinator; and Ms. Anne B. Littlefield,
Administrative Coordinator.
Members of the environmental team were Dr. Indrakumar L. Jashnani,
Mr. Edmund H. Dohnert and Dr. Richard Stephens (consultant).
Within the individual industry studies we would like to acknowledge the
contributions of the following people.
Iron and Steel; Dr. Michel R. Mounier, Principal Investigator
Dr. Krishna Parameswaran
Petroleum Refining; Mr. R. Peter Stickles, Principal Investigator
Mr. Edward Interess
Mr. Stephen A. Reber
Dr. James Kittrell (consultant)
Dr. Leigh Short (consultant)
xvii
-------
Pulp and Paper;
Olefins:
Ammonia:
Aluminum:
Textiles:
Cement:
Glass:
Chlor-Alkali;
Phosphorus/
Phos pho r i c Ac id:
Primary Copper:
Fertilizers:
Mr. Fred D. lannazzi, Principal Investigator
Mr. Donald B. Sparrow
Mr. Edward Myskowski (consultant)
Mr. Karl P. Pagans
Mr. G. E. Wong
Mr. Stanley E. Dale, Principal Investigator
Mr. R. Peter Stickles
Mr. J. Kevin O'Neill
Mr. George B. Hegeman
Mr. John L. Sherff, Principal Investigator
Ms. Nancy J. Cunningham
Mr. Harry W. Lambe
Mr. Richard W, Hyde, Principal Investigator
Ms. Anne B. Littlefield
Dr. Charles L, Kusik
Mr* Edward L. Pepper
Mr. Edwin L. Field
Mr, John W. Rafferty
Douglas Shooter, Principal Investigator
Robert M. Green (consultant)
Edward S, Shanley
John Willard (consultant)
Dr,
Mr,
Mr.
Dr.
Dr.. Richard F, Heitmiller
Dr. Paul A. Huska, Principal Investigator
Ms. Anne B. Littlefield
Mr.. J.. Kevin O'Neill
Dr, D. William Lee, Principal Investigator
Mr, Michael Rossetti
Mr, R, Peter Stickles
Mr, Edward Interess
Dr, Ravindra M. Nadkarni
Mr, Roger E. Shamel, Principal Investigator
Mr, Harry W. Lambe
Mr,, Richard P- Schneider
Mr. William V. Keary, Principal Investigator
Mr. Harry W. Lambe
Mr. George C. Sweeney
Dr., Krishna Parameswaran
Dr. Ravindra M. Nadkarni, Principal Investigator
Dr, Michel R. Mounier
Dr. Krishna Parameswaran
Mr. John L. Sherff, Principal Investigator
Mr. Roger Shamel
Dr. Indrakumar L. Jashnani
xviii
-------
ENGLISH-METRIC (SI) CONVERSION FACTORS
To Convert From
IP-
Metre2
Pascal
Metre3
t Joule
Pascal-second
Degree Celsius
Degree Kelvin
Metre
Metre /sec
Metre3
Metre2
Metre/sec
2
Metre /sec
0 Metre3
Ibf/sec) Watt
.c) Watt
Watt
Metre
Joule
Metre3
Metre
Metre
Metre
Pascal-second
Newton
Kilogram
Kilogram
Kilogram
Kilogram
Kilogram
Kilogram
Multiply By
4,046
101,325
0.1589
1,055
0.001
tซ = (t; -32;
fcK - tR/1'8
0.3048
0.0004719
0.02831
0.09290
0.3048
0.00002580
0.003785
745.7
746.0
735.5
0.02540
3.60 x 106
1.000 x 10~3
1.000 x 10~6
0.00002540
1,609
0.1000
4.448
0.4536
0.02916
1,016
1,000
907.1
1,000
Acre
Atmosphere (normal)
Barrel (42 gal)
British Thermal Unit
Centipoise
Degree Fahrenheit
Degree Rankine
Foot
3
Foot /minute
Foot3
Foot2
Foot/sec
Foot2/hr
Gallon (U.S. liquid)
Horsepower (550 ft-1
Horsepower (electric)
Horsepower (metric)
Inch
Kilowatt-hour
Litre
Micron
Mil
Mile (U.S. statute)
Poise
Pound force (avdp)
Pound mass (avdp)
Ton (assay)
Ton (long)
Ton (metric)
Ton (short)
Tonne
Source: American National Standards Institute, "Standard Metric Practice
Guide," March 15, 1973. (ANS72101-1973) (ASTM Designation E380-72)
xix
-------
I. INTRODUCTION
A. BACKGROUND
Industry in the United States purchases about 27 quads* annually, approx-
imately 40% of total national energy usage.** This energy is used for chemical
processing, raising steam, drying, space cooling and heating, process stream
heating, and miscellaneous other purposes.
In many industrial sectors energy consumption can be reduced significantly
by better "housekeeping" (i.e., shutting off standby furnaces, better thermo-
stat control, elimination of steam and heat leaks, etc.) and greater emphasis
on optimization of energy usage. In addition, however, industry can be
expected to introduce new industrial practices or processes either to con-
serve energy or to take advantage of a more readily available or less costly
fuel. Such changes in industrial practices may result in changes in air,
water or solid waste discharges. The EPA is interested in identifying the
pollution loads of such new energy-conserving industrial practices or proc-
esses and in determining where additional research, development, or demonstra-
tion is needed to characterize and control the effluent streams.
B. CRITERIA FOR INDUSTRY SELECTION
In the first phase of this study we identified industry sectors that have
a potential for change, emphasizing those changes which have an environmental/
energy impact.
Industries were eliminated from further consideration within this assign-
ment if the only changes that could be envisioned were:
energy conservation as a result of better policing or "housekeeping,"
better waste heat utilization,
fuel switching in steam, raising, or
power generation.
*1 quad = 1015 Btu
**Purchased electricity valued at an approximate fossil fuel equivalence of
10,500 Btu/kWh
-------
After discussions with the EPA Project Officer and his advisors,
industry sectors were selected for further consideration and ranked using:
Quantitative criteria based on the gross amount of energy (fossil
fuel and electric) purchased by industry sector as found in U. S.
Census figures and from information provided from industry sources.
The primary copper industry purchased Q.081 quads out of the 12.14
quads purchased in 1971 by the 13 industries selected for study, or
0.3% of the 27 quads purchased by all industry (see Table 1-1).
Qualitative criteria relating to probability and potential for
process change, and the energy and effluent consequences of such
changes.
In order to allow for as broad a coverage of technologies as possible, we
then reviewed the ranking, eliminating some industries in which the process
changes to be studied were similar to those in another industry planned for
study. We believe the final ranking resulting from these considerations identi-
fies those industry sectors which show the greatest possibility of energy con-
servation via process change. Further details on this selection process can be
found in the Industry Priority Report prepared under this contract (Volume II).
On the basis of this ranking method, the primary copper industry
appeared in tenth place among the 13 industrial sectors listed.
TABLE 1-1
SUMMARY OF 1971 ENERGY PURCHASED IN
SELECTED INDUSTRY SECTORS
SIC Code
In Which
Industry Sector 10 Btu/ir. Industry Pound
1. Blast furnaces and steel mills 3.t9(1) 3312
2. Petroleum refining 2.96 2911
1. Paper and allied products 1.59 26
ป. Olefins 0.9S4^3' 2818
5. Ammonia 0.63^* 287
6. Aluminum 0.59 3334
/. Textiles 0.54 22
H. Cement 0,52 3241
9. Class n.3.1 3211, 3221, 3229
10. Alkalies and chlorine 0.24 2.312
11. Phosphorus and phosphoric f,^
acid production 0.121 ' 2S19
12. Primary copper 0.081 3331
13. Fertilizers (excluding ammonia.) 0.078 287
^ Estimate for 1967 reported by FEA Project Independence Blueprint, p. 6-2,
USCPO, No-venter 1974.
Includes captive consumption of energy from process byproducts (FEA Project
Independence Blueprint)
01ซflQS only, includes energy of feedstocks: ADL estimates
^ ^Amonia feedstock energy included: ADL estimates
153 ADL estimates
Source: 1972 Census of Manufactures, FEA Project Independence Blueprint,
USCPO, Hovenber 1974, and ADL estimates.
-------
C. CRITERIA FOR PROCESS SELECTION
In this study we have focused on identifying changes in the primary
production processes which have clearly defined pollution consequences.
In selecting those to be included in this study, we have considered the
needs and limitations of the EPA as discussed more completely in the Indus-
try Priority Report mentioned above. Specifically, energy conservation has
been defined broadly to include, in addition to process changes, conserva-
tion of energy or energy form (gas, oil, coal) by a process or feedstock
change. Natural gas has been considered as having the highest energy form
value followed in descending order by oil, electric power, and coal. Thus,
a switch from gas to electric power would be considered energy conservation
because electric power could be generated from coal, existing in abundant
reserves in the United States in comparison to natural gas. Moreover, pollu-
tion control methods resulting in energy conservation have been included
within the scope of this study. Finally, emphasis has been placed on process
changes with near-term rather than long-term potential within the 15-year
span of time of this study.
In addition to excluding from consideration better waste heat utiliza-
tion, "housekeeping," power generation, and fuel switching, as mentioned
above, certain options have been excluded to avoid duplicating work being
funded under other contracts and to focus this study more strictly on "process
changes." Consequently, the following have also not been considered to be
within the scope of work:
Carbon monoxide boilers (however, unique process vent streams
yielding recoverable energy could be mentioned);
Fuel substitution -in fired process heaters;
Mining and milling, agriculture, and animal husbandry;
Substitution of scrap -(such as iron, aluminum, glass, reclaimed
textiles, and paper) for virgin materials;
Production of synthetic fuels from coal (low-and high-Btu gas,
synthetic crude, synthetic fuel oil, etc.); and
All aspects of industry-related transportation (such as trans-
portation of raw material).
D. SELECTION OF PRIMARY COPPER INDUSTRY PROCESS OPTIONS
Within each industry, the magnitude of energy use was an important cri-
terion in judging where the most significant energy savings might be realized,
since reduction in energy use reduces the amount of pollution generated in
the energy production step. Guided by this consideration, candidate options
for in-depth analysis were identified from the major energy consuming process
steps with known or potential environmental problems.
-------
After developing a list of candidate process options, we assessed
subj ectively
pollution or environmental consequences of the process change,
probability or potential for the change, and
* energy conservation consequences of the change.
Even though all of the candidate process options were large energy users,
there was wide variation in energy use and estimated pollution loads between
options at the top and bottom of the list. A modest process change in a major
energy consuming process step could have more dramatic energy consequences
than a more technically significant process change in a process step whose
energy consumption is rather modest. For the lesser energy-using process
steps process options were selected for in-depth analysis only if a high
probability for process change and pollution consequences was perceived.
Because of the time and scope limitations for this study, we have not
attempted to prepare a comprehensive list of process options or consider all
economic, technological, institutional, legal or other factors affecting
implementation of these changes. Instead we have relied on our own back-
ground experience, industry contacts, and the guidance of the Project Officer
and EPA advisors to choose promising process options (with an emphasis on
near-term potential) for study in the primary copper industry.
Copper extraction from sulfide ores is traditionally divided into four
segments:
mining- where ore containing 0.6-2% copper is mined;
beneficiation - where the copper-containing minerals are separated
from the waste rock to produce a concentrate containing about 25%
copper;
smelting - where concentrates are melted and reacted to produce 98%
pure "blister" copper or "anode" copper (copper 98-99% pure requir-
ing further refining); and
refining - where blister copper is refined electrolytically to
produce 99.9% pure "cathode" copper.
The conventional approach and the potential process changes are shown in
Table 1-2. Of these, the following candidate process options were considered
in order to illustrate as wide a spectrum as possible of the consequences of
process change:
Outokumpu flash smelting,
Noranda process,
Mitsubishi process,
-------
TABLE 1-2
COPPER EXTRACTION FROM SULPIDES: CONVENTIONAL PROCESSING AND PROCESS OPTIONS
Stage
Mining
Process Options
Ben.ef iciation
Smelting
Refining
Conventional Smelting - Slag Discard*
e Outokumpu Smelting*
Noranda Smelting*
Mitsubishi Smelting*
ซ Oxygen Use in Smelting*
Metal Recovery from Slag*
Brixlegg Process
Other Continuous Smelting Processes
Conventional Electrorefining*
Arbiter Process*
Ferric Ion Leaching of
Sulfides
IM
Bacterial Leaching
Sulfuric Acid Leaching of
Sulfides
Roas't-Leach-Electrowin
Nitric Acid Leaching
Process options studied in detail in this report.
-------
Oxygen use in smelting,
Metal recovery from slags (flotation or electric furnace),
Arbiter process,
Sulfuric acid leaching of oxide ores, LIX, electrowinning,
Ferric ion leaching,
Brixlegg process,
Bacterial leaching, and
Sulfuric acid leaching of sulfides.
From this listing we chose (with the concurrence of the Project Officer)
the first six for detailed analyses for the following reasons:
The first three processes, with a high specific capacity, are
representative of the newer pyrometallurgical processes which
reduce total energy requirements by 30-50%. From the smelting
unit they produce a concentrated stream of sulfur dioxide which
can be economically converted to sulfuric acid. The adoption of
this technology would avoid the use of reverbs which produce
dilute S02~containing gas streams. Sulfur capture, 50-70% for
current technology, would increase to more than 90%.
* Using oxygen in the smelting process is one way of decreasing fuel
and increasing the specific capacity of the newer processes. We
believed that an examination of oxygen use might provide an assess-
ment of the fuel savings and the decrease in capital requirements
through an increase in specific capacity.
In conventional smelting, the slag from the smelting furnace is
discarded. Since the metal content of the slag increases with
matte* grade, the matte grade has to be held low. The ability to
economically recover metal from slags makes it possible to use the
newer smelting processes mentioned above, all of which operate with
much higher matte grades.
There is considerable interest in the United States in the hydro-
metallurgical extraction of copper from sulfide concentrates to
avoid air pollution problems associated with conventional smelting.
We selected the Arbiter process because it is ahead of others in
terms of plant construction and plant size and because detailed
information about the process has been published.
"*
See Glossary
-------
The remaining processes are less important for the following reasons:
Sulfuric acid leaching of oxides is not a primary copper recovery
process but can be considered as an acid neutralization/disposal
technique that also recovers copper from resources previously con-
sidered as marginal.
Ferric ion (ferric chloride) leaching and sulfuric acid leaching
are processes about which there are not sufficient data for a
detailed evaluation.
Bacterial leaching has long-terra potential only in conjunction with
solution mining techniques.
The Brixlegg process has been used only on a small scale. Even
fewer data are available on impurity behavior in this process than
with the other pyrometallurgical processes, such as Outokumpu,
Noranda, and Mitsubishi.
For each process, we evaluated the capital and operating costs to pin-
point economic factors which would influence the adoption of new technology.
Operating costs for the base line were also calculated on the same bases.
In each case we assumed that the plant would meet existing EPA standards for
ambient air quality (S02 and particulates) and the 1983 Effluent Limitation
Guidelines for aqueous effluents. Our calculations do not include any
allowances for future standards affecting trace metals since such standards
have not been selected and because there are little or no data on the potential
emissions of such trace metals from these processes. In all pyrometallurgical
processes, the containment and control of airborne pollutants is a difficult
achievement and there will always be certain low levels of uncontrolled fugi-
tive emissions.
There are two issues of controversy in the copper industry, relating
primarily to existing smelters. The first is the trade-off between permanent
controls and intermittent controls for meeting ambient air quality standards.
The technological requirements for meeting ambient air quality standards vary
greatly from location to location and there does not appear to be a single
technological approach which will be the most cost-effective solution in all
cases. The second is the trade-off between single absorption and double
absorption acid plants for retrofit applications. Although both these issues
will be resolved in the future, they do not affect the evaluation of the
process options selected here since we are considering the general applica-
tion of new technology in new locations where New Stationary Source Standards
(double absorption acid plants) apply.
-------
II. FINDINGS, CONCLUSIONS, RECOMMENDATIONS
A. PYROMETALLURGICAL PROCESS
Tables II-l and II-2 summarize cost/energy/environmental consequences of
process options studied in this report.
The new pyrometallurgical processes (Outokumpu, Noranda, and Mitsubishi)
have two characteristics which make them more energy efficient and less pollut-
ing compared to conventional reverb smelters:
They utilize the heat of oxidation of sulfur and iron to supply part
of the process energy requirements.
They produce steady-concentrated streams of S02 suitable for sulfuric
acid manufacture.
The reduction in energy consumption is significant, about 30-50% (the latter
when coupled with oxygen enrichment). The processes are quite flexible in their
ability to use any form of energy: gas, oil, or coal. The reduction in S02
emissions is also significant. Compared to sulfur capture of 50-70% for
'conventional smelting, the new processes achieve a sulfur capture of over 90%.
Since only concentrated S02 streams are produced, the cost of controlling over
90% S02 in the new processes is about the same as that for 50-70% control for
conventional smelting.
The forces which would drive the U.S. industry to adopt this technology
would be EPA regulations limiting S02 emissions, and plant modifications.and
high energy costs in the future. Other factors should also b^ considered,
however. The U.S. copper industry has traditionally increased capacity by
expanding and modifying existing smelters and refineries because of the large
capital investment per dollar of revenue potential, relatively lot growth in
demand, cyclical nature of demand, and the international trade in ^' ined metal.
In an inflationary economy the production costs from existing, partially
depreciated facilities are much lower than those for new facilities. New
Source Performance Standards affecting the copper industry will constrain
capacity expansion at existing smelters.
The sulfuric acid produced would have to be utilized or disposed of. The
typical western U.S. smelter locations are distant from major sulfuric acid
markets, and the sulfuric acid has been utilized for leaching of marginal
resources (mine dumps, tailings, oxide ores, etctf) or for making wet process
phosphoric acid locally. The leaching of dumps and surface deposits without
contamination of groundwater is possible in the arid west but might not be
possible in other parts of the United States. As a last resort, neutralization
with limestone or the reduction of concentrated 862 streams to elemental sulfur
would have to be considered. These options would exert their own impacts.
8
-------
TABLE II-l
QUALITATIVE SUMMARY OF COSTS/ENERGY/ENVIRONMENTAL CONSEQUENCES OF PROCESS OPTIONS IN THE COPPER INDUSTRY
Base Line Process: Conventional Smelting and Refining of Concentrates
COSTS
VWPRfY
fil^EiIVvJ.
ENVIRONMENTAL
PROCESS OPTIONS
Outokumpu
Smelting
Slightly higher
capital cost.
Lower operating
cost because of
energy savings.
Slightly lower
air pollution
control cost be-
cause more con-
centrated so2
streams are
handled .
Lower energy
cost use.
Overall saving
of 30-50Z.
Higher levels of
sulfur recovery
are possible
(over 90%).
Minimal or no
water pollution.
Solid vaste prob-
lem unchanged.
No change In feed-
stock but possible
changes in impurity
distributions which
might require
different product
treatment .
Noranda
Smelting
Same as
Outokumpu
Same as
Outokumpu
Same as
Outokumpu
Mitsubishi
Smelting
Same as
Outokumpu
Same as
Outokumpu
Same as
Outokumpu
Oxygen Use
in Smelting
Slightly lower
capital cost.
Slightly higher
maintenance re-
quirements.
Lower operating
costs.
Lower energy
use even after
energy used in
oxygen manu-
facture is taken
into account.
No major change.
Possible change
in impurity dis-
tributions which
might require
different product
treatment.
Metal Recovery from Slaps
Electric Furnace
Lower capital
cos t . Lower
operating cost,
but lower copper
recovery.
About the same
energy require-
ments .
Solid waste
problem un-
changed. No
water pollution.
Flotation
Higher capital
cost. Higher
operating cost,
but higher copper
recovery.
About the same
energy require-
ments.
Solid waste of
finer particle
size. Potential
water pollution
problem.
Arbiter Process
Lower capital cost
at smaller plant
size. Higher op-
erating cost.
Higher energy re-
quirements. Most
of the energy re-
quired is in the
form of electricity.
Minimal air pollution.
Change in type of
aolid wastes generated.
Some aqueous effluents.
VO
-------
TABLE II-2
QUANTITATIVE COMPARISON OF BASE LINE AND ALTERNATIVE PROCESSES IN THE COPPER INDUSTRY
Outokumpu
Reverb Electro- Smelting Slag Cleaning
Smelting Refining No Norahda 'Mitsubishi Electric
(Base Case) (Base Case) Oxygen Oxygen Smelting Smelting Furnace Flotation Arbiter Process
Environmental
Pollution Control
Costs ($/ton of
Copper) 54-64 - 59 59 46 59 15 1.90
Energy Consumption
106 Btu/ton
of Copper3 23-26 4.6 15 13.2 12.5 12 1.2C 0.8ฐ 61
Process Economics
Investment ($/
annual ton) 650-750 450 750 500 750 750 9.51" 20^ 1000
Pollution Control
and Operating Coat . H
($/ton)b 340-370 230 336 259 336 333 10.47 ' 12.40 ' " 636
alncludes electricity at 10,500 Btu/kWh; oxygen at 360 kWh /ton.
Includes pre-tax return on investment - excludes concentrate cost.
c$/ton of slag.
Equivalent to about $200-360/ton of recovered copper.
-------
Like conventional processes, the newer pyrometallurgical processes
offer significant economies of scale, the smallest economic size being
approximately 100,000 ton/yr of copper.
The major shortcoming of the new processes is that their applicability to
impure concentrates (concentrates high in As, Sb, Bi, Pb, Zn, Se, Te, etc.) is
unproven. Until this issue is resolved, the new processes would be utilized
for building large smelters to smelt clean concentrates in regions where acid
markets are available.
In spite of this, we believe that 3-4 new pyrometallurgical plants
(Outokumpu, Noranda, etc.) will be built in the United States by 1985-1990,
which would account for about 500,000 ton/yr or about 15% of anticipated pro-
duction in 1990.
B. OXYGEN USE IN SMELTING
The use of oxygen in smelting decreases fuel requirements and is energy
efficient overall because the decrease in fuel requirements is usually larger
than the energy required for oxygen separation. The use of oxygen enrichment
can increase capacities of existing units and decrease capital costs per unit
of output from new units. Lesser benefits are the ability to melt more scrap
and the production of more concentrated S02 streams in some cases.
We believe that these advantages (and the absence of disadvantages other
than higher operating temperatures and consequently increased maintenance)
will lead to the widespread adoption of oxygen enrichment.
C. RECOVERY OF METALS FROM SLAG
The new pyrometallurgical processes are economically viable only if the
metal contained in the ^slag from the primary smelting unit can be recovered.
Thus, these processes are important adjuncts to the new smelting processes and,
in addition, might be used in existing smelters (e.g., treatment of converter
slags via flotation). The processes can decrease flux requirements, increase
the overall recovery of copper in some cases, and improve smelting unit
operations.
Of the two processes, electric furnace cleaning is somewhat cheaper, but
the higher copper recovery in slag flotation compensates for the higher costs.
The energy requirements are about equal for both processes.
While slag flotation produces a finely ground slag that is different
from the slag from conventional processing, this slag can be placed into land
disposal areas which have been especially prepared to mitigate the possibility
of long-term emissions to the environment. Although the particular handling
method will be unique to the geological aspects of the area, it is not expected
that slag disposal will present a significant environmental problem.
11
-------
D. THE ARBITER PROCESS
The Arbiter process is significantly more expensive for treating
chalcopyrite concentrates than conventional smelting and refining. We believe
that this process will be used on other concentrates with favorable mineralogy,
e.g., chalcocite concentrates, in locations distant from sulfuric acid markets
and where a full-sized pyrometallurgical plant is risky because of growth in
copper demand.
The Arbiter process and hydrometallurgical processes in general are not
energy efficient and utilize the same or slightly more energy than conventional
smelting and refining. The leached solid wastes will require land disposal
into areas prepared to prevent groundwater leaching of soluble substances and
to prevent airborne particulates. Because the plant will be located generally
in the semi-arid western United States, such a disposal area can be established
with a high degree of confidence that long-term, adverse environmental impact
will not occur.
E. RESEARCH AREAS
Research in the following areas would provide more information about these
processes and might resolve the problems preventing the adoption of these
technologies.
1. Impurity Distributions
The behavior of impurities in each process, their distribution between
gas, slag, matte and metallic phases, and forms in which impurities
leave the process units (as particulates, slag, etc.)-
In oxygen-enriched smelting, an examination and definition of the
changes in impurity distributions resulting from higher temperatures.
Impurity distributions in hydrometallurgy.
Impurity forms in waste streams from pyrometallurgical and hydro-
metallurgical processes.
Reasons: An understanding of impurity distributions is necessary to
determine emissions of trace metals and the need for
pollution control or process change to prevent these
emissions.
2. Impurity Removal
Methods for removing impurities (e.g., Bi) from blister copper via
modified fire-refining procedures. If impurities can be remoyed from
copper, one-step smelting can be used. This would significantly
decrease fugitive SC>2 emissions from smelters.
Methods for removal of arsenic from elemental sulfur produced by
S02 reduction.
Techniques for impurity removal from concentrates via pretreatment.
12
-------
3. Feedstocks
Verifying the applicability of new 'smelting technology to impure
concentrates.
Verifying the applicability of new smelting technology to smelt
calcines from roasters.
4. Metal Recovery/Separation
Methods for separation and recovery of metals from flue dust; in
particular, the treatment of flue dust containing arsenic, antimony,
lead, and bismuth.
Methods for the recovery of precious and trace metals from hydro-
metallurgical residues.
Methods for the recovery of copper (also zinc and lead) from
slag flotation tailings.
Methods for reduction of copper in slag during pyrometallurgical
slag cleaning, e.g., carbon reduction in a rotary converter.
Conversion of ferric oxide sludge from hydrometallurgical processes
to a form suitable for use in other industries.
5. Miscellaneous
High-temperature refractories for oxygen smelting.
Better process control techniques to reduce manpower requirements,
particularly in pyrometallurgy.
13
-------
III. INDUSTRY OVERVIEW
The United States has ranked first or second among world producers of
refined copper since before the turn of the century as shown in Table III-l.
Most of the copper mined in the United States is produced in five western
states Arizona, Utah, New Mexico, Montana and Nevada (Table III-2).
In the United States, 27 major mines account for over 95% of the copper output.
A major portion of the mine production is accounted for by producers such as
Kennecott, Phelps Dodge, Anaconda, Newmont and Inspiration, who are integrated
from mining through fabrication. Numerous small companies participate only in
mining and beneficiation and sell or arrange for toll treatment of their
concentrates at the custom smelters of Asarco. Two of the larger mining
companies are Duval and Cyprus.
TABLE III-l
LEADING WORLD PRODUCERS OF REFINED COPPER
(Short Tons)
Rank Country 1970
1 United States 2,242,700
2 U.S.S.R.1 1,185,000
3 Japan 777,500
4 Zambia 640,100
5 Canada 543,000
6 Chile 512,700
Estimate
Source: "Non-Ferrous Metal Data 1974", American
Bureau of Metal Statistics, N.Y.
14
-------
TABLE III-2
U.S. MINE PRODUCTION OF COPPER
(Short Tons)
State
Arizona
Utah
New Mexico
Montana
Nevada
Michigan
Other1
TOTAL 1,199,290 1,719,657 1,664,840 1,593,590
1968
631,300
228,300
92,300
64,862
72,870
74,590
34,708
1970
917,918
295,738
166,278
120,412
106,688
67,543
45,080
1972
908,612
259,507
168,034
123,110
101,119
67,260
37,198
1974
852,650
230,088
197,374
133,675
83,725
67,297
28,781
California, Colorado, Idaho, Maine, Missouri, Pennsylvania, Tennessee,
Oklahoma
Source: U.S. Bureau of Mines, Mineral Yearbook
Copper extraction is traditionally divided into four segments: mining,
beneficiation, smelting, and refining. A generalized flowsheet of copper
processing is shown in Figure III-l. This report addresses technological
changes in smelting and refining.
Smelting practice in the United States .is fairly uniform'from smelter to
smelter. About half of the copper smelters roast their charge prior to feeding
to the reverberatory (reverb) furnace (calcine smelting), while the other half
feed the concentrates directly (green feed smelting). The subsequent steps
consist of melting the charge in the reverberatory furnace to form matte, a
mixture of copper and iron sulfides and a slag (which is discarded); converting
the matte to blister copper and finally fire-refining to remove the oxidizable
impurities and/or electrolytic refining to remove and recover the precious
metal impurities in the copper. The major portion of the blister copper is
electrorefined.
Traditionally, the smelters have been situated near the mines in order to
minimize transportation costs for concentrates. At present, there are 18
primary smelters in the United States, thirteen of them west of the Mississippi.
There are 16 electrolytic refineries which have traditionally been located near
consumers on the East Coast where they account for about half of the electro-
refining capacity. Recently some refineries have been built in the west near
the smelters. Locations of U.S. smelters and refineries are shown in Figure
III-2.
15
-------
MINING
BENEFICIATION
SMELTING
REFINING
WASTE
ROCK
TO DUMP
WATERS.
REAGENTS
AIR
FIRE-REFINED
COPPER
ANODE
MUD
Figure III-l. Generalized Flowsheet for Copper Extraction From Sulfide Ores
16
-------
REFINERIES
^SMELTERS
SOURCE: Mซlal!Weok, 1972
Figure III-2, Where the Smelters and Refineries are Located
-------
Fabricating companies are the principal consumers of refined copper. They
work the metal into semi-finished form such as sheet, strip, rod, tube, wire,
and extruded or rolled shapes which are the raw materials for the manufacturing
industries.
Although copper has many end uses, they can be classified into five broad
categories:
1. Electrical/electronic equipment
2, Building construction
3. Transportation
4. Non-electrical industrial equipment, and
5. Ordnance.
Other applications include use in chemicals, inorganic pigments, jewelry, and
coins.
Through subsidiaries or stockholdings, many large domestic producers
operate foreign properties in both the developed and developing countries and
are involved domestically in the production of other nonferrous metals such as
aluminum, lead, and zinc. The financial posture of some of the companies has
been changed by expropriation of certain foreign holdings, and nationalization
in other countries continues to be a threat.
The domestic copper, lead, and zinc industries are interdependent to the
extent that byproducts or residues from one industry form a part of the input
to the other. The western copper, lead, and zinc industry is a significant
producer of byproducts such as silver, gold, and bismuth. In certain instances,
especially at the western lead smelters, the value of byproducts can equal or
exceed that of the primary productlead. In general, the byproduct supply
and production is inelastic, i.e., not dependent on demand or price of the
byproduct but dependent only on the primary metal production. Any factor
(pollution-related or otherwise) that changes the output of the primary product
would automatically affect byproduct output. An apparent exception might be
the western lead smelter where, because of the high volume of coproducts,
higher coproduct prices (e.g., silver, bismuth, etc.) can decrease the sen-
sitivity of primary metal production to primary metal price.
An important aspect of the entire primary nonferrous industry is that
traditionally the smelters and refineries have been operated as service opera-
tions at a fixed and relatively low profit margin which is not very sensitive
to the price of the finished product. Hence, the impact of any change in
price of the primary metal has to be reflected back and affects directly the
value of the concentrate. In the 1960's, the traditional rule-of-thumb ,in deter-
mining concentrate value in the copper industry was to assume 4c/lb for smelting
charges and 5ฃ/lb for refining charges so that the value of copper contained
in the concentrate was very approximately 9/lb below the cathode or wirebar
market price. (The 1975 operating margin in the copper industry was about
16-18c/lb.)
18
-------
Because of this mechanism, any increase in smelting or refining costs
cannot "be "absorbed" by the smelter or refinery but can only be passed back-
ward to the mine, and the net-back (the "net concentrate value realized at the
mine, e.g., smelter payment minus transportation costs) would be decreased.
Should the market supply/demand constraints permit an upward adjustment in
primary metal price, this increase would then be reflected back to the mine.
The mechanism described above is of primary importance to custom and toll
smelters since it is possible that a decreased concentrate value can result in
mine closings and loss of smelter feed material. However, essentially the
mechanism operates in the case of producers integrated from mining through
smelting and refining since the concentrate transfer price is related to the
primary metal price and again the mines would have to absorb the increased
smelting and refining costs under adverse market conditions.
Given generally similar operating, tax, and pricing structures, and
because of the 'overlap in company participation in copper, lead and zinc mining,
smelting, and/or refining, the major companies are somewhat similar from the
financial viewpoint. Sales and earnings of the primary nonferrous metals
companies are cyclical. Major influences on earnings are the operating rate
and metal prices. The latter fluctuate more than annual consumption or demand
since prices tend to be sensitive to small imbalances between supply and demand
and varying international situations.
Over the past 20 years, world consumption of copper has been increasing
at an average annual rate of 4-4.5%. Despite competition from plastics and
aluminum, copper consumption is expected to increase at about the same rate
worldwide over the next decade with a slightly lower rate in the industrialized
countries. The United States is a leading producer and consumer of primary
copper, accounting for about one third of Free World production and consumption.
Despite this, the United States has been in a position of undersupply since
the early 1960's as shown in Figure III-3. Domestic mine production has been
increasing recently at an adjusted rate of about 3.5% per year, while world
mine production has been increasing at about 5% per year.
19
-------
2,000
1,500
CO
O
cc.
O
O
O
O
cc.
Q.
1,000
500
-500 L-
U.S. REFINED
COPPER PRODUCTION
1975
NET EXPORTS
SOURCE: Arthur D. Little, U.S. Bureau of Mines, American Bureau of Metal Statistics.
Figure III-3. Trends in Net U.S. Imports and Production of Refined Copper
20
-------
IV. COMPARISON OF CURRENT AND ALTERNATIVE PROCESSES
A. REASONS FOR CHOOSING OPTIONS TO BE ANALYZED IN DEPTH
1. Problems of Current Technology
The smelting technology in the United States evolved in a framework of
low energy costs in locations distant from sulfuric acid markets and urban
population centers. Thus the technology was not aimed at recovering sulfur
values as sulfuric acid (as is the case abroad) and was not particularly
efficient in its use of energy. Several changes have occurred in the past
five years on the economic and regulatory scene which may influence the
adoption of new technology for construction of new smelters or for switching
to coal firing in existing reverbatory furnaces. These changes are as follows:
Energy costs for smelting have increased rapidly. Cheap natural
gas, the fuel used by most smelters, is not now available to the
smelters particularly in peak demand months (winter) and might not
be available at all in the future.
Emissions of 862 to the atmosphere have to be controlled. After
several years of debate, all issues relating to S02 emissions are
yet to be resolved. These issues are as follows:
- At present the New Source Performance Standards for primary
copper smelters contains an exemption for reverbatory furnaces
smelting "impure" concentrates (concentrates containing As,
Sb, Bi, etc.). About 30% of the feed sulfur is emitted from
the reverbs.
- Emissions from streams containing high concentrations of S02
(converter and roaster gases) have to be controlled by tech-
nology such as sulfuric acid plants. Double absorption plants
(or equivalent) are mandatory for new sources. This combination
of uncontrolled reverbs but controlled roasters and converters
can recover from about 50-70% of the sulfur in the feed in the
form of sulfuric acid. New smelting technology can recover a
larger fraction (over 90%) of the sulfur in the feed.
21
-------
- Federal Ambient Air Quality Standards which define permissible
ground level concentrations of S02 have to be met at all times
by a combination of permanent controls, e.g., acid plants, and
production curtailment. The issue of when production curtail-
ment should be used has not been resolved. The degree of per-
manent controls necessary to meet Ambient Air Quality Standards
varies greatly with location. The control of roaster or con-
verter emissions is adequate for meeting Airbient Air Quality
Standards only in certain locations.
Compared to the cost of SC>2 control, the costs of complying with
other existing pollution control legislationt e.g., water and
solid wastes, is quite small. However, this situation could
change as new standards are proposed for controlling emissions
of other substances.
The acid produced by smelters cannot be economically transported to
traditional acid markets, and it has to be disposed of (by selling
it at a low price) in the general vicinity of the smelter. This
resulting availability of low-priced acid in the west has made it
possible to use it to make wet-process phosphoric acid from low
grade western phosphate rock or to use it to leach mine dumps and
low grade deposits of copper ore which cannot be treated conven-
tionally for the extraction of the contained copper. An alternative
is its neutralization with limestone. This is being practiced
indirectly on copper ores high in limestone.
Copper smelting and refining have been capital intensive in the past,
and the capital intensity has increased significantly in recent
years because of pollution control requirements and inflationary
pressures. Also, pyrometallurgical operations involved significant
economies of scale, and the smallest economic size is over 100,000
ton/yr of copper. Because of the small growth rate in copper con-
sumption, an increment in capacity of 100,000 ton/yr or more for
any United States company implies significant risk since their market
share would have to increase by 20% or more. This is why smelting >
and refining capacity has generally increased by modifications of
existing plants rather than by the construction of new plants in
remote locations. We believe that the New Source Performance
Standards will affect this traditional mode of capacity expansion
since emissions from the new portion of the plant may require differ-
ent treatment in order to comply with New Stationary Source Perfor-
mance Standards,
2. Reasons for Choosing Options for Detailed Analysis
We selected an illustrative group of process changes which we believe are
likely to be implemented by the U.S. industry in the near term. The set of
process changes is not exhaustive but is representative of industry responses
to one or more of the problems described above in Section 1.
22
-------
The topics selected for detailed analysis are:
New smelting processes
Outokumpu flash smelting
Noranda process
Mitsubishi process
The use of oxygen in smelting
Slag cleaning processes
Arbiter process
The first three are illustrative of new -high-intensity pyrometallurgical proc-
esses. The use of oxygen in smelting, particularly in the context of these
new processes, is an interesting development in that it not only decreases net
energy use but also increases smelting intensity. The latter can significantly
decrease the capital cost ,of a smelter per unit of output. The slag cleaning
processes (flotation and electric furnace cleaning) are important adjuncts to
the new processes since these processes would not be economically viable
unless residual copper (up to 12%) in the slag was recovered. The Arbiter
process is the earliest of several hydrometallurgical processes being adopted
for the extraction of copper from sulfide concentrates. These processes
produce cathode copper directly and release sulfur from the concentrates in
forms other than S02. It is feasible to build plants sized around 40,000 tons
of copper per year at a unit cost of about $1000/annual ton of copper, which
is about the same as the unit cost of a large (over 100,000 ton/yr) copper
smelter and refinery. However, such a hydrometallurgical process does not
offer significant economies of scale, and larger plants would require the
same unit capital investments.
The sections that follow discuss the base line and the process options.
Background literature references used in evaluating each process are given at
the end of the report.
B. BASE LINE: CONVENTIONAL SMELTING AND REFINING
1. Introduction
Conventional smelting and refining is the base line against which poten-
tial process alternatives have been compared in order to evaluate pollutional
and energy consequences of these process alternatives.
Conventional smelting of sulfide concentrate involves the smelting of
concentrates in the reverberatory furnace (reverb) either directly (green
charge smelting) or after roasting (calcine smelting). The mixture of molten
sulfides from the reverb is converted to blister copper in converters. A
23
-------
detailed description of conventional smelting is presented in Appendix A.
Both green charge and calcine smelting form the base line for comparison of.
other new smelting approaches since both approaches are widely used in the
United States.
Conventional electroref ining (also described in Appendix A) purifies the
smelter output to cathode quality copper. We have used smelting plus refining
as the base line for comparison of the Arbiter hydrometallurgical process since
it produces cathode copper directly from concentrates.
2. Effluents
Figure IV-1 is a schematic flow diagram of the conventional smelting and
refining process. The emission sources of pollutants are shown in four
categories: air, solid, water, and fugitive. The latter are air emissions
that come from diffuse sources. Table IV-1 shows the magnitude of these
streams and the major constituents.
a. Air Pollution
(1) Current Methods
At present, there are three methods in use at copper smelters for reduc-
ing the sulfur dioxide concentrations in the vicinity of a smelter. These
are: the use of a tall stack to disperse dilute gas streams; the production
of sulfuric acid by the contact process from concentrated gas streams to
a'chieve a degree of reduction in emissions; and production curtailment.
The tall stack discharges sulfur dioxide at such heights that the gas is
diluted when dispersed into the lower atmosphere. It is possible to add
preheated air into the stack to achieve additional dispersion and dilution.
Because tall stacks and preheated air can achieve dispersion and dilution when
used, in conjunction with other means of limiting emissions, there is no simple
relationship which can predict ambient concentrations as a function of percent
sulfur recovery. The overall control strategy has to be well defined, and
local weather patterns have to be considered.
The contact sulfuric acid process is well established for treating
containing off -gases from metallurgical plants. Modern contact acid plants
require at least 4.5-5% sulfur dioxide in the feed gas in order to operate
autogenously (i.e., without external heat). For handling lower concentrations
of S02, an additional fuel input is required. The acid plant size is primarily
a function of the volume of gas handled. Hence, for a constant acid output,
an acid plant operating on more dilute gases is much larger (and more expen-
sive) than an acid plant operating on more concentrated gases. With the
currently used vanadium pentoxide catalysts, the upper level of S02 concentra-
tion in the feed gas to an acid plant is between 7% and 9%. Gas streams more
concentrated than this require dilution.
24
-------
CONCENTRATE
10
Ol
!
RECYCLED DUST
A'MST ,.
r n
OPTIONAL
ROASTING
t 1
SLAG
GAS COOLING
DUST CATCHING /^
DUST
REVERB
SMELTING
ฉ
MATTE.
ป GAS " ป CTซ ^fATlSTACr
r tUULINtj -" S^J/ "
|_ * /v DUST
CONVERTING
BLISTER
ANODE FURNACE
1
CASTING
ELECTROLYTIC
REFINING
ELECTROLYTIC
COPPER
H2S04
ฉ
ฉ
EFFLUENT TYPES: ^^ -AIR; ^J -WATER; ^\ -SOLID WASTES; ^M -FUGITIVE
'ESP - ELECTROSTATIC PRECIPITATOR
Figure IV-1. Sources of Emissions in Conventional Copper Smelting and Refining
-------
TABLE IV-1
EMISSIONS FROM CONVENTIONAL SMELTING
Stream
Air Pollution
A-l - Reverb Gas
A-2 - Acid Plant Tail Gas
A-3 - Anode Furnace Gas
Stream Size
82,000 scfm
38,800 scfm
NA
Major Constituents
x>
S02: 1-2%; 02: 5%; Participates: 50 yg/NMJ
SO : 0.2%
Flue Gas, Some SO
N>
Water Pollution
W-l - Slag Granulation
W-2 - Acid Plant Slowdown
W-3 - Contact Cooling
W-4 - Black Acid Bleed
1200 gal/ton
340 gal/ton
180 gal/ton
17 gal/ton
TDS, SSS
TDS, TSS, Acidity
TDS, TSS
Acidity, TDS, TSS
Solid Wastes
S-l - Reverb Slag
S-2 - Dust Bleed
3 ton/ton Cu
0.3 ton/ton Cu
Iron Silicates
Copper Oxides, Minor Elements
-------
The third method for controlling sulfur dioxide concentrations at ground
level is production curtailment when adverse weather conditions prevail. This
method has been referred to as "closed loop control" or a Supplementary Control
System (SCS) when it is based on the monitoring of sulfur dioxide concentra-
tions at ground level at various sites in the areas surrounding the smelter and
using this information to control the smelter operating rate. When ground
level concentrations increase as a result of adverse weather conditions, the
smelter operation is curtailed to reduce the emission rate.
(2) Sulfur Emission Source in a Conventional Smelter
With over 30% sulfur, most copper concentrates contain more sulfur than
copper. About 1-2% of the sulfur entering the smelter is lost in the slag
and perhaps 3-4% evolves as fugitive emissions. The remaining sulfur is in
gaseous effluents from roaster, reverb, and converters. Typical sulfur dis-
tributions in conventional smelting are:
Sulfur Distribution (%)
Source Calcine Smelting Green Charge Smelting
Roaster 20
Reverb 25 40
Converter 50 55
Slag and Fugitives 5_ 5_
Total 100 100
As mentioned in the previous section, the conventional smelting process
evolved in geographical areas where acid markets were unavailable and where
all S02~containing gas streams were vented to the atmosphere (after particu-
late control, if necessary). Thus, conventional technology uses gas-handling
techniques (e.g., use of dilution air for cooling of gas streams) which would
not be used if the stream were to be treated for SC>2 recovery. However,
streams from the roaster and converter can be handled to minimize air leakage.
This results in1 S02 concentrations over 4-5% which is adequate for autogenous
sulfuric acid manufacture the most cost-effective control technology for
removing SC>2 from such streams. The reverb gases are a high volume (up to
100,000 scfm) and low concentration (0.5-2% 862) stream, not amenable to
autogenous sulfuric acid manufacture and have to be discharged via tall stacks.
EPA's work during the promulgation of New Stationary Source Standards showed
that S02 control of reverb emissions was not economically feasible, (This
aspect is discussed further in the next section.)
With conventional smelting as well as with the new smelting processes,
S02 control is achieved via an "end-of-pipe" treatment facility, i.e., a
sulfuric acid plant. Changes in processing and in gas handling and gas cooling
are necessary, and no clear delineation is possible between process units and
pollution control units. For this reason, we have integrated the costs of SC^
27
-------
control into the calculation of operating costs for both the base line as well
as the new smelting process options. Furthermore, the base line case assumes
control of converter gases alone (equivalent to about 50% sulfur control) and
no control of reverb gases since this is typical of conventional smelting
operations today. The process options considered later do not incorporate
reverbs and recover more S02 than the base line. This is somewhat unique
among the various industries studied.
Table IV-2 shows how one would calculate the price of sulfuric acid manu-
facture if it was a separate entity receiving the off-gases from converting
furnaces. Converter gases fluctuate in volume and quality because converting
is a batch operation. For smaller smelters, the acid plant has to be over-
sized to accommodate these fluctuations; and the cost of air pollution control
can be more than what is shown in Table IV-2.
The reverb gas contains particulates in addition to S02ซ Technology is
presently available (electrostatic precipitators) which will usually permit
meeting the particulate emission levels established for copper smelters.
Electrolytic refineries have no air pollution problems.
The addition of an acid plant substantially increases the power demand
at a smelter. Without acid plants, most smelters generate more than ade-
quate power for their internal needs from the waste-heat boilers on the
reverberatory furnaces and power generation can be increased by installing
waste-heat boilers to cool the hot converter gases. With acid plants,
smelters become net purchasers of power.
Increasing sulfur capture by control of reverb gases requires scrubbing
or S02~concentration technology. Data published by the EPA* suggest that the
costs of this technology are very high. For example, the production of sul-
furic acid from roaster and converter gases costs about $70-80/ton of sulfur.
The recovery of sulfur from reverb gases via scrubbing or S02~concentration
costs in the range of $200-300/ton of total sulfur removed and $400-5007
incremental ton of sulfur recovered via the scrubbing or concentration
techniques. For this reason, EPA has concluded that control of reverb gases
is not economically achievable.
b. Solid Wastes
Slightly more than three pounds of solid waste per pound of pure copper
are generated in a copper smelter. These come in the form of a slag and
collected flue dust.
(1) Slag
The converter slag is recycled to the reverb in order to recover its'
copper content. The slag tapped from the reverberatory furnace (and granu- ;
lated in some cases) is disposed of as an inert rock. Reverb slag is
mainly an iron silicate, containing about 0.5-0.9% copper and minor elements
in rather dilute form. It is sent to landfill at an estimated cost of
$5/ton of slag, or $15/ton of copper produced.
* Background Information - New Source Performance Standards for Primary Copper,
Zinc, & Lead Smelters, EPA, Office of Air & Water Programs, August 1973.
28
-------
TABLE IV-2
COSTS OF OPERATING AN ACID PLANT: CONVENTIONAL SMELTING
Basis: 52,000 scfm
8% S02 Concentration
800 ton/day of Acid
330 Operating Days/Yr
Capital Investment (CI) :
Waste heat boiler 2.13
Electrostatic precipitator & cold gas cleaning 1,06
Double Absorption (DA) acid plant 12.88
Total 16.07
Expected H2SO& Production Cost: $/Yr $/Yr
Variable Costs
Direct Operating & Maintenance Labor
13 men @ $5.75/hr, 48 hr/week (L) 188,370
Supervision @ 15% of Labor (S) 28,255
Overhead @ 35% of Labor & Supervision 75,818
Total Labor 292,433
Operating & Main. Supplies @ 2.4% of Cap. Cost 385,680
Misc. Supplies 29,200
Total Supplies 414,880
Energy - Fuel oil 4,550 bbl @ $11.60/bbl 52,780
Electricity 180 kWh/ton @ $0.021/kWh 997,920
Total Energy 1,050,700
Other expenses - shutdown 255,J)OQ
Total Variable Costs 2,012,823
Fixed Costs
Plant Overhead @ 65% (L + S) I 140,807
Local Taxes & Insurance @ 2% of CI 321,400
Depreciation 14 years, straight line 1,147,857
Total Fixed Costs ' 1,610,064
Total Variable & Fixed Costs 3,622,887
BDI 20% of CI 3.214,000
Total charge against l^SO^ manufacture 6,836,887
Total Charge/Ton of acid $25.90
29
-------
(2) Flue Dust
The flue dusts result from entrained particles and condensed effluents
in the gas stream. Typically, 3-6% of the total weight of solids entering
the smelter are evolved as dust. Coarse particles are caught in the cooling
chambers, while fine particulates are removed by electrostatic precipitators
operating slightly above the dew point of the gas stream. While the electro-
static precipitator is highly efficient, the fumes from volatile metal
species can still be present in the gas stream. For example, the converter
gases are usually scrubbed after the initial particulate removal in order to
further reduce the concentration of minor elements which could create prob-
lems in the operation of a sulfuric acid plant.
All flue dusts contain entrained copper, and they are mostly recycled
to the reverberatory furnace. They may also contain volatile impurities in
a concentrated form. Excessive impurity build-up by dust recirculation
would impair the quality of the blister copper. At times, it is economical
to process these dusts further in order to recover such metals as zinc,
lead, etc. Depending on the composition of the feed, a fraction of the dust
generated may be diverted and either sold to other specialized smelters which
recover the contained metals or encapsulated for safe disposal. For example,
encapsulation could be accomplished by incorporating the dust into a matrix
such as a pozzolanic cement which is suitable for landfilling. To the best
of our knowledge, some dust has been stockpiled at certain smelters but none
has been added to a pozzolanic cement.
c. Fate of Minor Elements
A schematic diagram of the flows of minor elements in a copper smelter
is shown on Figure IV-2, and a breakdown is shown in Table IV-3. Impurities
are eliminated from the copper-rich stream by slagging and volatilization,
whereas they are diluted in the slag by the inert oxides such as Fe2SiC>4. The
amount of dust recirculated is therefore a key factor in the steady-state flow
of impurities. This is why the data in Table IV-3 show the distribution of
impurities on the basis of the total feed to the reverb rather than the impur-
ities in the ore concentrate only.
In quantifying these streams, the scarce data available from published
literature have been supplemented by our "best engineering judgment." Due to
the immense variety of possible concentrates and operating practices, these
numbers should be considered as "order of magnitude" values only.
The Impurities remaining in the blister copper (other than sulfur and
oxygen) are carried over to the refining where they distribute themselves
among:
the final product;
i
the anode slimes, which are further processed for precious metals
recovery; and
t'he electrolyte, from which nickel may be separated.
30
-------
CONCENTRATE INTAKE
REVERB SLAG
i
REVERB
RECYCLED SLAG
DUST
MATTE
CONVERTER
T
DUST
BLISTER
OR
ANODE COPPER
DISCARDED
DUST
Figure IV-2. Diagram of Impurity Flow Within a Copper Smelter
31
-------
TABLE IV-3
SUMMARY OF THE WEIGHT DISTRIBUTION OF MINOR ELEMENTS IN THE CONVENTIONAL
SMELTING PROCESS (MATTE GRADE = 40% COPPER)
u>
ro
Element
As
Sb
Bi
Hg
Pb
Zn
Se
Te
Total Input .- .
to the Reverb^ '
100
100
' 100
100
100
100
100
100
Granulated
Reverb
Slag
54
54
8
0
10
30
<10
<10
Total Reverb
Dust and
Volatiles
12
15
85
=100
80
60
.<10
3'
25
14
6
0
10
10
>45
>45
Remaining
in
Blister Copper
5
9
1
0
traces
traces
>45
>45
Notes;
(1) Recycled converter slag-., recycled dusts and new ore concentrate are the contributors
to this total input.
(2) Estimated reverb emissions are that fraction of total reverb dust not captured by electro-
static precipitators.
(3) The converter dust is captured in electrostatic precipitators. The residual fines are cap-
tured in the wet scrubbers preceding the acid plant and constitute less than 1% of the
total converter dust.
Source: Arthur D, Little, Inc. estimates
-------
d. Water Pollution
The primary copper industry must control water emissions from three
major sources, the mines, the smelters, and the refineries. In controlling
water pollution it is often necessary to remember that in controlling the air
pollution problems, a water pollution problem can be created since some of the
most effective air pollution technologies are based on the use of water in
scrubbing. Furthermore, the water drainage problem from tailings disposal
areas is of considerable concern to the industry, but much less than in the
coal mining industry. Although air and water pollution control have been
considered separately, it is mandatory that in arriving at solutions to one
problem, another one of equal or greater magnitude is not created.
The water pollution regulatory constraints on the copper industry arise
mainly as a result of Sections 304(b) and 306 of the Federal Water Pollution
Control Act Amendments of 1972. Under this Act, the EPA has conducted techni-
cal studies which are published as "Development Documents" and form the basis
for the Effluent Limitation Guidelines. These guidelines refer to three
specific discharge levels.
Best Practicable Control Technology Currently Available (BPCTCA) -
to be met by industrial discharges by 1977.
Best Available Technology Economically Achievable (BATEA) - to be
met by 1983.
New Source Performance Standards (NSPS) - to be applied to all new
facilities constructed after the promulgation of these guidelines.
In order to achieve the effluent limitations, the recommended treatment
technology for the copper industry must perform three functions:
remove suspended solids,
adjust pH, and
remove the specific heavy metals.
To perform these functions, the following treatment steps are recommended:
lime precipitation,
settling, and
pH adjustment.
Since all of the heavy metals included in the proposed effluent limita-
tions have very low solubility in the alkaline pH range, the addition of lime
causes the metals to precipitate out of solution as hydroxides and carbonates.
The metal precipitates, along with other suspended solids present in the
33
-------
wastewater, are separated from the wastewater stream by means of settling and
are withdrawn as a sludge. Since the wastewater is still at a high pH after
this step, it is necessary to lower the pH by injecting carbon dioxide gas or
other acid into the water. This step is usually performed in a separate basin.
Other recommended techniques for improving the effectiveness of the previously
mentioned end-of-pipe treatment are:
reuse of water in other operations;
control of mine water drainage by modification of mining techniques,
construction of diversion structures, or ditching; and
use of solar evaporation to eliminate the discharge of excess water.
In a copper smelter and refinery the sources of wastewater are:
slag granulation (if this is practiced);
acid plant blowdown (i.e., blowdown from wet scrubbers ahead of
the acid plant);
metal cooling;
spent electrolyte and washings; and
storm water commingling with process wastewaters.
Table IV-4 shows raw waste characteristics.
The Effluent Limitation Guidelines for primary smelters and refineries in
net evaporation areas is zero discharge of process wastewater pollutants,
based on recycle, reuse, and solar evaporation. This standard applies to both
the 1977 (BPCTCA) and 1983 (BATEA) guidelines. (The applicability of this
standard to a particular smelter has been challenged in court.) We have used
the zero discharge standard as applicable in all cases since we have assumed
that the new process options would probably be utilized in the Southwestern
United States, e.g., Arizona, where net evaporation occurs.
Zero discharge is achieved by:
neutralization of acidic streams;
settling (thickening) of streams containing suspended solids;
cooling of contact cooling water for recycle, and
solar evaporation from storage areas.
Furthermore, any blowdown from these systems is bled into the water
recycle system of a nearby mill. The costs of water pollution control are
estimated to be $3.25/ton as shown in Table IV-5.
34
-------
TABLE IV-4
RAW WASTE CHARACTERISTICS OF CERTAIN STREAMS IN COPPER SMELTING
(TYPICAL VALUES)
to
Ln
Parameter
pH
TDS
TSS
Sฐ4=
CN
As
Cd
Cu
Fe
Pb
'Kg
Ni
Se
Te
Zn
Oil and Grease
Slag Granulation
Water
0.091
0.588
0.240
0.069
0.0022
-
0.0017
-
0 . 0015
-
-
-
-
0.0077
-
Acid Plant
Slowdown Anode Castinj
0.024 0.09 - 0.1
2.94 0.212
0.023 0.004
0.431 0.003
0.002
0.0001
-
0.0005
0.0010
_
-
0.0001
-
0.0026
-
Flow/Production 599.20* 176.17**
* 1200 gal/ton
** 35 gal/ton
***15-180 gal/ton
Source: EPA Development Document - EPA 440/1-75/032-b
7.35 - 93.48***
-------
TABLE IV-5
ESTIMATED WATER POLLUTION COSTS - COPPER SMELTING
Basis: 100,000 ton/yr
Estimated Capital Investment: '$500,000
j/yr
Direct Operating Costs
Labor and Supervision 3,000 hr @ $14.75/hr
Electricity - 0.9 x 106 kWh I? $0.021 kWh
Lime - 1,500 tons @ $35/ton
Maintenance (4% of Investment/year)
Sludge Disposal - 9,000 tons @ $5.00/ton
Fixed Operating Costs
Local Taxes and Insurance - 2% of Investment $ 10,000
Depreciation (7%) 35,000
Return on Investment (20%/yr) 100,000
Total Fixed Operating Costs $145,000
Total Operating Costs $325,000
Unit Costs ($/ton Copper) $3.25
3. Economic Factors
a. Smelting
Table IV-6 shows estimates of capital and operating costs for conventional
smelting based on data in the literature, e.g., Schwartz (1975), Foard and
Beck (1971), and our files. The raw material and energy costs are representa-
tive of the Southwest, e.g., Arizona.
It should be noted that the costs in Table IV-6 are for a "typical"
chalcopyrite concentrate in a "typical" location. They have been derived on
a basis consistent with the costs shown later to be "typical" of the new
processes. Thus, these costs are for comparison purposes to evaluate the
advantages of new technology. They should not be interpreted as current costs
of actual smelters in Arizona.
36
-------
TABLE IV-6
OPERATING COSTS: CONVENTIONAL SMELTING
(Basis: Copper sulfide concentrates, 28.6% Cu, 29.3% Fe,
33.4% S; 100,000 ton/yr of anode copper)
CAPITAL INVESTMENT
(CI) $ /annual ton Cu
OPERATING COSTS
VARIABLE COSTS
Silica Flux
Limestone
H2SO, Credit
Fuel Oil
Natural Gas
Electricity
Water : Process
Cooling
Refractories
Direct Operating and
Maintenance Labor (L)
Supervision (S)
Maintenance Materials
Labo.r Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local' Taxes &
Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
Unit
ton
ton,
ton
106 Btu
106 Btu
kWh
103 gal
103 gal
ton
man-hr
L
CI
L+S
L+S
CI
CI
CI
$/Unit
8.00
10.0
10.00
2.00
0.65
0.021
0.20
0.05
300.00
5.75
15% L
4% CI
35% (L+S)
i
65% (L+S)
2% CI
7% CI
20% CI
Greenfeed
Smelting
Units/ $/Ton
Ton Cu
$650
1.14 9.12
0.14 1.40
1.93 (19.30)
22.7 45.40
1.3 0.85
347.00 7.29
1.06 0.21
2.0 0.10
0.01 3.00
6.2 35.65
5.35
26.00
14.35
129.42
26.65
13.00
45.00
130.00
215.15
349.57
Calcine
Smelting
Units/ $/Ton
Ton Cu
$750
1.14 9.12
0.14 0.40
2.45 (24.50)
18.60 37.20
1.3 0.85
441.00 9.26
1.06 0.21
2.0 0.10
0.01 3.00
6.5 37.38
5.61
30.00
15.04
124.67
27.94
15.00
52.50
150.00
245.44
370.11
37
-------
The following should be noted:
The operating costs exclude concentrate costs. This is consistent
with the concept of a smelter's "operating margin" where the spread
between the copper price and concentrate value is relatively con-
stant and independent of the copper price.
The concentrates are assumed to contain no precious metals or
impurities. Many concentrates contain one or the other or both.
The smelter margin is improved by overrecovery of precious metals,
i.e., the difference between metals recovered and metals paid for.
It might also be improved by penalties assessed for impurities in
the concentrates. In consideration of generalized processing
approaches, the usual industry practice is to ignore these factors.
Copper overrecovery is not included. For all of the pyrometallurgi-
cal processes under consideration, it would amount to about $15/
ton copper.
Fuel oil is the fuel selected for the smelting furnace. Any fuel
(pulverized coal or natural gas) can be used instead. Natural gas
is used for "poling" (reducing oxygen in anode copper).
Sulfuric acid has been credited at $10/ton, which is a reasonable
"disposal" price for this acid in the Southwest. As discussed later,
this does not cover the cost of producing this acid.
The net energy requirements involve the use of waste heat recovery
equipment for cooling hot gases. Typically, a reverb will recover
about 1.2 to 1.3 x 10" Btu/ton of charge. Since all recovered
energy is used within the plant, a waste heat credit is not shown.
The costs include the costs of meeting Federal Ambient Air Quality
Standards for SC^ and particulates and the Effluent Limitation
Guidelines for 1983 for water pollution for existing plants, which
are the same as the Guidelines for New Sources.
The total operating costs are about 17-18e/lb with direct costs of
about 5-70/lb depending upon whether acid credit is included. These
compare with custom smelter charges (at existing older facilities)
of approximately 9-llc?/lb for smelting alone.
Table IV-7 shows pollution control costs. Out of the total operating
costs of 17-18e/lb copper shown in Table IV-6, pollution control costs amount
to about 3c/lb.
38
-------
TABLE IV-7
COST OF POLLUTION CONTROLS IN CONVENTIONAL SMELTING
Basis: 100,000 Short Tons of Copper/Yr
Green Feed Smelting Calcine Smelting
Unit $/Ton 1
Acid Manufacture
Credit from Acid Sale
Slag Disposal
Water Pollution Control*
Dust Disposal
Total Smelting , 49.42 57.68
Unit
Ton
Ton
Ton
Ton
$/Ton
25.90
(10.00)
5.00
24.00
Units /Ton
1.93
1.93
3
0.02
$/Ton Cu
49.99
(19.30)
15.00
3.25
0.48
Units /Ton
2.45
2.45
3
0.02
$/Ton Cu
63.46
(24.50)
15.00
3.25
0.48
See Table IV-5
The amount of dust bled and disposed depends on the quality of the feed.
It was taken here as 10% of the 'total dust emissions (assumed to be 5% of the
total charge to the smelter).
Source: Arthur D. Little, Inc. estimates
b, Electrorefining
Table IV-8 shows estimates for electrorefining anode copper to cathode
copper. Most refineries melt and cast cathode copper to other shapes or
products. We have selected cathode copper as a basis in order to provide a
basis for comparison for hydrometallurgical processes which also produce
cathode quality copper. It should be noted that about 15% of anode copper is
returned as anode scrap for remelting and casting.
Electrolytic refineries.have only one major effluent stream spent
electrolyte. The treatment costs for neutralization and disposal of this
stream are estimated to be about $0.50/ton of copper.
39
-------
TABLE IV-8
OPERATING .COSTS: CONVENTIONAL ELECTROREFINING
(Basis: 100,000 ton/yr cathode copper; 15% anode scrap recycle)
CAPITAL INVESTMENT (CI) $/annual ton
OPERATING COSTS
VARIABLE COSTS
Sulfuric Acid
Fuel Oil
Electricity
Water: Process
Direct Labor CL)
Supervision (S)
Maintenance Materials
Labor Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
Unit
ton
106 Btu
kUh
103 gal
Man-hr
L
CI
L+S
L+S
CI
CI
CI
$/Unit
10.00
2.00
0.021
0.20
5.75
15% L
ซ CI
35% (L+S)
65% (L+S)
2% CI
7% CI
20% CI
Units/Ton
$450
0.01
2.00
250.00
1.00
6.00
$/Ton
Cathode Cu
0.10
4.00
5.25
0.20
34.50
5.18
18.00
5.95
73.18
25.79
9.00
31.50
90.00
156.29
229.47
40
-------
C. OUTOKUMPU FLASH SMELTING
1. Flash Smelting - Concept and Operations
Flash smelting combines the separate roasting and smelting operations of
conventional copper extraction into a combined roasting-smelting process. The
heat generated by the exothermic roasting reactions can be used for smelting
so that little or no extraneous fuel is needed. A characteristic of this
method is that fine-grained concentrates are used and the smelting takes place
in suspension, which allows for rapid reaction rates. The major advantages of
the method are a reduction in fuel used for smelting and the production of a
stream of gas high in S02 which is suitable for sulfuric acid manufacture.
In the late 1940's Outokumpu Oy of Finland developed-the flash smelting
process, and has been using this process at its Harjavalta works since 1949'.
Other flash smelting techniques have been developed, e.g., oxygen flash
smelting in Canada, cyclone smelting in Russia. Of these, only the Outokumpu
approach has been adopted widely in other parts of the world.
The processing of copper concentrates by flash smelting at the Harjavalta
works of Outokumpu is shown in Figure IV-3. From storage, the feed materials
(concentrates and silica sand) are fed by automatic weighers in the right
proportions onto a belt which conveys the charge to a drier. In the drier,
which is a direct oil-fired rotary kiln, the concentrate is thoroughly dried.
The finest particles leave the kiln as flue dust but are collected in an
electrostatic precipitator (ESP) and returned to the main concentrate flow.
The charge is transported by means of a pneumatic elevator to the feed hopper.
From the feed hopper, the dried charge is fed by a conveyor into the concen-
trate burner. In the concentrate burner the charge is mixed with preheated
air at about 930ฐF. It reacts in the reaction shaft (the portion of the flash
furnace, beneath the concentrate burner) and the temperature rise is enough to
melt the particles. The molten particles are separated from the gas phase
in the settler (the horizontal portion) thus forming matte and slag. The matte
grade is controlled by the air/concentrate ratio.
From the uptake, the gases are led into a waste heat boiler - a forced
circulation type of water wall radiation boiler. The main function of the radi-
ant section of the boiler is to cool the gas containing molten dust particles
to such a temperature, 1290ฐF-1470ฐF, that they solidify and do not cause dif-
ficulties due to sintering in the heat exchangers. The steam pressure in the
boilers must be high enough to ensur'e that the wall temperature of the boiler
tubes is above the dew point temperatures of the gases which cqntain S02 and
803. The dust in the boiler is easily removed by automatic soot-blowing
equipment.
From the radiation boiler the gases are led at a temperature of 1290ฐF-
1470ฐF to the heat exchanger. The heat exchanger preheats the air needed for
smelting to a sufficiently high temperature by using the available heat content
of the waste gases. The heat exchanger type used by Outokumpu Oy consists of
groups of finned tubes cast from heat-resisting steel. The gases flow outside
and the air inside the tubes. The dust is removed from the elements by
41
-------
K>
AIR
SLAG
1. CONCENTRATE STORAGE
2. BELT CONVEYOR
3. DRYER
4. DRYER COTTRELL
5. REDLERS
6. FEED HOPPER
7. CONCENTRATE BURNER
8. FLASH SMELTING FURNACE
9. WASTE HEAT RADIATION BOILER
10. HEAT EXCHANGER
11. PRIMARY AIR FAN 15. CONVERTER
12. ELECTROSTATIC PRECIPITATOR 16. ANODE FURNACE
13. DAMPER FOR AUTOMATIC 17- CASTING WHEEL
18. SLAG CLEANING FURNACE
DRAFT CONTROL
14. GAS FANS
19. SLAG GRANULATING
SOURCE: Flash Smelting of Copper Concentrates, P. Byrk Et al.
Figure IV-3. Flowsheet of the Harjavalta Outokumpu Smelter
-------
automatic soot-blowing equipment. Typically, the heat exchanger heats the air
to about 930ฐF, and the waste gases are cooled to 660ฐF. The waste gases are
then led to hot cottrells where most of the 'remaining dust is removed.
The flue dust collected from the waste heat boiler and ESP can be trans-
ported by a pneumatic elevator and then returned to the flash smelting furnace
through a separate feed hopper.
From the ESP, the gases move through the exhaust gas fans which transport
them to the sulfuric acid plant. These fans operate with an automatic draft con-
trol system so that the flash smelting furnace operates under a very small draft.
From the settler of the flash smelting furnace, the molten copper matte
is tapped and transported in ladles to converters.
At Harjavalta,' the converter department consists of three 11-1/2 ft x
21-1/2 ft converters. Since the matte grade is high (about 60% copper), only
one converter at a time is necessary for blowing the flash smelter matte. All
the smelter reverts and considerable amounts of scrap are smelted in the con-
verter. Due to the high matte grade and the short blowing time, it is neces-
sary to keep the converter hot between charges. This is accomplished by adding
coke while waiting for the next charge^ except for this, the converter opera-
tion is conventional.
Because of the high grade of matte, the copper content of the flash smelter
slag is quite high. It is necessary to recover copper from this slag. This
can be done in an oil-fired or an electric furnace, or by flotation techniques.
(These are discussed in Section G "Metal Recovery from Slags.")
Blister copper is transferred to an oil-fired rotating anode furnace,
fire-refined and.cast into anodes.
2. Current Status of Flash Smelting
Since the establishment of the first flash smelter by Outokumpu at
Harjavalta in 1949, a number of smelters have been built worldwide or are under
construction. The following is a list of installations in operation or under
construction:
Japan; Furukawa Mining Company, Dowa Mining Company, Nippon Mining
Company, Mitsui Mining Company, Sumitomo Mining Company,
Hubo Kyodo Smelting' Company
Rumania; State Enterprise
.Turkey; Kardeniz Bakir Isletemeteri
India: Indian Copper Corporation, Hindustan Copper Corporation
Australia: Peko-Wallsend Ltd.
U.S.A.; Phelps Dodge Hidalgo
U.S.S.R.; Norilsk Mining Metallurgical Combine
43
-------
3. Effluents from Flash Smelting
Figure 17-4 is a schematic flow diagram of copper smelting using the
Outokumpu flash furnace. The sources of pollutants are shown in four cate-
gories: air, solid, water, and fugitive. Table IV-9 shows the magnitude of
these streams and the major constituents.
a. Air
The concentration of S02 in flash smelter gas is high, containing up to
13% S02- Conventional reverberatory furnace gas, on the other hand, typically
contains 1.5-2.0% SC^. These high-strength gases are most suitable for the
manufacture of sulfuric acid. The variable strength/variable volume SC>2 gas
stream from the converters can be mixed with the steady stream of flash
smelter gas to provide a stream high enough in SC>2 for acid manufacture.
Table IV-10 shows how the SC>2 emissions from the flash furnace and con-
verter change with different matte grades. At high matte grades, a large
amount of sulfur is eliminated in the flash furnace. This improves acid plant
performance since the volume and strength of the input stream are more con-
stant. The overall distribution of sulfur in the smelter is shown in Table
17-11. It can be seen that the total emission to the atmosphere is only 2.8%
of the total sulfur in feed. This is considerably lower than the emissions
from conventional smelting where sulfur recovery from reverbs cannot be prac-
ticed economically. Because the EPA has concluded that reverb gases could
not be treated economically for sulfur removal or sulfur recovery (while the
gases from the flash furnace can be so treated), the comparison of these new
smelting processes with the base line is not for equivalent degrees of sulfur
control. The base line case recovers only 50-70% sulfur, while the new tech-
nology recovers over 90% sulfur.
Table 17-12 shows the cost of manufacturing sulfuric acid' based on the
assumption that the flash smelter gas and the converter gases are combined to
form a steady flow of 56,900 scfm containing 10% S02.
b. Solid Wastes
Both flash smelter slag and converter slag are presently milled and
floated at the Harjavalta smelter. In this case, the same amount of solid
waste is generated as in the base case, but it consists of finely ground
tailings. In calculating pollution control cost, we assume that 'three pounds
of fine wet particles (90% <300 mesh) per pound of copper are pumped by pipe-
line to lined tailing ponds. In order not to tie up capital in a large pond,
a new dike can be built every year in order to create a new settling basin of
475 ft in area by 24 ft in depth. The basins stay water covered until ready
for final covering by three ft of dirt. The corresponding cost was estimated
at $190,000/yr, or $1.90/ton of copper produced. The alternative approach is
to use an electric furnace for slag cleaning. These two techniques are dis-1
cussed separately in Section G.
44
-------
CONCENTRATE
RECOVERED MATTE
32) DUST BLEED
STACK
EFFLUENT TYPES: (A) -AIR; (w) -WATER; (j) -SOLID WASTES; (?) -FUGITIVE
"ESP- ELECTROSTATIC PRECIPITATOR
SLAG (!0
figure IV"4, Sources of Emissions in the Outokumpu Flash Smelting Process
-------
TABLE IV-9
S tream
EMISSIONS FROM OUTOKUMPU SMELTING
Stream Size
Air Pollution
A-l - Acid Plant Tail Gas 55,000 Scfm
*
A-2 - Anode Furnace Gas NA
Major Constituents
S02 - 0.05%
Flue Gas, Some
Water Pollution
W-l - Slag Milling
W-2 - Acid Plant Slowdown
W-3 - Contact Cooling
1200 gal/ton
720 gal/ton
180 gal/ton
TDS, TSS
IDS', TSS, Acidity
TDSf TSS
Solid Wastes
S-l - Cleaned Slag
S-2 - Dust Bleed
3 ton/ton Cu
0.3 ton/ton Cu
Iron Silicates
Copper Oxides, Minor Elements
NA = Not Available
TABLE IV-10
SULFUR DISTRIBUTION IN FLASH FURNACE AND CONVERTER GASES
Sulfur Distribution
Matte Grade (%)
Feed
Furnace Gas
Converter Gas
45
100
55
45
55
100
65
35
65
100
72
28
75
100
76
24
Source: "New Developments in Flash Smelting,"
S.U. Harkki and J.T. Juusela, The Metal-
lurgical Society of AIME, New York, 1974.
TABLE IV-11
SULFUR LOSS IN VARIOUS PROCESSING STEPS IN FLASH SMELTING
Step
Drying
Smelting
Converting
Anode Furnace
Subtotal, Atmospheric
Slag Losses
Total toss
0.7
1.0
1.0
0.1
2.8
1.2
4.0
Source: "Hew Developments la Flash Smelting",
S,U. Barkki end J.T. Juusela, The
Metallurgical Society of AIHE, New
York, 1974.
46
-------
TABLE IV-12
OPERATING COSTS FOR AN ACID PLANT: OUTOKUMPU SMELTING
Outokumpu: 56,900 scfm
10% S02 Concentration
1094 ton/day of Acid
330 Operating Days/Yr
100,000 Tons of Anodes/Yr
Capital Investment (CI):
$10
Waste heat boiler
Electrostatic precipitator & cold gas cleaning
DA acid plant
Total
Expected H2 SO^ Production Cost:
Variable Costs
Direct Operating & Maintenance Labor
14 men @ $5.75/hr, 48 hr/week (L)
Supervision @ 15% of Labor (S)
Overhead @ 35% of Labor & Supervision
of Cap. Cost
Total Labor
Operating & Main. Supplies @ 2.'
Misc. Supplies
Total Supplies
Energy - Fuel oil 5,000 bbl @ $11.60/bbl
Electricity 180 kWh @ $0.021/kWh
Total Energy
Other expenses - shutdown
Total Variable Costs
Fixed Costs
2.26
1.12
13.65
17.03
$/Yr
Plant Overhead @ 65% (L + S)
Local Taxes & Insurace @ 2% of CI
Depreciation 14 years, straight line
Total Fixed Costs
Total Variable & Fixed Costs
ROI 20% of CI
Total charge against H^SO^ Manufacture
Total Charge/Ton of Acid
202,860
30,429
81,651
408,720
32,000
58,000
1,364,655
$/Yr
314,940
440,720
1,422,658
280.000
1,702,655
151,638
340,600
1,216,428
1,708,666
3,411,321
3.406.OOP
6,817,321
$18.88
47
-------
Flue dusts are generated in amounts comparable to those mentioned for
the base line (3 to 6% of the total weight of solids entering the smelter).
Dusts are normally recirculated to the smelter, but again part of the dust
stream may be diverted to avoid impurity build-up. The cost of disposing of
dusts is the same as for the baseline: $2A/ton of dust.
A schematic illustration of the flows of minor elements in a copper smelter
using flash smelting reactors is shown in Figure IV-5. When making matte con-
taining 60% copper, Table IV-13 gives a semi-quantitative idea of the distri-
bution of the impurities among various streams.
c. Water
The water effluent streams from Outokumpu smelting are associated with the
acid plant, the anode cooling, and the slag flotation (or granulation).
The aqueous effluents from the acid plant are similar in nature to those
of the conventional smelter, but their volume is somewhat larger. The same
control technology (lime and settle) can be applied at a cost of about $1.78/
ton of copper produced. These costs are part of the costs shown in Table IV-16
for the acid plant.
The anode cooling stage is identical to the one implemented in conventional
smelters.
CONCENTRATE INTAKE
RECYCLE
DUST BLEED
TAILINGS
ANODE
See Table IV-13 for explanation of ( ).
Figure IV-5. Diagram of Impurity Flow within a Smelter
Using Outokumpu Flash Furnaces
48
-------
TABLE IV-13
STREAM #
ESTIMATED DISTRIBUTION OF MINOR ELEMENTS AMONG THE VARIOUS
STREAMS IDENTIFIED IN FIGURE IV-5
(Basis: Flash Furnace Impurities Input = 100)
Flash Furnace
Converter
Slag
(2)
Matte
(6)
Dust
(8)
Slag Blister Dust
(3) (7) (9)
Total
Slag
Tailings
to
Pond Total
(4) Dust
Bi
As
Sb
Pb
Zn
2
10
30
10
30
18
20
50
10
10
80
70
20
80
60
0.2
2
11
2
3
0.7
3
13
Traces
Traces
17
15
75
9
7
2.2
12
41
12
33
Traces
4
29
10
29
97
85
45
89
67
Note: The solids discharged to the environment include the tailings and a fraction
of the dust.
The water from the flotation tailings pond is returned to the flotation
circuit. Some fresh water has to be added in order to compensate for the
losses by evaporation, etc.
4. Technical Considerations
(1) Impurities
U.S. smelters treat a variety of materials which are characterized by
their copper, iron, sulfur, and impurity content. The copper-bearing feed
to smelters includes: 1) concentrates, 2) precipitates, 3) scrap, 4) ashes,
5) lead plant byproducts. When the feed to a smelter contains certain impuri-
ties such as arsenic, antimony, bismuth, lead, etc., in significant quantities,
the feed is characterized as "dirty." In the United States, Asarco's El Paso,
Texas and Tacoma, Washington smelters and the Anaconda smelter in Montana
handle dirty concentrates (in levels above those routinely handled in Outokumpu
flash smelting units in other countries).
To handle these impurities successfully, not only must they be removed
from the product copper but they must also be split into two streams, an
arsenic-rich stream and a lead/antimony-rich stream, so that these streams
can be treated separately for byproduct recovery. The arsenic-containing
stream is usually roasted in the copper smelter to produce arsenic trioxide
49
-------
and recover the copper in the dust while the latter stream is usually sent
to a lead smelter for the recovery of lead, zinc, bismuth and antimony. Any
arsenic in this latter stream is not recovered at the lead smelter but is
returned to the copper smelter as a copper arsenide or "speiss" for further
recovery of arsenic via roasting. To the best of our knowledge, the recovery
of impurities from a mixed stream has not been practiced industrially.
Our discussions with Outokumpu indicate:
High-impurity charges (6-8% As, 1-2% Sb) have not been treated in
the flash furnaces. If such concentrates were charged, the flue
dust from such an operation would be a mixture of Pb, As, and Sb
compounds which could not be treated conventionally and which could
present environmental difficulties.
Outokumpu recommends roasting of the high arsenic materials in
multiple hearth roasters for arsenic removal and the charging of
calcines to the flash furnace. While this had been done success-
fully in a large-scale test, it has not been practiced routinely
at any smelter.
(2) Scrap and Non-Sulfide Copper-Containing Materials
For the smelting of clean concentrates, the Outokumpu flash smelting
process is capable of producing quality blister (over 99% copper). Flash
smelting results in the production of high-grade matte. The capacity for
smelting non-sulfide materials such as precipitate copper and scrap in the
converter is reduced, since heat release in converting is less. However, with
modifications in> operating practice, such as oxygen enrichment of converter
air and coke additions to the converter, scrap melting rates equivalent to
conventional smelting are possible.
(3) One-Step Processing
The flash furnace also can be operated to produce copper (sul-fur-
saturated or oxygen-saturated) in one step by controlling the rate of oxida-
tion in the furnace (Harkki and Juusela 1974). This approach is applicable
only to clean concentrates; otherwise, impurities such as bismuth are concen-
trated in the metallic copper.
(4) Elemental Sulfur from Flash Furnace Off-Gas
As noted, the flash smelting furnace gas is high in sulfur dioxide and
low in oxygen. The S02 can be reduced to elemental sulfur with pulverized
coal and liquid or gaseous hydrocarbons at high temperature in the vertical
uptake shaft, where the main part of the reduction is carried out. Since
elemental sulfur is cheaper and safer to store and to transport than sulfuric
acid, this process might be applicable when the smelter is distant from
sulfuric acid markets.
50
-------
The main problem in S02 reduction has been attaining a sufficiently
rapid reaction rate to reach complete conversion of sulfur dioxide at the
short retention times (2 to 4 seconds) required when handling large gas
volumes. The reaction rate can be increased either by catalytic action or
by raising the gas temperature. The latter approach is used in the Outokumpu
process. Normally the temperature of the exit gas is sufficiently high, but
it can also be raised by burning oil at the uptake of the settler. A thorough
mixing of the reductant and process gas is necessary, and the choice of
reductant also affects the reaction temperature.
The elemental sulfur process requires a flash furnace of slightly dif-
ferent shape and geometry. The gas-handling system comprises the following
stages: A waste heat boiler cools the gases to 660ฐF. Particulates are
removed in an electrostatic precipitator. The gases are then subjected to
two stages of catalytic conversion to increase the yield of sulfur. The gases
are reheated to 790ฐF before entering the first catalytic stage, where CO,
COS, CS2, and H2, if any, are reacted. The gases are then cooled to 340ฐF
in a boiler to reduce the temperature and condense elemental sulfur. The
gases are then reheated to 465ฐF for the conversion of S02 and E^S in the
cold catalyzer. Afterwards, the sulfur is recovered in a scrubber and a
demister. The recovery of purified sulfur from the flash furnace gases is
about 90%.
By producing sulfuric acid from the converter gases and elemental sulfur
from the flash smelter gases, the total recovery of sulfur in the copper
smelter exceeds
This elemental sulfur process is being used only in two locations on
copper concentrates: In Botswana, Africa on copper-nickel concentrates using
coal as the reductant, and at Phelps Dodge's Hidalgo smelter (under construc-
tion) using light naphtha as a reductant. (Outokumpurs Pori plant operates on
a feed of pyrites using a naphtha reductant.)
5. Economic Factors
a. Capital and Operating Costs
Table IV-14 shows estimates of capital and operating costs (including
pollution control) based on data presented by Schwartz (1975) and by Outokumpu.
The capital costs are about the same as for roast-reverb smelting. The fol-
lowing basis has been used: ,
Costs include pollution control and credit for acid sales.
. Oxygen enrichment is not utilized. This case was selected to serve
as a basis for the subsequent evaluation of oxygen enrichment in
Section F. Oxygen enrichment is preferred in many locations with
existing flash smelters, since the capacity of a unit can be increased
up to 50% with oxygen with a consequent decrease in fuel costs and
fixed charges per unit of output.
The calculations procedure is consistent with the calculation for
the base case.
51
-------
TABLE IV-14
OPERATING COSTS: OUTOKUMPU FLASH SMELTING
(Basis: Copper sulfide concentrates, 28.6% Cu, 29.3% Fe, 33.4% S
100,000 S ton/yr of anode copper)
CAPITAL INVESTMENT (CI) $ /annual ton
OPERATING COSTS
VARIABLE COSTS
Silica Flux
Limestone
H2SO, credit
Oxygen
Fuel Oil
Natural Gas
Electricity
Water: Process
Cooling
Refractories
Direct Operating and
Maintenance Labor (L)
Supervision (S)
Maintenance Materials
Labor Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
Unit
ton
ton
Con
ton
106 Btu
106 Btu
kWh
103 gal
103 gal
ton
Man-hr
L
CI
L+S
L+S
CI
CI
CI
$/Unit
8.00
10.00
10.00
15.00
2.00
0.65
0.021
0.02
0.05
300.00
5.75
15% L
4% CI
35% (L+S)
652 (L+S)
2% CI
7% CI
20% CI
Units/Ton
$750
0.86
3.33
12.7
1.3
366.40
1.06
2.00
0.014
6.2
$/Ton CU
6.38
(33.30)
25.40
0.85
7.69
0.21
0.10
4.20
35.65
5.35
30.00
14.35
97.38
26.65
15.00
52.50
150 . 00
244.15
' 341.53
52
-------
b. Comparison with the Base Line
(1) Energy Use
Table IV-15 shows that energy use is decreased compared to a roast-reverb
smelter. Fuel oil consumption decreases from 18.6 to 12.7 10^ Btu/ton copper,
electricity consumption decreases from 440 to 370 kWh/ton copper. This is
because the heat of oxidation of sulfur and iron is used to melt the charge and
energy is recovered from all high temperature waste heat streams. Since both
technologies can use any fossil fuel in the smelting unit, there is no further
conservation of energy of a different form.
The energy data discussed above are those estimated by us on a basis com-
parable to the base case. Outokumpu has published energy consumption data
based on their operating experience at the Harjavalta smelter (Juusela 1974).
Because the Harjavalta smelter uses oxygen enrichment, the results are not
directly comparable. Also the power requirements for the acid plant are not
included. We present Outokumpu's data without any alterations for interested
readers in Table IV-15.
(2) Pollution Aspects
The flash smelter and its associated acid plant recover a higher amount
of S02 (95%) compared to roast-reverb smelting (70%). Table IV-16 shows the
pollution control costs for the process. Comparison of this table with Table
IV-7 shows that because the Outokumpu process produces a much more concentrated
S02 stream,, the pollution control costs per ton of copper are comparable in
spite of the higher degree of sulfur recovery.
For all smelting processes> the transfer of matte to the converter and
converter operations are a major source of fugitive emissions.
(3) Byproducts
When acid markets exist near a smelter, the conversion of the S02-containing
streams to sulfuric acid is the most cost-effective way of preventing their
emission to the atmosphere. However, a market for acid is not always assured,
particularly because of the wide swings in price of elemental sulfur with time
and the availability (in the future) of increasing quantities of byproduct sul-
fur from crude oil desulfurization. Alternative ways of disposal are neutral-
ization with limestone which may cost l-2c/lb copper or sulfur and the conver-
sion of S02 to elemental sulfur (for storage, if markets are unavailable), which
may cost more than 4/lb copper or sulfur.
(4) Cost Comparison
Comparison of Tables ;IV-6 and IV-14 indicates that the cost of flash
smelting is slightly lower, primarily because it is more energy efficient. The
energy costs for roast-reverb smelting are about $44/ton of copper vs $28 for
flash smelting. The total operating costs for both examples are about 17ฃ/lb.
Even if acid credits are ignored, the process is still somewhat cheaper in new
facilities based on clean concentrates.
53
-------
TABLE IV-15
ENERGY BALANCE FOR OUTOKUMPU FLASH SMELTING
BASIS; 1000 Metric ton/day Wet Concentrate (25% Cu, 8% H^O)"
Transformed to Anode Copper.
Requirements by Energy Form/Ton of Charge
Process Step
Drying
Flash Smelting
Slag Flotation
Converting
Anode Casting
Superheating of Steam (70 atm)
Sub-Total
Energy Equivalents
Heat Recovered As Steam
Less Process Steam Used
Less Steam Used For Oxygen
Plant
Surplus Steam
NET ENERGY REQUIRED (kg oil)
(106 Btu)
If Acid Plant is Included
NET ENERGY REQUIRED (kg oil)
(106 Btu)
Steam
(kg)
110
110 --
900
110 <-''
23U
560
ฐ2 Electricity Oil
(NM3) (-kWh) (kg)
4 8
100 15 15
50
10 15
2 6
9
ll(j 86 38
I |
1
1
1
1(2)
I
i
f A j ป
26
1
58
2.2
NOTES;
1. Excludes acid production. For single contact plants, power
requirements ara equivalent to about 125 kWh/ton of feed.
2, Oxygen to steam equivalence based on actual operating data.
3. Electricity to oil equivalence based on standard power plant
efficiency of 10;"500 Btu/kWh.
4. Steam to oil equivalence based on standard boiler efficiency.
5. Original table in metric units; not converted to English units.
Source: Outokumpu, private communications
54
-------
TABLE IV-16
COST OF POLLUTION CONTROL IN THE OUTQKUMPU FLASH SMELTING PROCESS
Units $/Unit Units/Ton $/Ton of Copper
Acid Manufacture Ton Acid 18.88 3.33 62.87
Credit for Acid Sales Ton Acid (10.00) 3.33 (33.30)
Slag Disposal (Granulation) Ton Slag 5.00 3 15.00
Water Pollution Control Ton - - 3.25
Dust Disposal Ton Dust 24.00 0.02 0.48
r
Total Pollution Cost 48.30
Source: Arthur D. Little, Inc. estimates.
D. THE-NORANDA PROCESS
1. Concept and Operation
a. Introduction
The Noranda process combines in a single reactor the three operations of
roasting, smelting, and converting of copper concentrates. The heat losses
suffered during the transfer of concentrate from the roaster to the reverbera-
tory furnace are suppressed, as well as the heat losses occurring during the
transfer of the matte from the reverberatory furnace to the converter. In
addition, the net heat of oxidation is used for smelting. Liquid copper, matte,
and slag coexist in three layers resting on top of each other. Air is blown
into the bath through tuyeres, and oxygen enrichment to 35-40% tends to make
the process autogenous, so that extraneous fuel can be kept at a minimum during
the blow. A continuous flow of high strength (over 10% 802) sulfur dioxide gas
is generated, which is suitable for the manufacture of sulfuric acid.
b. Description of the Reactor and Plant Layout
The following description and figures are largely abstracted from various
papers published by Noranda's staff members. Figure IV-6 shows a schematic
drawing of the Noranda process reactor. A flowsheet of the process appears
as Figure IV-7.
The reactor is a horizontal, cylindrical furnace with a series of tuyeres
located along the reactor in the smelting and converting zone. The reactor can
be rotated to bring the tuyeres out of the bath when a stoppage in the process
is needed for maintenance or because of loss of converting air pressure.
55
-------
SO2 OFF-GAS
CONCENTRATE AND FLUX
FEEDER
BURNER
iiiiiiiniiiniiiiiiiiiiHiiimiiiiimiiimif
7
AIR TUYERES
BURNER
-SLAG
COPPER
SOURCE: Weddick, 1974.
Figure IV-6. Schematic of the Noranda Process Reactor
GASES
TO STACK
H2SO< PLANT
33.3 TONS CONCENTRATE IORY)
22.5% Cu. 33.3% S, 30.VX, FE
30,000 LB STEAM
WASTE HEAT BOILER
AIR
FILTRATION
25.000 sclm
200 scfm NAT SAS_
1800 scfrn AIR
1.7 TON FLUE OUST
FLUE CAS (MaiTFt
78,300 icfm
(5.5% SO2. 7.5% 02, 9,0% HjO
SLAG
32.3 TON
SLAG
MILLING
CIRCUIT
SLAG
*- CONCENTRATE
7.5 TON
4 HITIfllfl
1 1 1 1 M i M i
I 7,5 TON COPPER
[98% Cu)
mi J|
CONVERTING AIR
32,600 Mfm
7.6 TON SLAG
CONCENTRATE
6.J TON FLUX
SLAG TAILINGS
24.8 TON
Figure IV-7-
Material Flowsheet (per hour) for Noranda
Process Plant (800 ton concentrate/day)
56
-------
An opening in the roof of the reactor, away from the tuyere zone, allows
off-gases to escape into a hood, spray chamber, or waste heat boiler and elec-
trostatic precipitator. The cleaned gases are then converted to sulfuric acid.
Other design features are burners at each end of the furnace, a feed port for
charging copper concentrate, flux, and slag concentrate by means of a belt
slinger, and separate tapholes for tapping slag and copper.
The reactor bath consists of layers of slag, high-grade matte, and metallic
copper. Copper concentrate, silica flux and concentrate recovered from slag
milling are charged on the surface of the bath in the smelting and converting
zone, which is strongly agitated by air or oxygen-enriched air injected through
the tuyeres.
Under the dynamic conditions existing in the bath, copper is produced even
though the bath contains more iron than an equilibrium system. The copper
produced by oxidation settles by gravity to the bottom of the vessel, and the
flux combines with oxidized iron to form a slag which floats on a large volume
of matte. This volume is normally kept constant by blowing air or oxygen-
enriched air through the tuyeres at a rate proportional to the rate of copper
concentrate addition so that all the input sulfur and iron is oxidized to pro-
duce copper and slag. At the same time, the rate of flux addition is controlled
proportionally to the concentrate input rate to maintain an Fe/Si02 ratio in
the slag of 1.6. The slag concentrate addition is also kept at a constant ratio
to the copper concentrate. Copper and slag are tapped at regular intervals so
as to maintain relatively constant levels of matte, slag and copper. These are
the basic control parameters of the Noranda process.
By introducing the air at a sufficient depth below the surface of the
matte, 95% or more of the oxygen reacts with the matte. This consistently high
utilization of oxygen makes it possible to predict accurately the amount of
oxygen required for each ton of concentrate of a particular composition. The
intense mixing action of the air jets maintains the bath in turmoil and thus
provides a high heat transfer rate from the copper sulfide matte, where heat
is generated by the converting reactions, to the slag phase and to the concen-
trate charge on the surface of the bath.
The turbulence generated by air jets, and the fact that the slag is in
contact with a high grade matte, causes the slag to contain 10-12% copper,
most of which is present as entrained metal globules. The tapped slag is
allowed to cool slowly over a period of days, crushed to 90% -324 mesh and sub-
jected to flotation. A tailing containing 0.5% copper is obtained. The copper
loss in the tailings is equivalent to the copper loss with a 0.35% copper slag
produced in a reverberatory furnace. A slag concentrate containing 50% copper
is also produced. It is mixed with flue dust in a pelletizing machine and
recycled to the reactor.
The raw copper produced by the Noranda process contains 1.5 to 2% sulfur,
which is somewhat higher than that contained in blister copper from the con-
ventional smelting process. This copper is transferred to a converter until a
charge of about 120 tons is collected. The sulfur is then oxidized by blowing
air through the tuyeres for 10-15 minutes, followed by fire refining in a rotary
anode furnace, and then casting into anodes for electrolytic refining.
57
-------
When the concentrate feed contains impurities such as bismuth, it is pre-
ferable to produce a high grade matte in the Noranda reactor and convert the
matte separately. Bismuth is volatalized and collected in the converter dust.
Under these conditions, the copper content of the slag is lowered to 6-7%, but
the reactor slag is still amenable to the same flotation treatment. The reactor
matte is charged to conventional Pierce-Smith converters and is further proc-
essed as in the conventional process.
2. Current Status
This process was developed by Noranda Mines Ltd. during the sixties. A
semi-industrial scale pilot plant with a rated capacity of 100 tons of con-
centrate per day has been in operation since May 1968, at the Noranda, Quebec
smelter. Based on the results of the experimental installation, an 800 ton/day
industrial plant was built in Noranda and went on stream in March 1973, using
air for converting. Oxygen enrichment has since improved, the productivity of
this unit. It was shown that increasing the oxygen content of the air blown
through the tuyeres 'from 21% to 35% more than doubled the instantaneous pro-^
ductivity of the reactor.
The reactor presently operated at Noranda, Quebec is now making high grade
matte containing 70-72% copper. This practice has the following advantages:
The refractories around the tuyeres last longer. The longest
campaigns of copper-producing operations were of the order of 72
days, whereas the campaign of matte making was in its 105th day on
October 29, 1975. It was expected to last at least another two weeks.
The level of impurities in the final blister copper is much more
favorable. This has far reaching implications concerning the energy
consumption during refining and the quality of the cathodes
themselves.
A greater flexibility of the integrated operation may be achieved,
as mattes of chosen grades can be made.
Kennecott Copper Corporation has planned the construction of a plant with
a rated capacity of 250,000 ton/yr of blister copper. This plan includes three
reactors of the same dimensions as the present Noranda unit (800 ton/day when
making copper with nonenriched air).
Enough experience has been accumulated so that one may assess the main
features of the process. A comparison of the highlights of the Noranda con-
tinuous smelting process as published by Noranda and conventional smelting
appears in Table IV-17. Table IV-18 shows a material balance. These data
have been used by us in our estimation of process costs.
3. Effluents
Figure IV-8 is a schematic flow diagram of copper smelting using Noranda
reactors. The sources of pollutants are identified for four categories: air,
solid, water, and fugitive. These streams are essentially similar to those
described earlier. Table IV-19 shows the magnitude of these streams and the
major constituents.
58
-------
TABLE IV-17
COMPARISON OF THE NORANDA PROCESS VS.
CONVENTIONAL REVERBERATORY-CONVERTOR WORKS
.(1)
Process Type
Capacity, metric tons concentrate/day
Converters
3
Internal furnace volume/m
Matte grade
Sulfur content of blister'
Smelter slag, % Cu
Smelter slag milled
Power for milling slag, kWh/ron slag
Total copper in slag 2 Cu
Oxygen consumed, ton/ton cone.
Total fuel consumption, 10 Btu/ron
Smelter gas, continuous 7, SO,, cone.
Reverb,
Batch
Convert-
ing
830
2
1485
20-40
0.01
0.3-0.6
no
0
0.3-0.6
0
24
1-2
Noranda
(making matte)
Continuous Matte
Making
Batch converting
900-1800
1
68-72
0.01
0.5-0.6
yes
26
0.3-0.6
0-0.3
n.a".
10-12
Noranda
(making copper)
Continuous Reactor
Copper Making
720-1450
1
1-200
no matte produced
0.01
10-12
yes
26
0-0.3
0-0.3
min 12
10-12
NOTE: The Noranda reactor copper contains 1.5-2% S. It has to be finally converted in a.
few minutes to 0.01% S, This operation can take place either in a converter
or during an. extended-fine refining operation.
Source: Noranda Mines Ltd.
59
-------
TABLE IV-18
TYPICAL COMMERCIAL OPERATION WITH AIR
(Tons)
Tuyere blowing rate, scfm
Fuel type
27,667
oil
% oxygen in tuyere air
% oxygen in combustion air
oxygen,tons/day
21.0
21.0
0.0
MATERIALS BALANCE
Throughput
dry tons/
day
Composition %
Cu
Fe
SiO,
Zn
H2O
REACTOR INPUT
Copper Concentrate
Slag Concentrate
Flux
Oust
800.0
1B2.3
169.7
32.0
24.9
50.0
0.0
20.0
29.0
17.4
4.4
10.0
32.2
5.6
0.9
12.0
4.1
10.9
68.7
0.0
4.9
1.5
0.0
19.3
10.0
12.0
7.0
0.0
REACTOR OUTPUT
Copper
Slag
Dust
200.8
784.7
32.0
97.8
12.0
20.0
0.2
34.5
10.0
2.0
1.3
12.0
0.0
21.6
0.0
0:0
4.9
19.3
0.0
0.0
0.0
MILL INPUT
Slag
784.7
12.0
34.5
1.3
21.6
4.9
0.0
MILL OUTPUT
Slag Concentrate
Slag Tail
182.3
602.4
50.0
0.5
17.4
39.7
5.6
0.0
10.9
24.8
1.5
5.9
12.0
12.0
Copper Loss, % of copper in new metal bearing material: 1.51
HEAT BALANCE
106 Btu/
ton cone.
Distribution
HEAT INPUT
Converting Reactions
Net Heat From Fuel
Total
3.56
1.69
5.25
67.8
32.2
100.0
HEAT OUTPUT
Converting Gas (excluding
combustion gas)
Heat Content of Copper
Heat Content of Slag
Heat Loss
Total
3.35
0.16
1.20
0.54
5.25
63.8
3.0
22.9
10.3
100.0
Total fuel required, million BTU/ton copper concentrate: 4.8
Useful steam from waste heat boiler: 101,420 pounds at 660ฐF and 600 PSIA
OFF GAS DATA
Off gas continuously available to acid plant
(including 75% dilution). Dry scfm
94,631
Moisture content (% wet basis)
6.6
Composition (% dry basis)
N2
81.5
02
9.8
S02
4.5
CO2
4.2
Source: Noxanda Mines, Ltd.
60
-------
-o
EFFLUENT [VPE5 (T) -AIR (^) .WATER (7) -50LIDWASTES (T) FUGITIVE
Figure IV-8. Sources of Emissions in the Noranda Process
Stream
TABLE IV-19
EMISSIONS FROM NORANDA SMELTING
Stream Size Major Constituents
Air Pollution
A-l Acid Plant Tail Gas
A-2 Anode Furnace Gas
Water Pollution
W-l Slag Milling
W-2 Acid Plant Slowdown
W-3 Contact Cooling
Solid Wastes
S-l - Cleaned Slag
S-2 Dust Bleed
55,000 scfm
*
NA
<1200 gal/ton
720 gal/ton
180 gal/ton
3 ton/ton copper
0.3 ton/ton copper
S02 0.05%
Flue Gas, Some SO-
TDS, TSS
TDS, TSS, Acidity
TDS, TSS
Iron Silicates
Copper Oxides, Minor Elements
NA = Hoc Available
61
-------
a. Air
Off-gases leave the reactor through a water-cooled hood. The coarse dust
particles are collected in the cooling chamber connected to the hood, while
fine particulates are collected in an electrostatic precipitator. Part of the
dust is treated for minor elements elimination, while most of it is sent back
to the pelletizer and recycled to the furnace.
The sulfur dioxide concentration in the reactor atmosphere is around 23%
on a dry basis. Because of air infiltration around the hood, the gas stream
entering the acid plant contains 10-13% SC^. This stream is only interrupted
5% of the time during tapping, and can be mixed with the off-gases of other
reactors and Pierce-Smith converters. Air in-leakage (thereby preventing
atmospheric emissions) is high but does not affect the subsequent H2S04 plant
since its operation is at or below the 10-13% SC>2 concentration. This steady,
high S02 gas generation level is a significant advantage over the conventional
reverberatory process.
After dry gas cleaning, wet gas cleaning equipment is required to scrub
out the remaining fine particulates. The gas can then be treated in a double
contact acid plant. The total sulfur recovery is as high as the one achieved
with the Outokumpu flash smelting process, and Tables IV-10 and IV-11 would
apply to the Noranda process.
The fugitive emissions occur during matte transfer and converter opera-
tions and consist of SC>2 and particulate emissions.
b. Solid Wastes
Both converter slag and reactor slag need bensficiation, and this is
presently done by milling and flotation. For each pound of copper, four pounds
of slag containing an average 8% copper are treated; the products are three
pounds of tailings (90% <300 mesh) and one pound of slag concentrate returned
to the smelter's feed. These tailings require lined tailing ponds. As noted
earlier, a new dike can be built every year in order to create a new basin of
475 sq ft in area by 24 ft of depth to accept the calculated volume of discharge.
These basins stay water-logged until ready for final covering by three feet of
dirt. Supernatant water is recovered and reused in the milling circuit. The
corresponding cost for solid waste disposal was estimated at $190,000/yr, or
$1.90/ton of copper produced.
Flue dusts are generated in amounts comparable to those mentioned for the
base line (2 to 5% of the total weight of solids entering the smelter according
to Noranda). Dusts are recirculated to the smelter, unless part of it must be
diverted to avoid impurity build-up or to recover some of these impurities
for commercial purposes. The cost of disposal of impure dust is estimated at
$24/ton of dust.
The flow of minor elements in a copper smelter using the Noranda reactor
is similar to that shown in Figure IV-5. But, when making matte containing
70% copper by the Noranda process, the distribution of the impurities among the
various streams shown in Figure IV-5 is as shown in Table IV-20.
62
-------
TABLE IV-20
NORANDA PROCESS - DISTRIBUTION OF MINOR ELEMENTS WHEN MAKING 70% GRADE MATTE
Basis: 100 units of each Element entering the reactor
Reactor Reactor Volatilized Staying Residual
Input Tapped Slag Tailings in Reactor in Matte in Blister
Stream H (1) (2)_ (4) (6) (9) (5)
Pb
Zn
As
Sb
Bi
Remarks :
100
100
100
100
100
(1) The
13
68
7
28
21
environment
10
60
4
19
16
receives
74
27
85
57
70
the tailings
13
6
9
15
9
(Stream tii)
Traces
Traces
0.8
4.5
2.3
and the
portion of the dusts that is not recycled, i.e., usually
10-20X of -the total dust discharge.
(2) The impurities remaining in the matte are largely volatilized
in the converter and collected as converter dust and in acid
plant blowdown.
Source: Noranda Mines, Ltd., and Arthur D. Little, Inc. estimates
A detailed discussion of the impurity distribution in the Noranda process
appears in the section "technological factors" because impurities have major
impact on the quality and mechanical properties of the product copper.
c. Water Pollution
The water streams in Noranda smelters are those associated with the acid
plant scrubbers, anode cooling, and the slag milling and flotation circuit.
The combined off-gases go through a wet cleaning stage before entering the
acid plant. The nature of the aqueous effluents generated is the same as for
the base case or for Outokumpu smelting. Costs and dimensions given for the
Outokumpu flash smelting process ar(e identical to those for the Noranda process.
Anode cooling waters are also identical to those generated in conventional
smelters,
The water from the slag milling circuit flows with the tailings to the
lined settling ponds. The water is recycled to the milling circuit. The
flotation agents are mainly absorbed on the concentrates and subsequently
pyrolyzed in the smelting furnace.
63
-------
4. Technical Considerations
a. Impurities
A number of minor elements are usually present in the feed to the smelter.
Their overall elimination and possibility of recovery are governed by thermo-
dynamic, kinetic and process parameters. Depending on the prevailing oxygen
and sulfur potential, they may be present as sulfides, oxides, or dissolved
metals.
During the smelting process, elimination takes place by two mechanisms,
slagging and volatilization. Those impurities not eliminated by these mechan-
isms remain in the anodes and affect the operation of the electrolytic refinery.
Elements such as lead, bismuth, arsenic, and antimony are difficult to remove
by oxidation and separation into an oxide slag phase. However, they can be
volatilized during smelting. Lead, bismuth, and tin have high vapor pressures
both as sulfides and as oxides, but are best removed under reducing conditions.
Arsenic, tin and antimony may be removed only as oxides or sub-oxides. Arsenic,
antimony, and bismuth are more soluble in copper than in matte. Thus, their
presence in significant amounts means that the Noranda reactor has to be operated
for making matte.
When making copper, the Noranda process combines a high oxygen potential
and the presence of metallic copper, causing elements such as Pb, Bi and Sb
to stay with the copper. These elements are not considered desirable in anodes
used for conventional electrorefining because in some cases they decrease the
hydrogen overvoltage and the cathodes evolve hydrogen instead of depositing
copper, with a corresponding increase in energy consumption. In other cases,
these impurities may transfer to the cathode, with detrimental effects on the
electrical and mechanical properties of the copper produced. Since conven-
tional electrorefining does not remove these materials very effectively, this
is an open area for research. For example, the synergistic effect of the com-
bined presence of bismuth, arsenic, and antimony in the anodes is not well
explored. These impurities might be rejected more effectively as anode mud
under some circumstances. Selenium and tellurium belong to the same chemical
group as sulfur and they tend to stay in the matte phase. Their transfer to
the copper phase is, therefore, lower than in conventional blister copper since
the matte is never tapped in making copper in the one-step process.
Cobalt, tin, nickel, zinc, and lead are easily oxidized in the slag. The
solubility of silver and copper in the slag is also favored by the high oxygen
potential that prevails when making copper. The reactor slag is presently
treated by milling and flotation at the Noranda smelter. Table IV-21 shows the
recovery of minor elements in slag concentrate by this process. Should it be
economically desirable to recover such valuable elements as lead, zinc, nickel,
cobalt, or other easily oxidizable elements, a pyrometallurgical slag treatment
could be adopted. Typically an electric arc furnace provides a suitable reduc-
ing environment and pyrite additions can provide the stoichiometric amount of
sulfur to recover dissolved metallics as sulfides from the slag to a matte
phase.
64
-------
TABLE IV-21
RECOVERY OF MINOR ELEMENTS IN SLAG MILLING
Elsซnt
Cu
S
KG
Pti
Zn
Cd
Sb
Bi
Si
Ic
Xi
Sn
1
Au
AR
IP. 51,S ro,>,cntr,rc
07
04
11
11,
11
2?
29
23
73
58
11
17
r,
91
Source: Hackey 1975
Lead and zinc are the largest contributors to the overall dust generation.
The dusts also contain significant amounts of the other elements, as noted
earlier. Bismuth, selenium, and tellurium volatilization is mainly achieved
by the flushing action of the tuyere gas. Arsenic and antimony would appear
to be volatilized during the early stage of concentrate smelting. Table IV-22
shows the observed distribution of the minor elements among the off-gases, the
slag, and the reactor copper. One way to avoid building up the amount of
impurities in the system is to treat part of the dust before recycling it to
the reactor. The amount of dust treated depends on the amount of impurities
fed with the reactor charge and also on the overall acceptable level from the
viewpoint of the quality of the final product. As usual, two separate treat-
ment circuits will recover arsenic in one stream and lead, arsenic, bismuth,
and antimony in the other stream. (
Fire refining of reactor copper is a further opportunity to reduce the
concentration of As, Sb, and Bi in the copper. This behavior of impurities
(particularly Bi, Sb, and Pb) means that under conditions of high oxygen poten-
tial the Noranda process cannot be applied to a concentrate containing these
impurities until electrorefining practice is improved for removal of these
impurities. This is the main reason why the initial applications of the
Noranda process focus on matte making and the elimination of the impurities
via volatilization in the batch converter operation. After converting the matte,
the concentrations obtained in the anode copper are similar to those resulting
from conventional copper smelting.
65
-------
TABLE IV-22
OBSERVED DISTRIBUTION OF MINOR ELEMENTS IN NORANDA PROCESS MAKING COPPER
(30% Oxygen Enriched Air)
Eleoent
Pb
Zn
Cd
As
Sb
Bl
Se
Percent Distribution In Reactor Streams
Off-gas(Dust)
(DO
21
14
95
39
IB
43
60
Tapped
Slag
(B,)
77
86
4.5
14
52
42
21
Reactor j
Copper
(D,)
2
0.2
0.5
47
30
15
19
Copper concentrate anaiyssis: 24.67. Cu, 28-ftX Kc,
Fe/SiOa ratio in slags 1.5
j.ons of X in orf-i:is (tltiiUj
tons of X in reactor fcc*l
tons of >: In Jjri*" !_Ji1 "'
tons of X in reactor
Source; Hackey 1975
b. Scrap
The Noranda process has the same limitations on scrap reuse as the
Outokumpu flash smelting process. In the matte version of the process, only
a limited amount of bulky scrap can be melted in the converters.
There is no thermal limitation on melting scrap in the Noranda reactor
since excess heat, if needed, can be supplied via external extraneous fuel or
increased oxygen enrichment. The limitation would be in terms of the particle
size of scrap and reverts which can be fed to the reactor.
c. The Use of Oxygen
The use of oxygen enrichment in the Noranda process increases unit capa-
city, decreases capital investment per unit of output, and improves energy
efficiency. Pertinent data published by Noranda are presented in Table IV-23
and Figures IV-9 and IV-10.
5. Economic Factors
a. Capital and Operating Costs
Modern smelters installed in the United States should be expected to cost
about $750 per annual ton of anode copper (rated capacity) for the major new
66
-------
TABLE IV-23
TYPICAL COMMERCIAL OPERATION WITH OXYGEN
(Tons)
lurca Uo-isig 'f\r, itlm
Full 1,1*
ปJซ
0>!
*l::?rr"1'
JSCh
* Si
i : f i s j i
00 MB It
ซ,0
IpO
170
*0
04
i;o
t;ป
Ceppir LOU 9' cappci m "(" wfu' oci'^g -"jrc"il 1 i'
HCAFBALAHCE
Cw ซ(, n,jci^.
Tcni
hฃAI OU1PUI
CorTidtfig Cปi (c nelud^ig
t04T>*U*lซfl CJ1)
'If at Cfl Aieซil ซl Copf*r
H*ir cooitfti ei SIปB
"*n Un
TQ!J|
Maian Dm/
tort taซ.
3W
040
4K
I3J
QIS
OJf
S8ซ
* D>ili.brl'en
!)?
SOS
40
30]
44
<ซ0
Tei4l(uei ttquifta irii iซi BlU.'ien cซ9pf'cil ta iar F3.ป9 pป-ซdi jl ซ0' ' iid CM "S">
OfPOASOlTA
|^
-------
I
100 ZOO 300 400 500
INSTANTANEOUS TONNAGE OXTGEN RATE (TON/DAY!
SOURCE Now* M.nn, Lid
F,g_ IV TO
Figure IV-10. Fuel Ratio Versus Tonnage Oxygen Added
pyrometallurgi'cal processes considered here. We have used this figure since
no company has yet built a complete smelter using Noranda reactors. Capital
costs published by Kennecott (Matheson, 1974 and Sharma, 1974), for modifying
their Utah smelter to a version of the Noranda concept suggest that this
assumption is justified.
It should also be noted that the capacity of the plant is increased by
oxygen enrichment.. The degree to which blast air should be enriched is sub-
ject to a number of parameters, not all of which are obvious. However, in
view of the fact that enriching the air to 35% oxygen doubles the capacity of
the Noranda reactor, it is safe to say that the Noranda process using enriched
air is less expensive, in terms of capital cost per unit of output, than the
conventional smelting process of the same capacity.
Table IV-24 shows estimates of capital and operating costs for the Noranda
process based on about 25% oxygen enrichment. This degree of oxygen enrichment
has been used for several campaigns on the Noranda reactor and represents the
oxygen usage when the plant capacity is about 100,000 ton/yr. (The same degree
of enrichment has been used for the evaluation of the Mitsubishi process in
Section E.) Higher degrees of enrichment raise the temperature of the bath and
increase refractory wear to an intolerable extent.
An unusual feature of the Noranda reactor is the maintenance requirements.
Maintenance costs are high and more critical with the Noranda process than with
the reverberatory process. Downtime schedules for repair are projected to be
the following:
68
-------
TABLE IV-24
OPERATING COSTS: NORANDA (MATTE) PROCESS
(Basis: Copper sulflde concentrates, 28.6% Cu, 29.3% Fe, 33.4% S
100,000 ton/yr of copper; 25% 0 enrichment)
CAPITAL INVESTMENT (CI) $/annual ton
OPERATING COSTS
VARIABLE COSTS
Silica Flux
Limestone
H2S04 Credit
Oxygen
Fuel "Oil
Natural Gas
Coal
Electricity
Water: Process
Cooling
Refractories
Direct Operating and
Maintenance Labor (L)
Supervision (S)
Maintenance Materials
Labor Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
, Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
*M _ in"
Unit
ton
ton
ton
ton
106 Btu*
106 Btu
106 Btu
kWh
103 gal**
103 gal
ton
Man-hr
L
CI
L+S
L+S
CI
CI
CI
$/Unit
8.00
10.00
10.00
15.00
2.00
0.65
0.70
0.021
0.02
0.05
300.00
5.75
15Z L.
4Z CI
352 (L+S)
65Z (L+S)
2Z CI
72 CI
20% CI
Units /Ton
$750
0.86
3.33
0.29
6.5
1.3
2.5
366.4
1.06
2.0
0.025
6.2
~
$/Ton Cu
6.88
(33.30)
4.35
13.0
0.85
1.75
7.69
0.21
0.10
7.50
35.65
5.35
30.00
14.35
94.38
26.65
15.00
52.10
150.00
244.15
341.53
**MM - 103
69
-------
16 days for every 72 days of continuous copper making
20 days for every 120 days of continuous matte making
The reactor has to be partially or totally relined during that time.
However, at this time there is a high degree of uncertainty in these figures.
Once a commercial plant starts operating, better data should become available.
b. Comparison with the Base Line
(1) Energy Use
Table IV-24 shows that energy use is decreased for the Noranda process.
As mentioned in Section C, there is no unique conservation in energy form
achievable with any of the new smelting processes.
It should be noted that the Noranda reactor is fitted with a main burner
at the feed end and a secondary burner at the slag tapping end. The heat trans-
fer from the flames to the bath is not optimum; the heat transfer is favored
in the vicinity of the main burner. Most of the heat is taken by the nitrogen
present in the air reacting with the fuel, and goes up to the hood. One way to
distribute the heat generated by extraneous fuels better is to mix carbon with
the feed in the form of coal or coke particles; 3% of the required fuel is
presently injected with the feed to the Noranda reactor in the form of low
quality, high sulfur coal. Table IV-24 takes this into account.
(2) Pollution Aspects
The Noranda smelte'r and its associated acid plant recover a higher amount
of SQ2 (joyer__?p%)_ compared to conventional smelting (50-70%). Table IV-25
shows the total pollution control costs and assumes conversion to sulfuric acid.
(3) Cost Comparison
Comparison of Tables IV-6 and IV-24 shows that the cost of Noranda smelt-
ing is lower because of its higher energy efficiency.
TABLE IV-25
COST OF POLLUTION CONTROL IN THE NORANDA SMELTING PROCESS
Acid Manufacture
Credit for Acid Sales
Water Pollution
Slag Disposal (Tailings)
Dust Disposal
Total Smelting and Converting Pollution Costs
Units
Ion
Ton
Ton
Ion
Ton
S/Unlt
18.88
(10.00)
0.63
24.00
Units/Ton
3.33
3.33
3
0.02
$/Ton of Copper
62.87
C33.30)
3.25
1.90
0.48
The sane amount of discarded dust is shown as for the conventional smelters.
70
-------
E. THE MITSUBISHI PROCESS
1. Concept and Operations
a. Description
The Mitsubishi process consists of three metallurgical stages, each of
which is carried out in a separate furnace (Figure IV-11). Thus, there is a
smelting furnace for concentrates, a converting furnace to oxidize iron in the
matte and make blister copper, and a slag-cleaning furnace. Intermediate
products in the molten state move continuously among the respective furnaces,
which are thus functionally connected with each other.
Copper concentrates, fluxes, and oxidizing air are charged into the smelt-
ing furnace through lances installed vertically through the roof of the fur-
nace and reaching just above the bath surface. Use of the lances, says
Mitsubishi, brings about a high smelting and oxidation rate and simplifies
furnace design and maintenance. Another advantage of this top-blowing method
is that, when necessary, oxygen or fuel oil can also be introduced through
the la,nces. Table IV-26 shows typical blowing conditions through the lances
to. the smelting furnace of the Onahama pilot plant.
The matte and slag produced during smelting continuously flow to the
slag-cleaning furnace as a dispersion of matte drops in the slag. Here, after
being cleaned, the slag is skimmed out continuously to be granulated (average
copper content of this waste slag is about 0.4%), while the matte is sent to
the converting furnace.
In the converting furnace, which is also equipped with lances that intro-
duce blow air, matte undergoes oxidation to blister copper. This blister
copper is continuously tapped out of the furnace and any slag formed during
the converting stage is returned to the smelting furnace via a moving bucket
system.
b. Pilot Plant Results
To establish basic operating techniques, Mitsubishi initially conducted
its test campaigns in the pilot plant by combining the smelting and "converting"
furnaces to obtain white metal: i.e., matte containing about 80% copper and
no iron. Soon afterward, the degree of oxidation was increased and production
of blister copper was started in the converting furnace. Table IV-27 shows
the composition of various types of concentrates smelted during four campaigns;
the fluxes used at the time were silica (85-90% Si02, 4-4.5% Al203) and lime-
stone (53-55% CaO, 1-2% Si02>.
Mitsubishi reports that the daily smelting rate of its pilot plant opera-
tion varied between 5 and 6 mt/m3 of furnace volume. Such a high smelting
rateabout six times greater than that of a reverb, and at least twice as
much as that of a flash smelting furnace (Table IV-28)is achievable for two
reasons. First, each furnace has only a single reaction zone, and the smelt-
ing rate can thus be increased far more than would be possible in a furnace
71
-------
RETURNING BY MOVING BUCKET SYSTEM OR AIR LIFTING
REVERT SLAG
BLISTER
.COPPER
SMELTING FURNACE
SOURCE: E&MJ, August 1972.
SLAG CLEANING
FURNACE | CONVERTING FURNACE
SLAG GRANULATION
Figure IV-11. Schematic View of Mitsubishi's Semi-Commercial Plant
TABLE IV-26
BLOWING CONDITIONS
Smelling Furnace
Feed rate of concei'Tates, mtph 3.0
Blowing rate, NmVhr 2,000-3,000
Oxygen mixed to the blow air, NmVrir 200-300
Combustion rate of fijel oil, liters per hr 150-200
Air for fuel oil combustion, NmVhr 1,500-2,000
Converting furnace
Blowing rat*, Nm'/hf 1.000
Oxygen mixed to thp blow air, Nm''/hr 20-50
Combustion rate of fuel oil, liters per hr' 20
In I hi- converting furnace of the pilot plant, the furnace temperature
could not be maintained h> nutic comcrtmi: alone, owinv to the rela-
IIM-IJ sm.ilt heat ^ซi"-Tiitiijn Fuel oil and nxygen were, tlwrediru, blown
tlinnnili the Mine lancet ,>nh the bUปw air. In a commercial furnace
having a larger capacil)-. the heal generated by converting wontd be
vy
-------
TABLE IV-27
AVERAGE COMPOSITIONS AND THROUGHPUT OF CONCENTRATES
Campaign A B C D1
Duration, hf 384 235 791 530
Concentrates fed, ml 1,031 632 2.102 1.392
Aug. feed rate, mlrjh' 2.69 2.69 2.66 2.&2
Average compositions
o( concentrates, %
Cu
F9
S
Pb
Zn
SiO,
AI.O.
(1) Avcranc feed r.iu w.is smaller Ih.in
downtmic lor tle.trmiji ill the II.i
(2) Tlw^C hi^h v.illifs .iri. .luc ;.i Mlic.i N.IIH! HUM.*' tu the ciiMn-rilr.iU".
(3) The \Uiii forrneil in ihe coim--rimt turiiiite wjs nol returned to the
^niclunt: luituct c\*-x.pi for can>pjij:n O.
194
233
29.4
2.3
6.6
14.2-
1.3
".n'j M"
19.6
23.6
29.1
1.9
7.8
12. 6:
26
;;;ซ",-;;;.
20,9
25.2
28.0
0.4
1.2
I5.bv
2.9
^ iu hec.ui^t.'
23.1
24.5
308
tr.
0.2
7.5
0.1
of ihu
Source: E&MJ, August 1972
TABLE IV-28
SMELTING RATE AND FUEL CONSUMPTION OF VARIOUS SMELTERS
<:nuLting rate, Fuel consumpt tun,
ntpd/ai^ of kg oil per
Process furnace volume me c^nce^trnCL- _
Smelting furnace
Low-grade matte (4QX Cu)1 4.7 H5
High-grade matte (70? CuK 6.0 82
High-grade matte C6CS Cu) 5.5.6.0 35-50
Re verbs rntory EuTaacc .,
Green cliarse, Onahama 0.8 160
Calcine charge, Nnoshlna 1.0 120
Flash srnซltlfiB furn.iro
Ashio5 2,0 50
(1) Thtse are the results of relatively shore campaigns <:.irri<-J out
flc the earlier period oE this project, uslnซ a sm^lttnR fyrruqe
of somGuhat different design. Oxygen w.it; not used.
(2) oxygen of about en Nn t^n >. &nf - V.M--- .idJod '. - clie hlnu .iU.
(3) Conbascion air is preheated ro 300-350eC.
(4) The temperature of calcine is .r "ui t-50ฐC,
(5> Estimaced from the published d,Tta.
Source: E&HJ, August 1972
73
-------
with two reaction zones, e.g., oxidation and settling, since oxidation is pro-
moted by good mixing but settling requires quiescent conditions. Second,
the concentrates and fluxes, injected into the bath through the lances, are
smelted so promptly that the reaction heat generated by the oxidation of sul-
fur and iron can be used effectively.
The composition range of slag formed in the smelting furnace was 3035%
Si02ป 5-7% CaO, 2-6% A^C^, and 40-45% Fe and Zn. The slag was fluxed to
obtain an'Al203/CaO ratio of less than 1.0. Copper content in the slag tapped
from the slag-cleaning furnace during campaign "C" is shown in Figure IV-12
plotted against the matte grade. As Figure IV-12 indicates, the slag factors
(apparently defined as 100 times % copper in slag/% copper in matte) are dis-
tributed mostly between 0.7 and 1.0whereas, in a conventional reverberatory
practice, slag factors are usually between 1.0 and 1.5. Table IV-29 shows
copper losses in the slag before and after cleaning.
c. Converter Slag
Copper content of matte entering the converting furnace varied from 57%
to 63%. Typical compositions of the converting furnace slag were as follows:
7-15% Cu, 40-50% Fe, 10-20% Si02 plus CaO. (Limestone was added to the slag
to obtain good fluidity.) Analysis of the slag also showed that 40-60% of
copper content in the slag was metallic and sulfide copper and the rest was
oxide. Most of the slag's iron content was in the form of magnetite. Before
being charged into the smelting furnace, the converting furnace slag was
cooled and crushed. In the larger plant the slag is granulated, dried, and
returned' to the furnace. Slag recirculation has no significant influence on
copper losses in the smelting furnace slag.
d. Blister Cooper
Typical analyses of blister copper, shown in Table IV-30, indicate a
copper content varying from 98-99% and a sulfur content of 0.4-0.8%. This
sulfur content was lower than in the blister copper that would be produced in
a furnace having three separate phases (i.e., slag, white metal or low iron
matte and copper phases). The low sulfur content in Mitsubishi's blister
copper seems to indicate that the converting furnace was operated under such
conditions that no matte layer existed as a separate phase between the slag
and copper phases. Otherwise, the blister copper would be nearly sulfur
saturated.
Other impurities such as Pb, As, and Sb were present but in small amounts.
e. Flue Dusts
Since the concentrates injected through the lances in the smelting fur-
nace were readily and promptly captured by the bath, the amount of mechanical
flue dusts was smallor nearly proportional to the Pb and Zn content of the
concentrates. These dusts, collected in the cooler, were sent back to the
process. Dusts rich in Pb and Zn were sent to a lead-zinc smelter.
74
-------
1.0 r~
20 3D 40 5Q 60 70 BO
THE GRADE OF MATTE FORMED IN THE SMELTING FURNACE, Cu%
Figure IV-12. Copper Losses in Smelting Furnace Slag
TABLE IV-29
COPPER LOSSES IN SLAG
% Cu in Slag
Sample No.*
1
2
3
4
5
6
7
Before Cleaning
1.53
0.69
0.82
0.80
0.77
1.21
0.92
After Cleaning
0.83
0.43
0.42
0.35
0.49
0.45
0.49
Sampling interval was one hour.
75
-------
TABLE IV-30
TYPICAL ANALYSES OF BLISTER COPPER PRODUCED
Analysis
Campaign
A
B
C
D
Cu
98.0
98.0
98.5
99.0
S^
0.6-0.7
0.6-0.8
0.4-0.8
0.7-0.8
Pb
0.22*
0.50*
0.05
0.01
Fe
0.08
0.001
0.008
0.004
Zn
0.05
-
-
*
These high values are due to the high Pb contents in the concentrates.
2. Current Status
Development work on this process "began in 1961. After a series of pre-
liminary experiments, Mitsubishi constructed a prototype pilot plant with a
monthly capacity of 500 tons of blister copper at the Onahama smelter.
Ishikawajima-Harima Heavy Industries was involved in this development work.
When pilot plant operations proved the technical feasibility of the process,
Mitsubishi decided to build a semi-commercial plant having a capacity of 1,500
metric tons/month of blister copper. This semi-commercial plant has been on
stream since November 1971. A commercial operation producing 50,000 ton/yr of
blister copper is now being started at Naoshima and a 130,000 ton/yr plant is
being designed by Montreal's SNC Consultants for the Kidd Creek smelter of
Texasgulf.
3. Effluent Control
Figure IV-13 is a schematic flow diagram of copper smelting using the
Mitsubishi process. The emission sources of pollutants are identified in four
categories: air, solid, water, and fugitive. Except for the slag treatment
method, the pollution controls required are identical to those necessary for
the Outokumpu flash smelting process. Table IV-31 shows the magnitude of
these streams and the major constituents.
a. Air
Upon leaving any of the three furnaces, the mixed off-gases are expected
to average over 10% S02 when the smelting furnace is operated with air enriched
76
-------
CONCENTRATE
DUST BLEED
EFFLUENT TYPES: -AIR; w -WATER
FUGITIVE
STACK
ANODES
Figure IV-13. Emission Sources in the Mitsubishi Continuous Copper Smelting Process
-------
TABLE IV-31
EMISSIONS FROM THE MITSUBISHI PROCESS
Stream
Air Pollution
A-l - Acid Plant Tail Gas
A-2 Anode Furnace Gas
Water Pollution
W-l - Slag Granulation
W-2 - Acid Plant Slowdown
W-3 Contact Cooling
Stream Size
55,000 scfn
NA
Major Constituents
S02 - 0.05%
Flue Gas,. Some SO.
50,000 liter/103 kg TDS, TSS
14,000 liter/103 kg IPS, TSS, Acidity
7,800 liter/103 kg IDS, TSS
Solid Wastes
S-l - Cleaned Slag
S-2 - Dust Bleed
3 ton/ton copper
0.3 ton/ton copper
Iron Silicates
Copper Oxides, Minor Elements
Fugitive Emissions
NA
NA - not available
to 25% oxygen. This steady, high SC^ gas generation is a significant advantage
over the conventional reverberatory process, as sulfur can be readily recovered
as sulfuric acid. Since the molten liquids flow continuously over very short
distances, minimum air pollution is generated in transfer operations, and
"converter aisle losses" typical of conventional operations are avoided. Thus,
fugitive emissions are expected to be lower than for conventional or for
Outokumpu and Noranda (matte) processes.
After cooling and dry cleaning in electrostatic precipitators or fabric
filters operated above the dew point, the collected off-gases usually require
a wet cleaning stage to remove any fine particulates and excess moisture. The
cleaned gases are then admitted to a double contact acid plant for H2S04 manu-
facture. The total sulfur recovery is over 90%, as with all these advanced
pyrometallurgical processes.
b. Solid Wastes
The converter slag is entirely returned to the smelting furnace, so that
the only slag to be disposed of comes from the slag^cleaning furnace. About
three tons of granulated slag per ton of anode copper produced must be land-
filled at a cost of about $5/ton. The total cost incurred to a smelter producing
100,000 ton/yr of anode copper is therefore $l,500,000/yr.
78
-------
Flue dusts are generated from all three furnaces in amounts expected
to be a little smaller than obtained in flash smelting. The dusts are normally
concentrated in volatile impurities and have to be bled to the extent necessary
to avoid impurity build-up. The remaining dust is recycled. The cost of dis-
posing of dusts is the same as for the base line - $24/ton of dust.
A schematic illustration of the flows of minor elements in a Mitsubishi
smelter is shown on Figure IV-14. When making matte containing 60% copper,
one may estimate that a great similarity exists between this process and the
Outokumpu flash smelting process. Distribution of impurities in these streams
is not yet known, although it is expected to be similar to that from the flash
smelting process.
c. Water
The water effluent streams in the Mitsubishi process are those associated
with the acid plant, anode cooling, and slag granulation; These streams are
identical to those discussed earlier for the Outokumpu process.
4. Technical Considerations
a. Impurities
By the nature of the process, the impurity distributions in the Mitsubishi
process appear to be somewhat similar to those in Outokumpu flash smelting.
One would expect precious metals to be retained in the metal. Cobalt, tin,
and nickel would tend to be retained in the matte due to the reducing conditions
of the electric furnace. Zinc and lead would be volatilized and recovered
mainly in the dusts of the settler. Figure IV-15 shows the distribution of
lead and zinc between matte and slag. Selenium and tellurium will follow the
sulfur in the matte. Arsenic, tin, and antimony will tend to be recovered as
oxides in the dusts of the smelting furnace; the dusts from the converting
furnace will contain more bismuth, lead, antimony, selenium, and tellurium.
5. Economic Factors
a. Capital and Operating Costs
Table IV-32 shows estimates of capital and operating costs (including
pollution control costs) for the Mitsubishi process based on about 25% oxygen
enrichmentthe same degree of enrichment assumed in the Noranda process. A
lime slag is also assumed for the converting reactor.
b. Comparison with the Base Line
(1) Energy Use
Table IV-32 shows that energy use is reduced from 18.6 to 9.8 million
Btu/ton of fossil energy. As mentioned in Section C, there is no further con-
servation of energy form since coal, oil, or gas can be used with any of the
new smelting processes.
79
-------
CONCENTRATE INTAKE
RECYCLE
GRANULATED
SLAG
ESLAG
DRYING
AND
SMELTING
1
DUST
MATTE
AND
SLAG
SLAG
CLEANING
i
DUST
MATTE
CONVERTER
DUST
DUST
Figure IV-14.
ANODE
Diagram of Impurity Flow Within a Smelter
Using the Mitsubishi Process
cc
tu
fc
cc
LU
CD
UI
X
o
i-
z
UI
8
ฃ
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
D
o
O
D
0
A
&
D
O
P A
^
D
O
Q
A
& c
O Pb
A As
) O
y\
G E
1 n
3
o
D
O 1")
0.10
006
0-04
n no
0
DC
LU
o
o
IT
LU
m
Hi
I
u.
O
I-
LU
I
0 5 10 15 20 25 30
COPPER CONTENT OF CONVERTING FURNACE SLAG, %
Figure IV-15. Slag Composition and Impurity Levels
80
-------
TABLE IV-32
OPERATING COSTS: THE MITSUBISHI PROCESS
CAPITAL INVESTMENT (CI) ?/annual ton
OPERATING COSTS
VARIABLE COSTS
Silica Flux
Limestone
H^SO, crudit
2 4
Oxygen
Fuel Oil
Natural Gas
Coal
Electricity
Water: Process
Cooling
Refractories
Direct Operating and
Maintenance Labor '(')
Supervision (S)
Maintenance Materials
v rhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
j
Unit
ton
ton
Con
ton
1C6 Ecu
1C6 Btu
106 Btu
kWh
103 gal
103 gal
ton
Man-hr
L
CI
L+S
L+S
CI
CI
CI
$/Unit
8.00
10.00
10.00
15.00
2.00
0.65
0.70
0.021
0.02
0.05
300.00
5.75
15% L
4% CI
35% (L+S)
65% (L+S)
2%
n ci
20% CI
Units/ton
$750
0.86
0.118
3.33
0.3
8.5
1.3
-
366.4
1.06
2.0
0.014
6.2
$/ton Cu
6.88
1.18
(33.30)
4.35
17.00
0.85
7.69
0.21
0.10
4.20
35.65
5.35
30.00
14.35
94.51
26.65
15.00
52.50
150.00
244.15
338.66
81
-------
(2) Pollution Aspects
The Mitsubishi smelter and its associated acid plant recover a higher
amount of S02 (over 90%) than does conventional smelting (50-70%). Table IV-33
shows the pollution control costs. These are about 14% of the total smelting
costs.
The Mitsubishi process potentially has less problems with fugitive emis-
sions than other new processes since all the transfer points are covered and
hooded.
(3) Cost Comparison
Comparison of Tables IV-6 and IV-32 shows that the cost of the Mitsubishi
process is lower than for the conventional smelting primarily because of its
increased energy efficiency. Also, because the process employs a different
smelting and converting unit, it should be more flexible in its ability to
treat impure concentrates.
TABLE IV-33
COST OF POLLUTION CONTROL IN THE MITSUBISHI CONTINUOUS
COPPER SMELTING PROCESS
Units $/Unit Units/Ton $/Ton of Copper
Acid Manufacture Ton 18.88 3.33 62.87
Credit for Acid Sales Ton (10.00) 3.33 (33.30)
Water Pollution Ton - - ' 3.25
Slag Disposal (Granulation) Ton 5.00 3 15.00
Dust Disposal Ton 24.00 0.02 0^48
Total Smelting and Converting Pollution Costs 48.30
F. THE USE OF OXYGEN IN SMELTING
1. Concept and Operations
Copper smelting can be conducted with pure oxygen or by using oxygen-
enriched air. Reasons for using oxygen or oxygen enrichment include:
Increasing processing temperatures and process heat rates.
Decreasing the nitrogen content of the flue gases (when high S02
concentrations are needed) and increasing fuel efficiency (particu-
larly where waste heat is not recovered).
82
-------
Increasing the specific capacity of furnaces so that production of
metal is increased significantly for a given size of reactor.
Examples of the above in copper smelting are as follows:
Fuel efficiency in conventional reverb smelting can be increased with
oxygen enrichment. The limitations on oxygen use arise from the
decrease in refractory life as operating temperature increases. Con-
siderations of refractory wear normally limit the enrichment of con-
verter air to about 27% oxygen for conventional bottom-blown converters.
The enrichment not only increases S02 concentrations but also increases
the scrap melting capability of the converter.
Furthermore, since the sensible heat in converter gases is not usually
recovered, and less nitrogen has to be heated to exit gas temperatures,
there is a significant saving in energy. The net heat generation
under these conditions can be used for melting scrap (Parameswaran
and Nadkarni, 1975). The converter is an ideal unit for melting and
refining several types of low-grade copper scrap which are not
normally handled by the secondary copper industry.
The specific capacity of furnaces such as the flash smelting furnaces
can be increased by the use.of oxygen. In this case, the increase in
energy efficiency is not as great as in the case of converter opera-
tions since waste heat recovery is already practiced. However, the
increase in smelting rate significantly decreases the capital costs
'per unit of output. This is advantageous for new plants and extremely
attractive for existing plants which need to increase production with
normal capital outlay.
We have chosen the example of Outokumpu flash smelting to illustrate
the beneficial effect of oxygen use in copper smelting.
2. Current Status
Oxygen enrichment of Outokumpu furnace gases has been practiced since about
1972 at Harjavalta. This has enabled almost completely autogenous operations
of the furnace and resulted in a large increase in furnace capacity.
3. Effluents
An oxygen plant is a user of significant quantities of cooling water for
indirect cooling but produces no known deleterious effluents. Cooling tower
blowdown will contain oil and grease, corrosion inhibitors, and chlorides con-
centrated from the feed water. The plant consumes electricity, and the genera-
tion of this electricity at a central power station has the associated pollution
consequences; however, these problems are readily identifiable and are not
considered further in this study.
As far as smelter effluents are concerned, we believe that the use of
oxygen will not significantly change the nature and amount of the effluent and
Table IV-9 will still apply.
83
-------
Because of higher operating temperatures, the generation of smelter dusts
might increase, resulting in an increased circulating dust load, however, there
do not appear to be any published data in this area. Higher furnace tempera-
tures might also result in more NOX formation in the furnace. However, since
furnace gases are scrubbed and treated in the acid plant, these nitrogen gases
are probably absorbed in the acid, as occurs in the Guy-Lussac towers in the cham-
ber process for making sulfuric acid. There are no published data in this area.
4. Technical Considerations
Interplay of air preheating temperature and oxygen enrichment in Figure
IV-16 illustrates the effect of the process air preheating temperature or
oxygen enrichment on the net energy consumption. As it appears from this figure,
a change in either of the above-mentioned two factors is of minor importance
only, because the lower the preheating temperature or the percentage of oxygen
enrichment, the greater the amount of oil required for smelting, which in turn
means increased off-gas amount and waste-heat credit. However, the figures
given indicate that an increase in the process air preheating temperature some-
what decreases the overall net energy requirements in smelting provided that
the thermal efficiency of the air preheater is adequate. The. use of oxygen
does not lead to an improved saving of energy but it reduces the consumption
of extraneous fuel and considerably increases the smelting capacity.
The flash smelting furnace capacity can be increased by raising either
the preheating temperature of process air or the percentage of oxygen enrich-
ment. The simultaneous raising of the preheating temperature and oxygen enrich-
ment rapidly decreases the off-gas flow and gives more capacity. Once the
autogenous point of the reaction shaft is reached, this point cannot be exceeded
without overheating the reaction shaft. In such a case, increase in capacity
by the use of additional oxygen is possible only if the process air tempera-
ture is lowered.
OFF GAS
VOLUME
(Nm3/ton CONCI
2.000-
1,330
1.000 -
25ฐC
ENERGY
CONSUMPTION
kป AT/TON CONC
-30
-20
21
25
30
35
IW02
Figure IV-16.
Off-Gas Volume and Energy Consumption at the
Harjavalta Copper Smelter
84
-------
5. Economic Factors
a. Capital and Operating Costs
Table IV-34 presents estimated capital and operating costs (including
pollution control) for the Outokumpu process using air and using about 35%
oxygen (autogenous operation in the reaction shaft). The table shows that
the operating costs decreases mainly because of lower fixed costs. In Table
IV-34, we have shown oxygen as a raw material purchased from an "across the
fence" plant. If the cost of such a plant were included in the fixed charges,
one would still show a saving in operating cost.
G. METAL RECOVERY FROM SLAG
1. Background
a. Introduction
In conventional copper smelting, converter slag is recycled to the reverb
and all the slag tapped from the reverb is discarded. The copper contained in
this discard slag is lost. The amount of copper lost with the slag is quite
significant, about 1.5-3% or more of the copper in the feed materials.
It is well known that the concentration of copper in the reverb slag
varies directly with matte grade (see Figure IV-17). With newer processes
where there is significant sulfur oxidation in the smelting step, the matte
grades are higher and consequently the concentration of copper in the slag is
also higher (explained in b). The oxidation of sulfur in the smelting step is
beneficial from a pollution control viewpoint since the off-gases from these
smelting units contain 10-15% SC>2 (suitable for sulfuric acid manufacture) com-
pared to SC>2 concentrations of 1-2% normally found in reverb gases. However,
the slag contains an unacceptable high concentration of copper and cannot be
discarded. The availability of technology for metal recovery from slag has
made it possible to smelt at high matte grades, recover much of the copper in
the slag, and produce a slag stream that can be discarded without undue economic
penalty. Thus, these processes for recovering metal from slag are considered
an essential factor in all of the new smelting technologies.
b. Mechanism of Metallic Losses in the j>lag
In copper pyrometallurgy, the molten feed materials separate into two
phases - slag and matte. The distribution of copper between these phases is
determined by solubility, kinetic factors, surface phenomena, and other para-
meters governing the separation of the 'melt into two distinct phases:
The liquid phases have some limited mutual solubility: part of the
copper and other metallic values are chemically "lost" by solution
to the slag phase. The amount and nature of these losses are governed
by the thermodynamics of the system, and our knowledge of the chemistry
of this process is still incomplete. Sulfides, oxides, and elemental
copper are partially soluble in fayalite slags in contact with copper
mattes.
85
-------
TABLE IV-34
OPERATING COSTS: OUTOKUMPU FLASH SMELTING USING AIR AND 35% OXYGEN
(Basis: Copper sulfide concentrates, 28.6% Cu, 29.3% Fe, 33.4% S
Initial unit built for 100,000 ton/yr of copper)
Without Oxygen
With Oxygen
(35% enrichment)
OPERATING COSTS
VARIABLE COSTS
Silica Flux
H2S04 Credit
Oxygen
Fuel Oil
Natural Gas
Electricity
Water: Process
Cooling
Refractories
Direct Operating and
Maintenance Labor (L)
Supervision (S)
Maintenance Materials
Labor Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
Unit
$/Unit
innual ton
ton
ton
ton
106 Btu
106 Btu
kWh
10;? gal
10J gal
ton
Man-hr
L
CI
L+S
L+S
CI
CI
CI
8.00
10.00
15.00
2.00
0.65
0.021
0.02
0.05
300.00
5.75
15% L
4% CI
35% (L+S)
65% (L+S)
2% CI
7% CI
20% CI
Units/ $/Ton
Ton Cu
$750
0.86 6.88
3.33 (33.30)
12.7 25.40
1.3 0.85
366.4 7.69
1.06 0.21
2.0 0.10
0.014 4.20
6.2 35.65
5.35
30.00
14.35
97.38
26.65
15.00
52.50
150.00
244.15
341.53
Units/
Ton
0.86
3.33
1.3
5.9
1.3
366.4
1.06
2.0
$/Ton
Cu
$500
6.88
(33.30)
19.50
11.80
0.85
7.69
0.20
0.10
0.014 4.20
6.2
35.65
5.35
20.00
14.35
93.28
26.65
10.00
35.00
100.00
171,65
264.93
86
-------
co
2
CL
UJ
CL.
CL.
8
1.5
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Figure IV-17.
0 10 20 30 40 50 60 70 80 90 100
COPPER IN MATTE {%)
SOURCE: Ruddle, 1953.
Copper Content of Reverberatory Slags Plotted Against
Matte Grade
The best correlation so far has been proposed by Splra and Themelis
(1969). Their findings are summarized in Figure IV-18. The most
important variables governing the solubility of copper in fayalite
slags are the matte grade and the oxygen potential. The higher the
matte grade, the higher the solubility of copper in the slag. Also,
the higher the oxygen potential for a given matte grade, the higher
the copper loss in the slag. The practical consequence is that
processes operating at high matte grade and high oxygen potential
(continuous processes, late stage of converting) enhance the chemical
dissolution of copper in the slag. Figure IV-18 also shows the
theoretical limit to which a slag can be cleaned of its copper con-
tent at a given oxygen potential by pyrometallurgical techniques
such as electric arc furnace settling.
Copper is believed to dissolve as oxide (Cu20) in the slag during the
later stage of converting. Ruddle and .Taylor (1966) have proposed
t;he correlation between Cu20 in slag and oxygen potential in the
Cu-Fe2Oo-FeC03-Si02 system. Their findings appear in Figure IV-19.
Copper is not the only element chemically soluble in the slag. Some
ores contain sizeable amounts of other easily-oxidized metals. The
oxides of lead, zinc, tin, nickel, and cobalt are highly soluble in
fayalite; oxidizing conditions also favor their retention in the slag.
This is one way in which impurities leave the system.
87
-------
NUMBERS INDICATE %Cu IN MATTE
0 123456789 10
Figure IV-18.
SOURCE: Spira and Themelis, 1969.
Solubility of Copper in Silica-Saturated Slag as a Function
of Oxygen and Copper Content of Matte
16
12
O o
CM O
3
\ I i r
IRON
SATURATION
MAGNETITE
SATURATION
-12
-10 -8 -6
LOG PQ2 (ATM.)
SOURCE: Ruddle and Taylor, 1966.
Figure IV-19. Effect of Oxygen Pressure on Cu20 Content of Slag
88
-------
The separation of the slag phase and matte (or copper) phase is
slowed down by the combined action of several physical parameters:
turbulence, viscosity of the slag, surface tension, and gravity.
During the smelting operation, droplets of matte coalesce and sink
to the bottom of the furnace to form the matte layer. The violent
turbulence created by submerged gas injection as well as the sulfur
dioxide evolution from the matte or copper phase tend to promote a
certain amount of dispersion, carrying droplets of copper or matte
into the slag. This tendency of matte particles to become buoyant
is favored by increased matte grade. The settling speed is also
fairly dependent on the slag viscosity, which in turn depends on the
silica content of the slag (higher silica means higher viscosity).
As it turns out, this mechanical entrapment of copper and other
metallic values in the slag accounts from anywhere between 25 and
75% of the losses occurring during converting or under other turbulent
conditions.
Copper mattes contain other valuable elements, in addition to copper.
By the same entrapment mechanism, proportional amounts of gold, silver,
platinum, etc. are retained in the slag.
Flash smelting slags contain 1-2% copper, because the fast smelting
rate does not allow sufficient time for settling and the finer matte-droplets
disperse in the slag. Converter slags may contain from 3-6% copper in most
cases, depending on the practice followed. The Noranda reactor slag contains
from 6-7% copper when making matte, and 10-12% copper when making copper.
2. Slag Cleaning Via Flotation
a. Concept and Operation
(1) Slag Composition
The presence of sufficient amounts of silica is a requirement for copper
smelting, since silica is the fluxing agent that ties up iron in the concentrate
as iron silicate. However, silica is a strong glass-forming component, and
therefore low-silica slags (20-25% Si02) are preferable in the milling process
because they are easier to grind. Also, there is less slag produced per ton of
concentrate smelted. Ultimate copper loss in discard slag is proportional to the
quantity of slag, hence there is incentive to minimize slag quantity.
Alumina and lime also tend to promote the formation of the amorphous,
vitreous phase and thus inhibit the precipitation of copper sulfide particles..
Zinc in the slag above the 6% level has the same effect.
"Magnetite" amounts to a sizeable percentage of converter and oxygen
smelting slags, sometimes up to 30%. One advantage of the milling and flotation
process over simple recycling to the smelting furnace is the decrease in the
quantity of magnetite fed to the smelting unit; otherwise, magnetite tends to
build up as solid accretions in the cooler parts of the furnace, thus diminish-
ing the available smelting volume. It should be noted that in copper smelting,
"magnetite" refers to any of several compounds such as magnetite, chrome spinels,
and olivines which behave similarly.
89
-------
(2) Cooling Rate
It is important to cool the slag as slowly as possible, usually over a
period of several days. The advantages of this practice are as follows:
The solubility of copper (metallic or as sulfide) decreases with _
decreasing temperature. Slag quenching would keep copper in metastable
solution, whereas slow cooling favors the precipitation of copper
sulfides and metallic copper as larger, discrete particles.
Slow cooling gives more time for surface tension forces to pro-
mote coalescence of entrapped and newly precipitated copper-bearing
microscopic droplets.
The optimum temperature for these two diffusion-related phenomena to be
most effectively enhanced is just below the slag freezing temperature. Slow
cooling can typically decrease the copper content of the tailing by a factor
of two. In addition, slow cooling promotes the crystallization of fayalite as
prisms of 100-200 microns with ferrite particle intergrowth of 10-80 microns.
This configuration is easier to crush than quenched amorphous slag.
(3) Process Description
After the slow cooling stage, the slag is crushed and ground and copper is
recovered via froth flotation. This process is very similar to ore crushing
and flotation. In fact, a mill built for one purpose can easily be modified to
serve the other. However, due to differences in specific gravity, grindability,
precious metals recovery, copper content, etc., ore and slag cannot be simul-
taneously fed to the same circuit. A schematic diagram of such a circuit
appears in Figure IV-20.
After crushing, the slag is ground, usually to a range of -270 to -325
mesh. Although a few smelters like Harjavalta have been successful in achiev-r .
ing autogenous grinding, most others use steel or cast iron balls for grinding.
The consumption of grinding media is an important cost factor, as it is difficult
to go below 8 Ib of steel/ton of slag milled.
The flotation circuit is rather simple, as only one concentrate is to be
separated. Sodium amyl xanthate, sodium isopropyl xanthate, or other alkali
xanthates are used as a collector at an approximate rate of 270-300 g/metric
ton. A frothing agent is then added (Dowfroth 250 or an equivalent) and the
cell overflow is pumped to a thickener from which the thickened concentrate
slurry is fed to a disc filter. The moisture in the filtered concentrate is
fairly high, due to its fineness. At Harjavalta the slag concentrate contains
13% water. The concentrate is then fed back to the smelter, usually mixed
with ore concentrate, flue dust, etc.
(4) Efficiency of the Process
Table IV-35 summarizes the efficiency of slag flotation for selected com-
mercial operations.
90
-------
CONCENTRATE
TO
SMELTER
Figure IV-20. Typical Slag Milling Flowsheet
TABLE IV-35
EFFICIENCY OF METALS RECOVERY FROM SLAGS BY FLOTATION
Cu
Ag
Au
Pt
Aa
Sb
21
Fa
Si
S
N>
Zn
Cd
Sa
la
Hi
Sn
Co
HORAfflA
Continuous
Process
97
93
95
32
29
23
11
91
IS
11
22
73
58
11
17
NIPPON HIKING
HITACHI
SAGANOSEK1
92.7 ! 91.6
93.7
94.8
89.2
93.6
MITSUBISHI
NAOSHIHA
93.4
97.3
100.0
56.7
12,4
OUTOKUMPU
HARJAVALIA
90.1
CAHANEA
COHVEP.TEP,
SLAG
95.3
94. 8
21.7
16.14
Note: Thla table shous only the data released by che respective companies.
Blanks Indicate that liata are not available.
91
-------
Copper and precious metals are recovered from slag with over 90%
efficiency. Sulfur is also recovered in the concentrate. Elements that
remain with the slag are oxides such as iron, lead, zinc, nickel, etc. These
are combined with silica and are discarded as tailings.
The overall importance and wide diversity of the secondary copper stream
coming to and leaving the flotation operation is illustrated in Table IV-36.
The range is from a low of 4.4% for a conventional plant in Naoshima to over
5p% with the Noranda reactor.
TABLE IV-36
SLAG FLOTATION OPERATIONS
Coppe- in slag (2)
Copper in slag concentrate (Z>
Copper in tailings (wX)
Percentage of the copper in
smelter feed lost in tail-
ings
Percentage of copper in
smelter feed entrained
in slag before treatment
Noranda
Continuous Process
Making Copper
12(1)
50
0.5
1.S6
51
Naoshima
3<2)
23
0.32
negligible
4.4
Bar j aval ta
2<3)
12.2
0.36
0.27
1.7
6<2>
35
0.52
0.42
5.6
NOTES: (l) Reactor slag
(2) Converter slag
(3) Flash furnace slag
(5) Treatment of Tailings
The usual treatment for the tailings is to pump them into a lined tailings
pond near the plant. The clear water is reused together with the overflow of
the thickener, the filter water, and the water from the vacuum pumps. The net
water consumption is quite small, less than 10% of the circulating water load.
The tailings still contain up to 0.5% copper, and the very fine particle
size of this material has led a number of investigators to think that further
copper recovery could be achieved by sulfuric acid leaching under aerated con-
ditions. Indeed, such a procedure carried out in the laboratory for 1 hour'at
pH 3.0 and 70ฐC was shown to bring down the copper concentration of the taiiings
from 0.5% to 0.25%. Copper could then be recovered from the liquor by cementa-
tion on iron, and the barren solution could be recycled or neutralized with
lime. However, such an operation has not been commercialized nor evaluated on
a larger scale.
92
-------
b. Current Status
Nippon Mining Company apparently was first to develop this process in
connection with its oxygen converter smelting. It has since been generally
adopted by several other Japanese copper smelters. This process is also used
for treating slag at flash smelters such as Outokumpu, and Noranda is presently
using it to treat the slag from its continuous copper/matte-making unit.
c. Effluents
(1) Air
No air pollution is generated by this process. Since slag return ports
are eliminated from the smelting furnace, this reduces air infiltration and
simplifies emission controls on reverberatory or other types of smelting furnaces.
(2) Water and Solid Waste Disposal
As noted, the process produces finely ground tailings which are stored in
lined ponds. The water from the tailings ponds is recycled as shown in
Figure IV-^20.
As long as a tailings pond is in use, its surface is water covered and
dust emissions are not a problem. After a pond is abandoned, the pond must be
covered with dirt and revegetated. For a smelter producing 100,000 ton/yr of
copper, an area approximately 475 sq ft and 24 ft deep is required annually.
The cost for solid waste disposal is about $1.90/ton of copper produced.
3. Slag Cleaning in Electric Furnaces
a. Concept and Operations
The metallics exist in slag either as entrapped globules of a second
phase (matte or copper) or as metals dissolved in the slag. The electric fur-
nace recovers metallic values by counteracting both mechanisms of loss into the
slag:
By keeping the slag molten and fluid at 2190-2370ฐF for an
extended period of time, it allows for small dispersed entrapped
particles to coalesce, settle, and form a matte pool at the bottom
of the furnace.
By providing a reducing environment, it decreases the solubility of
copper and other easily oxidizable elements in the slag. To maximize
this effect, coal can be added to the bath to supplement the action
of the carbon already present in the electrodes. Also, pyrite is
usually added to match the sulfur requirement of the metals going
from oxides in the slag to sulfides in the matte and reduce the
oxygen potential of the slag.
After sufficient retention time has elapsed, the matte phase is tapped
and sent to the converter together with the matte coming directly from the
smelting furnace. Because of the pyrite additions, the grade of the cleaning
93
-------
furnace matte is somewhat lowered. Typically, if a 50-60% Cu matte is
produced in the smelting unit, a 20-40% Cu matte will come out of the electric
furnace. The slag tapped from the electric furnace is discarded. A common
practice is to granulate and then landfill it.
The fumes of the furnace are collected in bag houses or electrostatic
precipitators and recycled to the smelting unit unless they are high in
impurities.
The electric furnace can be operated either as a resistance furnace or
an arc furnace. The latter mode of operation increases power consumption by
almost a factor of two but also increases copper recovery.
The overall copper recovery from slag by electric furnace is comparable
to or slightly below that achieved by slag flotation. Typically, a power usage
or 100-200 kWh/ton leads to a discard slag containing 0.3-0.5% copper. Precious
metals (silver, gold, etc.) are recovered almost entirely in the matte phase.
Metals with a high vapor pressure such as zinc and cadmium are volatilized and
recovered in the dust as oxides. Bismuth, antimony, and lead tend to volati-
lize as sulfides. Nickel, cobalt, tin, and lead are consistently recovered in
excess of 50% of that in the matte phase.
b. Current Status
Electric slag cleaning has been used in Finland, Sweden, and Japan. At
Harjavalta, Outokumpu has retained electric furnace slag cleaning for its
nickel plant and is using flotation in the copper plant. In many smelters, both
electric furnace and flotation are used for slag cleaning. The slag from the
smelting unit is directed to the electric furnace. The converter slag may be
recycled to the smelting unit or treated via flotation.
Other ways to use electric arc furnaces to clean slags have been envisioned.
The most recent development in this direction is the Mitsubishi process, where
both matte and slag flow from the smelting unit to an electric arc settler.
(See discussion of the Mitsubishi process.)
c. Effluents
(1) Air
The fumes coming off the furnace must be collected, and the dusts are
further processed for recovery of metallic values (zinc, cadmium, etc.). The
volume of cleaned gas is insignificant and can be merged with the tail gas of
the acid plant prior to venting or mixed with the feed gas to the acid plant.
(2) Water and Solid Waste Disposal
The slag from the electric furnace can be disposed of via dumping in the
molten state or dumping after granulation. Since the slag is similar in all
respects to reverb slag, disposal is identical to that in the base case and
has already been discussed in Section B.
94
-------
4. Economic Factors
In this section, both slag flotation and electric cleaning are discussed
together.
a. Capital and Operating Costs
Based on data in the literature, it appears that a 600 ton/day slag flota-
tion mill would cost about $4.2 million. This would include facilities for
slow cooling of the slag.
Table IV-37 shows capital and operating costs for both slag flotation and
electric furnace cleaning (resistance furnace operation). The costs are shown
on the basis of tons of slag into the plant because this is the critical param-
eter that determines plant size and operating costs. Because copper contents
of the slag can vary widely, it is inappropriate to normalize these costs on
the basis of copper recovered. As a rough estimate, the cost of recovered
copper is 10-18ฃ/lb. In slag flotation, the amount of copper in the slag has
little influence on the residual copper in the tailings where fineness of grind
is the important factor. The optimum grind is achieved when the incremental
copper recovery achieved by additional grinding balances the marginal cost of
grinding. In many cases, the optimum grind is about 90% in the -270 to -325
mesh range.
Table IV-37 also presents capital and operating costs for electric furnace
cleaning. An electric furnace is less expensive, about $2 million for a
600 ton/day furnace. Also, land and building requirements favor the electric
furnace over the flotation route. As the slag volume increases and several
electric furnaces are needed, these cost differences become less divergent.
Table JV-37 shows that the slag milling process is more expensive than
electric furnace cleaning. However, it usually allows recovery of a sufficient
amount of extra copper to pay for the higher processing cost. For example, if
the electric furnace discard slag is 0.4-0.6% Cu and mill tailings are 0.3% Cu,
the incremental copper recovery in a 600 ton/day mill is Q.6-8.1 tons, equiva-
lent to about $720-2200/day. The difference in operating cost is about $700,
which balances the incremental recovery. The above suggests that each slag
flotation scheme has to be tailored to the specific operation and optimized.
H. THE ARBITER PROCESS
Since the Arbiter process is the only hydrometallurgical process selected
for detailed analysis, we have included in this section several general com-
ments that would apply to most hydrometallurgical processes for sulfide
concentrates.
1. Concept and Operations
The basic process consists of five separate stages for treatment of copper
concentrates. A block flow diagram of the process is shown in Figure IV-21.
In the first stage, the copper concentrate slurry is leached with ammonia and
oxygen in two parallel trains of reactors. Each of the reactors is well mixed
95
-------
TABLE IV-37
OPERATING COSTS: SLAG GLEANING PROCESS
(Basis: 600 tons/day of Slag)
Electric
Furnace
Slag
Flotation
CAPITAL INVESTMENT (CI)
OPERATING COSTS
VARIABLE COSTS
Grinding Media
Electrodes
Electricity
Water
Direct Operating &
Maintenance Labor (L)
Supervision
Maintenance Supplies
Labor Overhead
TOTAL VARIABLE
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED
TOTAL
Unit $/Unit
Ib 0.30
ton 300
kWh 0.0021
103 gal 0.20
Man-hr 5.75
15% L
4% CI
35% L
65% L
2% CI
7% CI
20% CI
Units/ $/Ton
Ton Slag
$2,000,00
0.0038 1.14
100 2.10
0.31 1.78
0.27
0.38
0.72
6.39
1.33
0.19
0.66
1.90
4.08
10.47
Units/ $/Ton
Ton Slag
$4,200,000
5 1.50
75 1.58
0.08
0.2 1.15
0.17
0.80
0.46
5.74
0.86
0.40
1.40
4.00
6.66
12.40
96
-------
AMMONIA RECVCLE
MAKE UP
AMMONIA
OXYGEN
CONCENTRATE
FEED
FROM SMELTER
COOLING
WATER
11=
L
r
PREGNANT LIQUOR
FILTRATION
DILUTION & LAUNDER
SPRAY WATER
RETURN TO SMELTER
Figure IV-21. Anaconda Arbiter Plant - Block Flow Diagram
using the mechanisms from a froth flotation machine which aerates the pulp and
increases the dissolution rate. The reaction in this stage is:
CuFeS2 + 4.25
4 NH3 + K20 = Cu(NH3>4
0.5
+ 250^' + 2 H
The leached residue and pregnant solution are discharged from the leach
tanks to a series of countercurrent decantation (CCD) thickeners. The purpose
of the CCD section is to separate the pregnant solution from the leached solids.
Wash water is added countercurrent to the flow of solids. The pregnant solu-
tion from the first CCD thickener is filtered in pressure filters while the
washed residue from the last thickener is pumped to a froth flotation section
for the recovery of unleached metal values. The concentrate from the flota-
tion section is pumped to the smelter for treatment. At Anaconda this stream
contains about 20% of the copper in the initial concentrate and most of its
precious metal values.
The clarified pregnant solution from the pressure filtration stage is
pumped to the fourth stage of the process, solvent extraction. During solvent
extraction, the pregnant solution is contacted with an organic reagent (General
Mills LIX 65N reagent) which selectively extracts copper from these solutions.
The loaded organic is washed to remove any entrained ammonia and is then fed to
the stripping side of the circuit where .the copper-loaded organic is stripped
with spent electrolyte from electrowinning . The stripped organic is returned
to the extraction side of the circuit. 'The loaded electrolyte is taken to a
conventional electrowinning stage where Copper in the electrolyte is plated on
97
-------
starter sheets of pure copper to produce copper cathodes according to the
reaction Cy*"1" + H^O = Cuฐ + 2H + 1/2 02- Insoluble anodes of antimonial
lead are used.
The ammonia recovery system regenerates ammonia for use in the leach.
The inputs to this system are the raffinate (the barren stream from the liquid
ion exchange system containing mainly ammonium sulfate) and exhaust gases from
the reactors, thickeners, and solvent extraction stages. Those streams which
contain ammonia compounds (mainly the raffinate) are treated in lime boil pots
to yield ammonia and calcium sulfate. The ammonia is stripped from the solu-
tion by steam flowing countercurrent to the slurry. The slurry from the last
lime boil pot is pumped to tailing storage while the stripped ammonia (and the
streams containing free ammonia) go to the ammonia fractionator column. The
fractionating system contains a fractionator, reboiler, condenser, reflux drum,
storage tanks and pumps. Vent gases from other parts of the plant are scrubbed
with water which is then fed to the ammonia fractionator.
The plant typically also includes auxiliary facilities such as a boiler
plant for making steam and an air separation plant for making oxygen used in
the leaching process.
2. Current Status
The Anaconda Arbiter plant is the first large hydrometallurgical process
to be placed in commercial operation in the United States for the hydrometallur-
gical extraction of copper from sulfide concentrates. This plant, located in
Montana, was built in 1973-1974 and has a capacity of 36,000 ton/yr of cathodes.
It was undergoing initial shakedown operations when the plant was mothballed
because of the low price of copper and the absence of feed materials. We
understand that the plant is also undergoing some modifications.
The concentrate used in Montana contains primarily chalcocite (Cu2S)
which is leached much more rapidly than chalcopyrite, the more abundant
mineral used as a basis for this study. Thus, the specific concentrate used
in Montana offers the following advantages:
The chalcocite can be leached rapidly, requiring smaller leach
reactors.
Because the precious metals are concentrated in the mineral enargite
(Cu-AsS^), which leaches slowly, the leach is conducted only to
dissolve approximately 80% of the copper. The solids from the
leaching step are treated by froth flotation to recover the remain-
ing copper sulfide minerals which are then smelted in a conventional
smelter. This enables an overall high recovery of copper (about
95-98%) as well as precious metals.
Because of the overall high copper-to-sulfur ratio in the leachable
fraction of the concentrate, the lime requirements in the ammonia
recovery section are much smaller than they would be for a standard
chalcopyrite concentrate.
The concentrate is also free of several troublesome trace impurities.
98
-------
In contrast to the above, chalcopyrite concentrate would require much
longer leach times, would require much higher copper recoveries in the leach
step, and would require a means for recovering precious metals if present in
the ferric oxide sludge, and a means for handling impurities. For the widest
applicability of this process, it might be desirable for the process not to
rely only on an adjacent smelter.
3. Effluents
Figure IV-22 shows the emission sources while Table IV-38 shows major
constituents of effluent streams from the Arbiter process by air, water, and
solid waste categories.
a. Air Pollution
The Arbiter process, like most hydrometallurgical processes, causes little
direct air pollution. The major emissions from the process are associated with
the ammonia recovery system, which would contain traces of ammonia and fugitive
emissions of sulfuric acid mist from the electrowinning tank house.
The process is energy-intensive and uses mainly electricity and steam.
Some pollution would be associated with the boiler plant in both cases.
COPPER
CONCENTRATE
OXYGEN NH3
IRON OXIDE
SLUDGE TO
TAILINGS POND
i
00ฉ
Ca(OH)2
^
LEACHING
\
1
[
COOLING
WATER (WO
STEAM
-^^
-
1
__ SOLID/LIQUID
*" SEPARATIONS
f ฉ
WASH WATER
RAFFINATE RECYCLE
SOLVENT
EXTRACTION
ฉ
I
SPENT
ELECTROLYTE
ฉ
ELECTRO-
WINNING
__ CATH
* COPF
l
LIME BOIL VEN3T
SYSTEM RECOVERY
,ฉ
NH3
FRACTIONATOR
0ฎ 1ฉ0 |
*- NH3
ป PROC
1 I ฉ
GYPSUM SLUDGE MISCELLANEOUS STEAM VENT GAS
FOR RECYCLE
TO TAILINGS POND
VENT GASES
CONTAINING
Figure IV-22. Sources of Emissions in the Arbiter Process
99
-------
TABLE IV-38
EMISSIONS FROM THE ARBITER PROCESS
Stream Stream Size Major Constituents
Air Pollution and Fugitives
A-l - NH, Vent Recovery System 10,000-50,000 scfm (Est.) Traces of NH3
ana Fractionator
F-l - Tank House Air NA* H2SC>4 Acid Mist
Water Pollution
W-l - Cooling Water Slowdown NA* Chlorides
Solid Wastes
S-l - Leach Residue 1.3 dry ton/ton of copper Iron oxide, insoluble minerals
such as silica, pyrites, bismuth
sulfides, lead, arsenic and
precious metals, if present
5-2 Gypsum Sludge 4.5 dry ton/ton of copper Calcium sulfate with traces of
arsenic, zinc and other heavy
metals
NA = Not Available
b. Solid Wastes
The two solid waste streams are the calcium sulfate sludge from the
ammonia recovery system and the iron oxide sludge from the leach residue.
These wastes would require storage in lined ponds with a wet top surface to
prevent dust emissions during the period the pond is in use and proper cover
after the pond is abandoned. The two sludges would contain impurities if
present in the concentrates. The behavior of impurities is discussed in the
next section.
c. Water Pollution
The Arbiter process can be operated so that there is no net aqueous
discharge to the environment in areas where solar evaporation is high, such as
Arizona, the location considered for this study. In areas where evaporation
is insufficient or with an excess rainfall, depending on the plant water balance,
it might be necessary to bleed some water from the tailings ponds. Treatment
necessary would be that typical for comparable waters.
100
-------
4. Technical Considerations
a. Products
The copper produced by the Arbiter process would be more or less equiva-
lent in quality to other electrowon copper. The sulfur values could either be
sold as ammonium sulfate fertilizer when and where market and transportation
conditions permit or be discarded as calcium sulfate sludge. In conventional
copper processing, the sulfur values are typically converted, to sulfuric acid
or discharged to the atmosphere. These options are not present for hydro-
metallurgy. The iron would be discarded as an iron hydroxide sludge. In all
probability, this sludge would be too impure to be considered as a feed to iron
and steel making. The major unknowns in the Arbiter process relate to its
ability to recover the precious metals from the iron oxide sludge by cyanida-
tion (dissolving precious metals in a solution of cyanide ions) and to the
behavior of trace metal impurities that are always present in the concentrates.
b. Impurities
Only limited data are available on the behavior of impurities during
processing. Available information indicates the following:
Arsenic - When the concentrates are high in iron, the arsenic is
precipitated as ferric arsenate. With concentrates low in iron
the arsenic is precipitated as calcium arsenate during the lime boil.
In the former case, arsenic would be present in the leach residue and
in the latter case, it would be present in the gypsum sludge, both
of which are ultimately disposed of to the land.
Cobalt and Nickel - These metals would be expected to dissolve in
the leach solution and build up in it unless they are removed by
precipitation or by ion exchange.
Selenium and Tellurium - Nothing is known at the present time. They
may remain in the leach residue.
Lead - Lead would be oxidized to lead sulfate or basic sulfate and
remain in the leach residue.
Zinc - Zinc is removed during the lime boil either as a hydroxide
or basic sulfate. Alternatively, zinc can be recovered by solvent
extraction/electrowinning, if present in concentrations such that
this incremental processing is economically justified.
o Precious Metals - Precious metals can be recovered in two ways.
At Anaconda's plant operations, only 80% of the copper will be
leached. This leachable fraction, chalcocite (Cu2S), is low in
precious metals. A slowly dissolving fraction, enargite (C^AsS^) ,
contains most of the precious metals. Thus, it is possible to
recover the unleached sulfides by a simple froth flotation step and
treatment of this flotation concentrate in the conventional manner
in the existing smelter. In general, it appears that when copper
extractions exceed 97%, the silver and gold recoveries by flotation
101
-------
are very low (under 25%). The silver and gold, however, can be
recovered by cyanide leaching of the residues. Alternatives would
be brine (NaCl) leaching. Anaconda claims that high silver and
gold extractions have been observed when "cyaniding" such residues
and that cyanide losses are related to the amount of copper remain-
ing in the residues rather than to the presence of hydrated ferric
oxides. However, no data have been released because Anaconda
considers this to be proprietary information.
Bismuth - Bismuth, present as bismuthinite, is converted either to
a heavily oxidized colloidal precipitate or to a sulfide with
oxidized surfaces during the ammonia leaching process. In either
case, this bismuth is not floated during subsequent flotation of
the leach residue and reports to the leach tailings fraction.
The process should be able to handle copper scrap as long as its particle
size permits. The leaching of copper scrap in similar solutions has been
practiced on a commercial scale and presents no obvious technical or
environmental problems.
5. Economic Factors
a. Basis for Comparison
The hydrometallurgical processes produce cathode copper usually equivalent
in quality to electrorefined cathode copper. Thus, they have to be compared
with conventional smelting plus refining. These processes claim the following
advantages (some of them still unproven):
They have a much smaller minimum economic size. The smallest
economic size for a conventional smelter is at least 100,000
ton/yr. It is even larger for an electrolytic refinery. A hydro-
metallurgical plant could be viable in the range of 20,000-40,000
ton/yr of copper.
The capital investment per unit of output is about the same. Large
smelters and refineries cost about $1200/annual ton. The smaller
hydrometallurgical plants would cost about $1000-1200/annual ton.
The hydrometallurgical processes are truly continuous processes as
opposed to the batch/continuous processing inherent in conventional
pyrometallurgical processing. This means that the hydrometallurgical
processes can be designed to have a significantly lower labor
component compared with the base line pyrometallurgical route.
In a hydrometallurgical plant, the sulfur in the concentrate is
converted to forms other than sulfuric acid, such as (NH4>2 $04, or
CaSO^ sludge. This is advantageous in locations where air pollution
regulations are very stringent.
102
-------
All the above suggests that the applications of these processes are
likely to be in locations where sulfuric acid markets are non-existent and
where the construction of a full-sized smelter represents an undue economic
risk because the growth in copper demand is sluggish. Because of this it
would not be proper to compare a smaller-sized Arbiter plant with a conventional
smelter-refinery complex sized at 100,000 ton/yr.
A comparison with conventional technology is feasible when custom smelting
is taken into account. A custom smelter offers smelting and refining services
to producers of concentrates from several sources. It is able to achieve
economies of scale by pooling of concentrates. Typical custom smelting and
refining charges at the present time using conventional technology would be
about 18<:/lb copper. A comparison of this figure with those in Tables IV-6 and
IV-8 shows that the custom smelting and refining charge is less than the cost
of smelting and refining in a new installation. This is because existing
custom smelters and refineries are old and have lower fixed costs while infla-
tion has substantially increased the capital investment requirements for new
facilities.
Another problem specific to the comparison of the Arbiter process with
the base line is related to the fact that the initial application of the
Arbiter process is not to the usual chalcopyrite concentrate but to a chalcocite-
enargite concentrate.
Nevertheless, for our comparison, we have used a clean chalcopyrite
concentrate as a basis because this is the most abundant copper concentrate in
the United States and in the world. This choice makes the comparison less
favorable for the Arbiter process because of the higher sulfur content, longer
leach times, and higher lime requirements for neutralization. Consistent with
industry practice and the base case, credit for precious metals and/or impurities
has been ignored. However, it appears that with the Arbiter process, the
incremental costs of treating leach residues for precious metal recovery
would be higher than the incremental cost of recovering the same metals in
conventional processing from anode mud residue.
Finally, it should be noted that a comparison of conventional smelting
and refining versus the Arbiter process cannot be made at constant or equiva-
lent emissions to the environment but rather can be made on the basis of both
processes meeting currently acceptable environmental standards (e.g., BATEA,
'BPCTCA and NSPS).
b. Capital and Operating Costs
Table IV-39 shows estimates of capital and operating costs for the Arbiter
process based on chalcopyrite concentrates.
c. Comparison with Conventional Technology
(1) Energy Use
Energy used for this process (mainly electricity and steam) amounts to
about 56 x 10*> Btu/ton (fossil fuel equivalent) according to Arbiter (1974).
103
-------
TABLE IV-39
OPERATING COSTS: THE ARBITER PROCESS
(Basis: Chalcopyrite concentrates 36,000 ton/yr of copper)
CAPITAL INVESTMENT (CI)
OPERATING COSTS
VARIABLE COSTS
Ammonia
Oxygen
CaO
LIX and Organic
Miscellaneous Chemicals
Purchased Steam (from Fuel Oil)
Purchased Electricity
Water: Process
Cooling
Direct Operating and
Maintenance Labor (L)
Supervision (S)
Maintenance Materials
Labor Overhead
TOTAL VARIABLE COSTS
FIXED COSTS
Plant Overhead
Local Taxes & Insurance
Depreciation
Capital Recovery
TOTAL FIXED COSTS
TOTAL COSTS
Unit
S/ annual ton
ton
ton
ton
gal
103 Ib
kWh
103 gal
103 gal
Man-hr
L
CI
L+S
L+S
CI
CI
CI
S/'Jnit
150.00
15.00
25.00
4.00
3.06
0.021
0.20
0.05
5.75
152 L
4% CI
352 (L+S)
65% (L+S)
27. CI
7% CI
202 CI
Units/Ton
$1000
0.15
2.5
2.3
3.4
20.00
3000.00
1.6
10.0
7.00
S/Ton Cu
22.50
37.50
57.50
13.60
15.00
61.20
63.00
0.32
0.50
40.45
6.04
40.00
16.27
373.88
26.65
15.00
70.00
150.00
261.65
635.53
104
-------
On the same basis, he reports 55.4 x 106 Btu/ton for conventional processing
and 30.4 x 10 Btu/ton for flash smelting. Out figures for the pyrometal-
lurgical processes are lower.
The hydrometallurgical processes are inherently less energy efficient
than the newer pyrometallurgical processes because in pyrometallurgical
operations the heat of oxidation of sulfur and iron is utilized to melt the
concentrates. In the Arbiter process, sulfur and iron are oxidized at about
200ฐF. This heat has to be carried away from the leach reactors using cooling
water. This low-grade heat cannot be utilized in the plant and has to be
dissipated.
(2) Pollution Aspects
As mentioned earlier, there is considerable alteration in the effluent
profile with the Arbiter process compared to the base line. Air pollution
is reduced to very low levels. New forms of solid wastes (leach residue and
calcium sulfate sludge) are produced. These, along with the associated liquid
effluent, would contain several of the trace impurities in the concentrates and
would require proper storage. In our calculations, we have assumed storage
ponds with water recycle for both tailings. If a water stream has to be bled,
we have assumed that the overflow from the calcium sulfate pond could be
discharged. Depending upon evaporation/rainfall ratio this discharge might be
eliminated.
Since hydrometallurgical processing, by its very nature, takes place in
closed containers, the emissions from a plant of this type are at specific
points which (at least conceptually) might be easier to control than the
emissions from several of the pyrometallurgical operations in conventional
smelting. Thus, the hydrometallurgical plant might be in a better position
to weather the changing regulations regarding fugitive emissions and trace
metal emissions than the conventional plant.
The costs of solid waste disposal are about $33.06/ton of copper as
shown in Table IV-40.
(3) Byproducts
For most concentrates it might be economical to treat the iron oxide
sludge via cyanidation for precious metal recovery prior to its disposal
in ponds. In special situations, it might be possible to sell ammonium
sulfate for fertilizer use (assuming that trace impurities can be removed or
demonstrated to be non-deleterious, depending on their nature and the crop
to which the fertilizer is applied).
(4) Cost Comparison
Table IV-39 shows that direct operating costs are about l8/lb and total
operating costs are about 32/lb. These costs are for chalcopyrite
concentrates and, as mentioned earlier, the process economics are more favor-
able with chalcocite concentrates. These costs are not competitive with
custom smelting and refining charges of about 18/lb or even smelting and
refining costs in a new large smelter and refinery of about 25/lb. This
105
-------
TABLE IV-40
ARBITER PROCESS - SOLID WASTE DISPOSAL
36,000 ton/year of copper
Lined Disposal Sites - covered each year - lined with membrane
Required: 14.5 acre/yr - 10 ft deep
Estimated Yearly Investment (Expenditure)
In Site Preparation and Closure $ 450,000
Estimated Hauling Costs @ $2.00/Wet ton 740,000
Total Estimated Annual Costs $1,190,000
Unit Costs ($/ton copper) $ 33.06
would suggest that the Arbiter process .(and perhaps hydrometallurgical processes
generally) are applicable only for special concentrates, such as the chalcocite
concentrates being utilized by Anaconda, in cases where custom smelting
capacity is either not available or where the required pyrometallurgical plant
is too small to be economical. This would explain why much of the activity in
hydrometallurgical process development in the United States has been undertaken
by the smaller, non-integrated mining companies such as Cyprus and Duval.
(5) Commercial Considerations
Users of the process would expect to pay a royalty to Anaconda for the
process or operate under some sort of licensing agreement. These costs are
not included in Table IV-39. It should also be noted that hydrometallurgical
processes typically have a smaller inventory of metal in the process compared
to smelting. However, because the process operations are more sensitive to
concentrate mineralogy, more extensive blending and stockpiling of feed
materials might be necessary.
106
-------
V. IMPLICATIONS OF POTENTIAL PROCESS CHANGES
A. INTRODUCTION
The process changes considered by us for detailed analysis fall into the
following four categories:
New pyrometallurgical processes
The use of oxygen in smelting
Metal recovery from slags (electric furnace and flotation)
Hydrometallurgy (Arbiter process)
In this chapter we discuss the implications of the process changes
discussed in Chapter IV grouped into the four categories mentioned above.
B. PYROMETALLURGICAL PROCESSES
Of the process units used in conventional smelting, the reverb has become
obsolescent because of high energy costs and because of the need to curb
emissions of S02 to the atmosphere. The new pyrometallurgical processes use
smelting units which produce concentrated S02~containing streams. Such streams
are suitable for the manufacture of sulfuric acid or for the reduction to
elemental sulfur. This makes it possible to increase the degree of sulfur
capture from 50-70% with conventional smelting (achieved via control of con-
verter or converter plus roaster gases) to over 90% (achieved via control of
both smelting unit and converter gases).
While sulfur capture prevents the emissions of sulfur dioxide to the
atmosphere, this captured sulfur has to be utilized or disposed. In the United
States, the major copper deposits and the copper mining, beneficiation, and
smelting activities are mainly in the West, and are concentrated in particular
in the .Southwest. These locations are distant from the traditional sulfuric
acid market (Florida). The cost of shipping this acid to the traditional
markets outside the Southwest is prohibitively expensive; i.e. about $40/ton
of acid for transportation alone versus about $20-30/ton for acid production.
Thus local markets must be found or the acid must be disposed of in an alternate,
acceptable fashion. The transportation constraints have resulted in the
availabiltiy of low-cost acid in the West. Two trends have developed as a
result of the availability of this low-cost acid. The first is the construction
of several wet-process phosphoric acid plants based on smelter acid and the
lower grade western phosphate rock. The second trend is the increasing use
of sulfuric acid to recover copper from such marginal resources as tailings,
mine dumps, and oxide ores. Since many of these ores contain high concentra-
tions of acid-consuming materials, this approach not only disposes of the acid
107
-------
by neutralizing it, but also recovers copper at the same time. Because of the
arid nature of the Southwest, dumps and surface deposits can be leached with-
out contamination of groundwater. This would not be possible in other regions
of the United States. Alternative ways of disposing of the sulfur values as
calcium sulfate are also being examined (neutralization of sulfuric acid or
direct roasting of concentrates with limestone to fix the sulfur). In the
future it is conceivable that sulfuric acid would still be in surplus in the
West and that neutralization with limestone would be quite expensive. In that
case it might be more economical to construct smelters closer to the traditional
acid markets east of the Mississippi and transport the ore concentrates by rail
to such locations for smelting.
A few years ago, a custom smelter considered this approach. However, none
of the independent mines were willing to participate since smelting charges
from the new smelter would have been quite a bit higher than the charges from
existing, partially depreciated smelters.
It is possible to convert concentrated SC>2 streams to elemental sulfur
instead of converting them to sulfuric acid. Natural gas, light hydrocarbon
liquids, or coal can be used, depending on the process selected. The elemental
sulfur produced in this fashion can be shipped a longer distance than the
corresponding sulfuric acid, or such sulfur can be stored. The cost of pro-
ducing such sulfur is quite high, exceeding about 5c/lb of copper contained in
the concentrates, versus a pollution control cost of about 3/lb of copper when
making sulfuric acid. It therefore appears that elemental sulfur would be
considered only in cases where all other options are not feasible.
Compared to conventional smelting, the newer smelting processes are quite
energy efficient and save from 30 to 50% of the energy over conventional smelt-
ing. Since both conventional and newer processes are quite flexible in their
use of any form of fossil fuel, there is no additional conservation in form
value. The adoption of these new processes not only reduces emissions to the
atmosphere but also reduces fuel requirements and the pollution that is gener-
ated from the fuel.
The major handicap of the new pyrometallurgical processes is that their
ability to handle impure concentrates (concentrates containing As, Sb, Bi, Zn,
Pb, Se, Te, etc.) has not been proven. In many cases when impurity concentra-
tions are moderate (say, under 5% total impurities), impurity elimination is
possible by making a matte in the new smelting unit and using a batch convert-
ing operation to volatilize these impurities. When impurity concentrations are
higher, it is necessary to roast the concentrates and to smelt the calcines in
these new smelting units. This technology is also unproven.
In general, these new processes are capable of recycling less scrap than
with conventional processing. However, several techniques (oxygen enrichment,
use of coke in converters, etc.) are available. The use of these techniques
can enable the new processes to recycle comparable quantities of scrap.
108
-------
Pyrometallurgy, in general, offers signficant economies of scale.
Thus, another handicap of the new pyrometallurgical processes is that they
too need to be constructed at a large size. The smallest economic size is
approximately 100,000 ton/yr.
The growth rate in copper consumption in the United States has been 3-4% a
a year. Assuming that this growth rate continues and that U.S. metal produc-
tion is increased to keep up with this (as has been the trend in the past),
the United States would require new capacity of about 600,000 ton/yr each year
to fulfill these requirements by 1985-1990. We believe that some of this
increased production will come from hydrometallurgical operations on oxide
and sulfide ores and the number of new pyrometallurgical smelters might be
3-4 at over 100,000 ton/yr each. We believe that these smelters would utilize
the newer processes or possibly electric smelting (if favorable power costs
can be negotiated or if the impurity content of the concentrates is high).
These new pyrometallurgical processes are economically viable only if the
metal contained in the slag is recovered. The slag treatment processes are
an integral portion of the new smelting technology and their costs have been
included in considering the process economics. Their costs have been estimated
separately in Section G for illustrative purposes.
C. OXYGEN USE IN SMELTING
The use of oxygen in smelting decreases fuel requirements and is energy-
efficient overall because the decrease in fuel requirements is usually larger
than the incremental energy required in the oxygen separation plant. The most
beneficial aspect of using oxygen is the ability to increase production signi-
ficantly from existing units and reducing the capital costs per unit of
output from new units. A smaller benefit is the ability to melt more scrap.
However, most of the easily available scrap is already recycled, and the
marginal scrap is probably mixed copper-iron scrap which is smelted in limited
quantities in existing smelters. These advantages would provide the major
driving force for the widespread adoption of the use of oxygen enrichment in
industry for both existing and new processes.
D. RECOVERY OF METALS* FROM SLAG
As mentioned earliers the availability of these processes has made it
possible to use the new smelting technology such as Outokumpu flash smelting.
The additional benefits of slag cleaning, with less flux added for control of
viscosity and fluidity, are that the total slag volume and flux requirements
are decreased. Since converter slag is not recycled to the smelting unit,
problems associated with accretions of magnetite are avoided. This makes it
possible to operate longer campaigns in the smelting unit without the loss of
capacity that occurs when magnetite deposits on the bottom. The elimination
of converter slag recycle also gets rid of the converter slag launder which is
a major point for air infiltration into the smelting unit. This air infiltra-
tion is undesirable because it increases fuel requirements and dilutes the S02
in the smelting unit gases.
109
-------
These processes are also applicable in conventional smelting; for
example, converter slag can be treated via slag flotation instead of returning it
to the reverb.
E. THE ARBITER PROCESS
The Arbiter process, and hydrometallurgical processes in general, would be
adopted mainly in regions distant from sulfuric acid markets and where a full-
sized pyrometallurgical smelter is risky. The adoption of hydrometallurgical
techniques would depend on the availability of surplus custom smelting capacity
in the United States. Overall, we believe that 3-5 small hydrometallurgical
plants will be operating in the 1980's, particularly on concentrates with
favorable mineralogy. These processes have little or no air pollution, but
produce significant quantities of solid waste which are different from the
conventional solid waste from the smelter. These wastes can be stored in lined
ponds so that they have only a minimal interaction with the environment.
F. SUMMARY
The copper industry can be expected to increase metal production by about
600,000 ton/yr by 1985-1990 if past trends continue. A major portion of this
increase will result from the adoption of new pyrometallurgical processes with
or without oxygen enrichment but with associated slag cleaning technology.
These processes would result in a 30-50% reduction in energy usage per ton
of copper. The capital costs for these smelting processes are about $750/annual
ton of copper and operating costs would be about 15-l?ฃ/lb copper. This
,technology would increase sulfur capture from the current level of 50-70% to.
over 90%.
110
-------
REFERENCES
Anon., "Outokumpu Flash Smelting and Its Energy Requirement," Symposium on
Efficient Utilization of Fuels.
Arbiter, N., "Anaconda's Ammonia Leach Process," AIME Meeting, Dallas 1974.
Arbiter, N. , "New Advances in Hydrometallurgy in Efficient Use of Fuels in the
Metallurgical Industries," IGT, 1974.
Bailey, J.B.W. , A.J, Weddick, and G.D. Hallett, "Oxygen Smelting in the
Noranda Process," AIME New York Meeting, 1975.
Banks, C.C., "Recovery of Non-Ferrous Metals from Secondary Copper Smelting
Slags." Advances in Extractive Metallurgy & Refining, IMM, London 1972.
Battelle Columbus Laboratories, for USBM, "Energy Use Patterns in Metallurgical
and Nonmetallic Mineral Processing," (Phase 4 - Energy Data and Flowsheets,
High-Priority Commodities), PB-245 759, June 1975.
Byrk, P., et al, "Flash Smelting of Copper Concentrates," Outokumpu Oy,
Finland, undated.
Dasher, J., "Hydrometallurgy for Copper Concentrates?," CIM Transactions,
76, 1973.
Edlund, V.E., and S.J. Hussey, "Recovery of Copper From Converter Slags by
Flotation," USBM RI #7562 (Revised).
Foard, J.E. and R.R. Beck, "Copper Smelting Current Practices and Future
Developments-," AIME Annual Meeting, February 1971, New York, N.Y.
Hayashi, M. et al, "Cost of Producing Copper from Chalcopyrite Concentrate as
Related to S02 Emission Abatement," USBM RI #7957.
Harkki, S.U. and J.T. Juusela, "New Developments in Flash Smelting," The
Met. Soc. of AIME, New York, 1974 (A74-16).
Itakura, K. et al., "Converter Slag Flotation Its Effect on Copper
Reverberatory Smelting Process," Journal oZ Metals, July 1969
Juusela, et al, "Outokumpu Flash Smelting and Its Energy Requirement,"
Efficient Use of Fuels in the Metallurgical Industries, IGT, 1974.
Kellogg, H.H., "Prospects for the Pyrometallurgy of Copper," Proceedings of
the Latin American Congress on Mining and Extractive Metallurgy, Santiago,
Chile, August 1973.
Ill
-------
Kuhn, et al., "Anaconda's Arbiter Process for Copper," CIM Bulletin,
February, 1974.
Arthur D. Little, Inc., for US EPA, "Economic Impact of New Source
Performance Standards on the Primary Copper Industry An Assessment,"
(EPA-68-02-1349), October 1974.
Arthur D. Little, Inc., for US EPA, "Economic Analysis of Proposed Effluent
Guidelines - Nonferrous Metals Manufacturing Industries," (EPA-230/1-75-041),
March 1975.
Lukkarinen, T., "How Outokumpu Recovers Copper from Smelter Slag at Harjavalta,"
World Mining. July 1971, pp. 32.
Lukkarinen, T., (Editor), "The Concentrators of Outokumpu Oy Finland."
Outokumpu Oy, Helsinki, Finland 1972.
Mackey, P.J., G.C. McKerrow and P. Tarassoff, "Minor Elements in the
Noranda Process," Paper presented at 104th Annual Meeting AIME, New York,
New York, February 16-20, 1975.
Matheson, K.H., et al., "Emission Control Effort at Kennecott's Utah Smelter,"
Paper presented at the 78th AIChE National Meeting, Salt Lake City, Utah,
Auguts 18-21, 1974.
"Mitsubishi's Continuous Copper Smelting Process Goes On-Stream," Engineering
and Mining Journal, August 1972, pp. 66.
Parameswaran, K. and R. Nadkarni, "Energy Considerations in Copper, Lead &
Zinc Smelting," Energy Use and Conservation in the Metals Industry, AIME, 1975
Price, F.C., "Copper Technology on the Move," Chemical and Engineering News,
April 1973.
Ruddle, R.W., "The Physical Chemistry of Copper Smelting," IMM (London), 1953.
Ruddle, R.W., B. Taylor, and A.P. Bates, "The Solubility of Copper in Iron
Silicate Slags," IMM (London), 1966, 75 pp. Cl.
Schnalek, F., and I. Imris, "Slags from Continuous Copper Production,"
Advances in Extractive Metallurgy and Refining, M.J. Jones Ed., IMM (London)
1972.
Schwartz, W.H., "Economics of Pyrometallurgical Copper Extraction Process,"
AIME 1975 (A75-57) (Frankfurt, Germany), LURGI CHEMIE.
Sharma, SปN., et al, "Process Analysis and Economics of Flash Technology,"
Journal of Metals, August 1975, p. 7.
Spira, P. and N.J. Themelis, "The Solubility of Copper in Slags," Journal of
Metals, April 1969, pp. 35.
Subramanian, K..N. and N.J. Themelis, "Copper Recovery by Flotation," Journal
of Metals, April 1972, pp. 33.
112
-------
Subramanian, K.N., and N.J. Themelis, "Recovery of Copper from Slag by
Milling," Journal of Metals, April 1972
Suzuki, T. et al., "Behavior of Impurities in Mitsubishi Continuous Copper
Smelting and Converting Process," TMS Annual Meeting, 1975.
Suzuki, T. et al., "Computer Control in Mitsubishi Continuous Copper Smelting
and Converting Process," TMS Annual Meeting, 1974.
Themelis, N.J., and G.C. McKerrow, P. Tarassoff, and G.D. Hallett, "The
Noranda Process," Journal of Metals, April 1972.
Themelis, N.J., and G.C. McKerrow, "Production of Copper by the Noranda
Process," Advances in Extractive Metallurgy and Refining, IMM London, October
1972, paper 7.
Treilhard, D.G., "Copper - State-of-the-Art," Chemical Engineering, April 16, 1973.
Weddick, A.J., "The Noranda Continuous Smelting Process for Copper,"
Efficient Utilization of Fuel, Symposium of Institute of Gas Technology,
Chicago, Illinois, December 9-13, 1974.
113
-------
APPENDIX A
PRESENT COPPER TECHNOLOGY
1. FEEDSTOCKS
There are seven major copper producing areas in the world: (1) the
western United States; (2) the western slope of the Andes in Peru and Chile;
(3) the central African Copperbelt in Zambia and Zaire (Kinshasa) ; (4) the.
Ural Mountains and the Kazakstan region in the U.S.S.R. ; (5) the Precambrian
area of central and western Canada; (6) the Keweenaw Peninsula of Northern
Michigan; and (7) the Southwest Pacific (Australia, Bougainville).
Of the many copper minerals, only chalcocite, chalcopyrite, bornite,
chrysocolla, azurite, and malachite are important commercially. Copper ores
occur in many types of deposits in various host rocks. Porphyry copper
deposits account for about 90% of the U.S. production and much of the world
output, and contain most of the estimated commercial copper reserves of the
world .
From a processing viewpoint, copper ores can be classified in three
categories: sulfide ores, native copper ores, and oxide ores.
A sulfide ore is a natural mixture containing copper-bearing sulfide
minerals, associated metals, and gangue minerals (e.g., pyrites, silicates,
aluminates) that at times have considerable value in themselves (e.g.,
molybdenum, silver, gold, as well as other metals). Most sulfide ores belong
to one of three major groups, all of which are represented in the United States,
namely:
The porphyry copper and Northern Rhodesian type deposits that carry
copper mostly in the form of chalcocite (Cu^S) , chalcopyrite
and bornite (C^FeS^) . Copper ranges from a fraction of one percent
to several percent, and iron is generally low. The deposits in
the southwestern United States are of this type.
Deposits, such as those found in Rio Tinto in Spain, in Cyprus, and
in Tennessee, commonly known as cupriferrous pyrite, which generally
have 1-3% copper as chalcopyrite, and contain abundant amounts of
pyrite and pyrrhotite. Generally, copper-to-iron ratios and copper-
to-sulfur ratios are low.
114
-------
Arsenic-bearing copper ores, such, as enargite (C^AsS^ , with
deposits occurring in Butte, Montana; Yugoslavia; Tsumeb in
South West Africa; and the Philippines.
The sulfide ores are treated primarily by crushing, grinding, and froth
flotation to produce a concentrate (or several concentrates) of sulfide minerals
and reject the worthless gangue as tailings.
Native copper ores are those in which some of the copper occurs as the
native metal. The Lake Superior District in Michigan is the only major source
of ore of this type. Although the reserves of this ore are quite extensive, it
contributes only a small portion of the total U.S. mine production of copper.
All non-sulfide, non-native ores of copper are termed "oxide" ores, the
oxide copper content being measured by and synonymous with solubility in
dilute sulfuric acid. An oxide copper ore can contain copper oxide, silicate
or carbonate minerals and gangue. The oxide ores have been treated metallurgi-
cally in a variety of ways, the character of the gangue minerals having a very
important bearing on the type of metallurgical treatment used. Oxide ores in
the United States are treated primarily by leaching with dilute sulfuric acid.
Commonly associated with copper are minor amounts of gold, silver, lead,
and zinc, the recovery of which can improve mine profitability. Molybdenum,
lead and zinc are recovered as sulfides by differential flotation. Minor
amounts of selenium, tellurium, and precious metals are extracted in electro-
lytic refining. On the other hand, arsenic, antimony and bismuth in the ores
cause problems in standard pyrometallurgical processing and electrorefining,
and thus their presence is a cost penalty. Nickel and cobalt can interfere with
electrolytic refining, but they do not occur in significant amounts with the
U.S. copper deposits.
In 1964, the Bureau of Mines reported domestic reserves of 75 x 10 tons
of metal in ore averaging 0.86% copper, assuming recovery at 90% of gross
metal content. An additional 58 x 10^ tons of copper was estimated as
potential resources recoverable with technological or economic improvements.
Arizona, Montana, Utah, New Mexico, and Michigan accounted for more than 90%
of the total reserves.
A 1973 study*/estimated the total known domestic resources of copper
economically available at various copper prices, allowing for a 12% return on
investment: .
Resources
Price (106 ton)
$2.00/lb 180
0.75/lb 115
0.50/lb 83
*IC 8598, "Economic Appraisal of the Supply of Copper," U.S.B.M., 1973
115
-------
The 83 x 10^ tons of reserves indicated above represent 49 years of
supply at our present production rate of about 1.7 x 10 ton/yr.
A comparison of U.S. copper resources with those of the rest of the world
(see Table A-l) indicates that the United States has about 20% of the world's
copper resources. It is also evident, however, that many areas of the world
have significant copper resources. The major resources are in South America,
Africa, U.S.S.R., Canada, Mexico, and Europe.
2. PROCESSING
a. Mining
About 85% of the total copper ore mined comes from open pits; the rest
comes from underground mines. Underground mining methods for copper ores
involve caving and/or cut-and-fill mining.
b. Beneficiation
The different mineral forms (sulfides, carbonates, oxides, native copper,
etc.) require different processing techniques. Many methods have been used to
beneficiate the ores; generally only the sulfide ores are amenable to concentra-
tion procedures such as grinding and froth flotation.
TABLE A-l
IDENTIFIED AND HYPOTHETICAL COPPER RESOURCES
(106 Tons)
Area Identified 1 Hypothetical 2
United States:
Eastern 10 5
Western, except Alaska 64 75
Alaska 2 20
Canada 19 50
Mexico 18 20
Central America 1 6
Antilles 2 1
South America 80 50
Europe, excluding U.S.S.R. 25 20
Africa 53 50
U.S.S.R. 39 50
Middle East - South Asia 4 20
Chins 3 ?
Oceania, including Japan 21 30
Australia 3 3
Total 344 400
1. Identified resources: Specific, identified mineral deposits that
may not be evaluated as to extent and grade and whose contained
minerals may or may not be profitably recoverable with existing
technology and economic conditions. Based on all categories of
reserve figures plus estimates where no figures are available.
Amounts are tentative and accuracy will be refined in subsequent
publications.
2. Hypothetical resources: Undiscovered mineral deposits, whether
of recoverable or subeconomlc grade, that are geologically pre-
dictable as existing in known districts. Based generally on inden-
tified resource figures times a factor assigned according to
geologic favorabillty of the region, extent of geologic mapping,
and exploration.
Source: Geological Survey professional paper 820, "United
States Mineral Resources," Brobst and Pratt, 1973.
116
-------
(1) Sulflde Ores
These ores, the most important source of copper, are concentrated by froth
flotation. This procedure requires crushing and grinding and classification to
about 100 mesh or finer to liberate the particles. Grinding is usually the largest
cost item in the process. After grinding, the ore-water mixture is treated with
reagents to condition the sulfide particles so that their surfaces become air-
avid. The sulfides are then collected with the froth produced in the flotation
cells. The final concentrate may contain 11% to 32% copper. Typically flotation
is used to separate copper sulfides from pyrite, recover molybdenum from copper
concentrate, and recover copper concentrate from complex lead-zinc-copper ore.
A typical flowsheet for flotation of a sulfide ore is shown in Figure A-l.
(2) Oxide Ores
Oxide ores occurring in the United States are generally not amenable to
flotation, but are generally soluble in various leaching solutions.
Acid Leaching
The ore is properly sized, if necessary, and leached with acid, which dis-
solves the copper.. Depending on ore grade and characteristics, the ore is leached
in vats (by percolation or with agitation), in heaps, or in place.
Sulfuric acid is the usual solvent for oxidized copper minerals. The presence
of ferric sulfate in the leach solution can solubilize some sulfide minerals such
as chalcocite. For dissolution of the oxide minerals, about 1.5 Ib of acid/lb
of contained copper is required, but total consumption of acid is often much
greater because of reaction with gangue minerals.
Copper is recovered from dilute leach solutions by precipitation with scrap
iron, and from concentrated leach solutions by electrowinning.
Other minor methods include ammonia leaching, cyanide leaching, the segrega-
tion process, and oxide ore flotation.
(3) Mixed Ores
The treatment of mixed ore, that is, ore containing both sulfide and oxide
minerals, depends on the proportions of the two types of minerals. If sulfides
predominate, flotation is used, with reagents that favor flotation of oxide
minerals. When the ore contains almost equal amounts of sulfide and oxide
minerals, combinations of leaching and flotation are used.
c. Smelting
Because most of the U.S. copper is extracted from low-grade sulfide ores
that require concentration, current pyrometallurgical practice for recovery of
copper from its sulfide concentrates is fairly uniform from smelter to smelter
and is adapted to treating fine grained sulfide concentrates consisting mainly
of copper and iron sulfides and gangue.
117
-------
ELECTRO
MAGNETIC
SURGE BIN
BELT CONVEYOR
c
r
i
L
... .,
- 1 1
FINEOREBIN
I-
I 1 SHORT HEAD
< '"-a-*1 CONE CRUSHER
l
oU-J, '
L-J L J
ADDITIONAL GRINDING AND FLOTATION CIRCUITS
DEPENDING ON TONNAGE
r n
L.-I
I l
h-l
H
h-l
i i
H
I l
t-^
1--I
*
n
!-i
h-ซ
uJ
ti
t--
j-H
M
H
L J
n
I-H
J-i
r-l
L J
BALL MILL
-I CLASSIFIER
THICKENER
Q SAMPLER
"i "-TAILINGS
CONCENTRATES
DISC FILTER
Figure A<-1, Typical Flowsheet - Sulfide Copper Ore Flotation
118
-------
Copper's strong affinity for sulfur and its weak affinity for oxygen as
compared with that of iron and other base metals in the ore forms the basis for
the three major steps in producing copper metal from sulfide concentrates;
roasting, smelting and converting.
(1) Raw Materials
Flotation concentrates containing from 15-30% copper constitute the bulk of
the feed to the smelters. In addition, smelters will charge cement copper
(produced by acid leaching of oxide ores and precipitation with iron) containing
70-85% copper, siliceous flux and limestone and a quantity of direct smelting ore
containing 4-8% copper. This type of ore, when available, functions as a source
of copper as well as a flux.
(2) Drying
The flotation concentrates received by the smelter are in the form of a wet
filter cake and can contain 10-15% moisture. Cement copper can contain as much as
30% moisture. The charge to a reverberatory furnace can be dried so that its
overall moisture content is 4-8% without unduly increasing dusting problems
in the reverb. The removal of moisture in drying reduces the fuel requirements
in the reverb. Also, the drier acts as a blender for homogenizing the charge.
Rotary or multiple hearth driers are used for drying the feed materials.
(3) Roasting
About half the copper smelters in the United States roast their charge
prior to feeding in the reverberatory furnace. The older smelters use multiple
hearth roasters for this purpose while the new smelters use fluidized bed
roasters.
The object of roasting copper sulfide ores and concentrates is to regulate
the amount of sulfur so that the material can be efficiently melted and to remove
certain volatile impurities such as antimony, arsenic, and bismuth. However, in
modern practice, the grade of the concentrate produced from some sulfide ores is
sufficiently controlled at the concentrator to eliminate roasting prior to rever-
beratory smelting. In the case of custom or toll smelters, the composition of
feed materials can vary widely. Hence, roasting is practiced to blend and control
the sulfur content of the charge.
Elimination of some of the sulfur in roasting results in a higher grade matte
in the reverberatory furnace and hence decreases the oxidizing load on the con-
verters. Sulfide roasting is autogeneous and additional fuel is not required. The
charging of hot roasted calcines into the reverberatory furnace can decrease its
fuel consumption per ton of charge by about 40% and consequently increase reverb
capacity. In addition, roasting also reduces the emissions of sulfur dioxide from
the reverb. The reason for this is as follows: the two major constituents of a-
centrates utilized by almost all the U.S. smelters are chalcopyrite (CuFeS2) ?~
pyrite (FeS2). These minerals contain sulfur that is loosely held or "labile"
which is given off by melting the minerals.
2CuFeS2 = Cu2S + FeS + S
FeS2 = FeS + S
119
-------
CuฃS and FeS form matte, whereas the labile sulfur reacts with oxygen in the
reverb gases to form SC>2. Removal of the labile sulfur during roasting can reduce
emissions from the reverb. Also the lower fuel requirement per ton of charge
when using calcine smelting reduces the volume of reverb off-gases.
Both types of roasters (multiple hearth and fluidized bed) usually operate
around 1200ฐF. Sulfur dioxide concentration in the wet off-gas is usually 2-10%
with multiple hearth roasters because of dilution with air. With fluid bed
roasters the wet off-gases can run 12-14% sulfur dioxide. Both types of roasters,
therefore, can produce a steady stream of relatively rich off-gases suitable for
sulfuric acid manufacture after cooling and dust removal. Both types of roasters
involve handling and collecting of large quantities of hot abrasive dust, which
can lead to high maintenance costs.
(4) Reverberatory Furnace Smelting
Roasted and unroasted materials are melted after mixing with suitable fluxes
in reverberatory furnaces. Liquid converter slag is also charged into the rever-
beratory furnace to recover its copper content. Heating of the charge is accom-
plished by burning fuel in the furnace cavity, the heat being transmitted to the
charge primarily by radiation from the roof, walls and flame.
Almost all the reverbs in the United States use natural gas as a fuel and
only one plant uses powdered coal. Because of the impending domestic shortage
of natural gas, most smelters are now installing facilities to burn alternate
fuels. The maximum smelting capacity of a reverb is limited by the amount of
fuel that can be burned (a function of reverb shape and size) and the quantity
of heat required by a unit weight of charge. Reverb throughput can be increased
by drying the charge, preheating the charge by roasting and preheating the
combustion air.
In the reverberatory furnace, copper and sulfur form the stable copper
sulfide (Cu2S). Excess sulfur unites with iron to form a stable ferrous sulfide,
(FeS). The combination of the two sulfides, known as matte, collects in the
lower area of the furnace and is removed. Such mattes may contain from 15-50%
copper, with 40-45% copper content being most common, and impurities such as
sulfur, antimony, arsenic, iron, and precious metals.
The remainder of the molten mass, containing most of the other impurities
and known as slag, being of lower specific gravity, floats on top of the matte
and is drawn off and discarded. Slags in copper smelting are ideally represented
by the composition 2FeO.Si02, but contain alumina from the various charge material*
and calcium oxide which is added for fluidity. Since reverb slags are discarded,
the copper contained in the reverb slag is a major cause of copper loss in pyro-
metallurgical practice. The concentration of copper in the slag increases with
increasing matte grade. This behavior limits the matte grades normally obtained
in conventional reverberatory practice to below 50% Cu.
When using a reverb for green charge smelting, 20% to almost 45% of the
sulfur in the feed is oxidized and is removed from the furnace with the off-gases.
The wet off-gases can contain 1.5-3% sulfur dioxide. When using calcine smelting,
sulfur dioxide evolution is lower and about 10-15% of the sulfur in the unroasted
feed material is contained in the reverb off-gases. 862 concentration in the wet
off-gases in this case can vary between 0.5-1%. In neither case is recovery as
H-SO, practical.
120
-------
The hot gases from the reverb are cooled in waste heat boilers, which extract
up to 50% of the sensible heat in the gases. A considerable amount of dust is
removed in the waste heat boiler and the gases are further cleaned in electro-
static precipitators before venting to the atmosphere.
Reverberatory furnaces can vary in width from about 22 ft to 38 ft and in
length from about 100 ft to 132 ft. The roofs of the older reverberatory furnaces
are sprung arch silica roofs, while almost all the newer furnaces have suspended
roofs of basic refractory. Over the years two types of reverberatory furnaces
have evolved, each with its specific charging methods. The first and older is the
deep bath reverberatory furnace which contains a large quantity (in excess of
100 tons) of molten slag and matte at all times. In modern deep bath reverberatory
furnaces, the molten material is held in a refractory crucible with cooling water
jackets along the sides to greatly diminish the danger of a breakout of the
liquid material. In deep bath smelting, several methods exist for charging. Wet
concentrates can be charged using slinger belts (high speed conveyors) that
spread the concentrates on the surface of the molten bath. Dry concentrates or
calcines from the roaster can be charged through the roof or via a Wagstaff gun,
(an inclined tube). Roof charging (side charging) is rarely practiced in conjunc-
tion with deep bath smelting because of dusting problems with fine dry calcine
and explosion problems with green charge. Wagstaff guns minimize these problems
and are commonly used.
The second type of reverberatory furnace is the dry hearth type in which a
pool of molten material exists only at the tapping end. The dry hearth type
furnaces are charged with wet or partially dried concentrates (green feed
smelting) or with calcines through the roof. In the latter case the dusting
problem can be quite severe for fine concentrates.
(5) Converting
Matte produced in the reverberatory furnace is transferred in ladles to the
converters using overhead cranes. The converters used in copper smelting are of
the cylindrical Fierce-Smith type, the most common size being 13 ft by 30 ft.
Air is blown from the side through a series of openings called tuyeres. During
the initial blowing period (the slag blow) FeS in the matte is preferentially
oxidized to FeO and Fe^O, and sulfur is removed with the off-gases as S02ซ Flux
is added to the converter to combine with iron oxide and form a fluid iron
silicate slag. When all the iron is oxidized, the slag is skimmed from the furnace
leaving behind "white metal" or molten Cu2S. Fresh matte is charged into the
converter at this stage and the slag(blowing continued until a sufficient quantity
of white metal has accumulated. When this happens the white metal is oxidized
with air to blister copper during the "copper blow". The blister copper is
removed from the converter and cast or subjected to additional fire refining
prior to casting. Converter blowing rates can vary between 12,000 to 30,000
scfm air. Also, the S02 content of the off-gases is lower during "slag blow" than
during "copper blow".
Cooling of the hot converter gases is necessary in order to prevent thermal
damage to the dry gas cleaning equipment. Normally, this is accomplished by
adding dilution air that can vary in quantity from 1-4 times the converter off-
gas. With dilution air, S02 concentrations in the converter off-gases can vary
121
-------
from 1-7%. With close fitting hoods or with Hoboken converters, the off-gases
average 5-10% SC^. However, when dilution air is not used, such cooling devices
as waste heat boilers, air/gas heat exchangers, or water sprays are necessary.
The converter gases pass via a balloon flue or individual high velocity
flues to dry gas cleaning equipment such as cyclones or electrostatic precipi-
tators. When these gases are to be used for acid manufacture an electrostatic
precipitator for dry gas cleaning is not essential since the wet gas cleaning
system (wet scrubber, electrostatic deraister, etc.) removes all the particulates
from the gas stream. Thus, with proper hooding, the converter off-gas is suffi-
ciently high in sulfur dioxide to be suitable for sulfuric acid manufacture, but
converting by its very nature is a batch operation and the off-gas flow rates
vary widely. In the smaller copper smelters that use two or three converters, the
scheduling of converter blows in order to obtain relatively steady flows to the
acid plant is a difficult problem.
d. Refining
The blister copper produced by smelting is too impure for most applications
and requires refining before use. It may contain silver and gold, and other
elements such as arsenic, antimony, bismuth, lead, selenium, tellurium, and iron.
Two methods are used, for refining copper - fire refining and electrolysis.
The fire-refining process employs oxidation, fluxing and reduction. It is
based on the weak affinity of copper for oxygen as compared with that of the
impurities. The molten metal is agitated with compressed air, sulfur dioxide is
liberated, and some of the impurities form metallic oxides, which combine with
added silica to form slag. Sulfur, zinc, tin, and iron are almost entirely
eliminated, and many other impurities are partially eliminated by oxidation.
Lead, arsenic, and antimony can be removed by fluxing and skimming as a dross.
After the impurities have been skimmed off, copper oxide in the melt is reduced
to metal by inserting green wood poles below the bath surface (poling). Reducing
gases formed by combustion of the pole convert the copper oxide in the bath to
copper. In recent years, reducing gases such as natural gas or reformed natural
gas have been used. If the original material does not contain sufficient gold or
silver to warrant its recovery, or if a special purpose silver-containing copper
is desired, the fire-refined copper is cast directly into forms for industrial
use. If it is of such a nature as to warrant the recovery of the precious metals,
the fire refining is not carried to completion but only far enough to insure
homogeneous anodes for subsequent electrolytic refining.
In the electrolytic refining step, anodes and cathodes (thin copper starting
sheets) are hung alternately in concrete electrolytic cells containing the elec-
trolyte which is essentially a solution of copper sulfate and sulfuric acid.
When current is applied, copper is dissolved from the anode and an equivalent
amount of copper plates out of solution on the cathode. Such impurities as gold,
silver, platinum-group metals, and the selenides and tellurides fall to the
bottom of the tank and form anode slime or mud. Arsenic, antimony, bismuth, and
nickel enter the electrolyte. After the plating cycle is finished, the cathodes
are removed from the tanks, melted, and cast into commercial refinery shapes.
The copper produced has a minimum purity of 99.9%. The anode slime is treated
for recovery of precious metals.
122
-------
APPENDIX B
GLOSSARY
Anode copper - Copper 98-99% pure produced in a smelter cast into a specific
shape for electrorefining,
Blister copper - copper produced by a smelter. Chemically the same as anode
copper but cast into other shapes.
Cathode copper - copper 99%+ pure produced via electrolytic deposition.
Toll smelting - the smelting of concentrates owned by an independent mine for
a charge ("toll charge") and returning the copper.
Custom smelting - the purchase of concentrates produced by an independent mine
for smelting.
Matte - mixture of copper and iron sulfide produced during smelting.
Roasting - high temperature partial oxidation of sulfides without melting.
Calcine smelting - the charging of roasted concentrates to a smelting unit.
Green feed smelting - the charging of concentrates directly to a smelting unit
without roasting.
123
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing!
. REPORT NO.
EPA-600/7-76-034n
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
ENVIRONMENTAL CONSIDERATIONS OF
SELECTED ENERGY CONSERVING MANUFACTURING PROCESS
OPTIONS. Vol. XIV. Primary Copper Industry Report
5. REPORT OATE
December 1976 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
10. PROGRAM ELEMENT NO.
EHE624B
11. CONTRACT/GRANT NO.
68-03-2198
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA-OKD
16.SUPPLEMENTARY NOTES Vol. IH-XIII, EPA-600/7-76-034.C. through EPA-600/7-76-034h and XV,
EPA-600/7-76-034o refer to studies of other industries as noted below; Vol. I, EPA-
600/7-76-034a is the Industry Summary Report and Vol. IT. F.PA-60n/7-7fi-fn4'h ?ซ t-hp
16. ABSTRACT industry Priority Report
This study assesses the likelihood of new process technology and new practices being
introduced by energy intensive industries and explores the environmental impacts df
such changes.
Specifically, Vol, XIV deals with the primary copper industry and examines six
alternatives: (1) Outokumpu flash smelting, (2) Noranda process, (3) Mitsubishi
process, (4) oxygen use in smelting, (5) metal recovery from slags (flotation or
electric furnace), and (6) Arbiter process, all in terms of relative economics and
environmental/energy consequences. Vol. III-XIII and Vol. XV deal with the following
industries: iron and steel, petroleum refining, pulp and paper, olefins, ammonia,
aluminum, textiles, cement, glass, chlor-alkali, phosphorus and phosphoric acid, and
fertilizers. Vol, I presents the overall summation and identification of research
needs and areas of highest overall priority. Vol. II, prepared early in the study,
presents and describes the overview of the industries considered and presents the
methodology used to select industries.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b,IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Energy; Pollution; Industrial Wastes;
Copper
Manufacturing Processes;
Energy Conservation;
Smelting; Pyrometallurgv;
Hydrometallurgy
13B
18. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
unclassified
21. NO. OF PAGES
144
20. SECURITY CLASS (Thispage)
unclas sified
EPA Form 2220-1 (9-73)
124
------- |