EPA-650/2-74-085-b
SEPTEMBER 1974
Environmental Protection Technology Series
-
I
55
o
p
UJ
O
:?ff^ljl|S-Si|:-:fe^i|i^!-§-:-S^
-------
EPA-650/2-74-085-b
CONTROL OF SULFUR DIOXIDE EMISSIONS
FROM COPPER SMELTERS: VOLUME II -
HYDROGEN SULFIDE PRODUCTION
FROM COPPER CONCENTRATES
by
C.A. Rohrmann andH.T. Fullam
Battelle Pacific Northwest Laboratories
Battelle Boulevard,
Richland, Washington 99352
Contract No. 68-02-0025
Program Element No. 1AB013
ROAPNo. 21ADC-056
EPA Project Officer: L. Stankus
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
September 1974
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does mention
of trade names or commercial products constitute endorsement or recommen-
dation for use.
11
-------
-11-
TABLE OF CONTENTS
LIST OF FIGURES iv
LIST OF TABLES vi
ABSTRACT 1
SUMMARY 3
INTRODUCTION 6
BACKGROUND 9
PROCESS DESCRIPTION 14
EXPERIMENTAL PROGRAM 18
COPPER CONCENTRATES STUDIED 18
NEUTRAL ROASTING OF CONCENTRATES 21
ACID LEACHING OF NEUTRAL ROASTED CONCENTRATES 22
ANALYTICAL PROCEDURES 25
RESULTS AND DISCUSSION 28
NEUTRAL ROASTING 28
Composition of Neutral Roasted Concentrates 28
X-Ray Analysis of Neutral Roasted Concentrates 33
Thermal Analysis of Concentrates 39
Trace Impurities 48
Sulfur Dioxide Formation 49
ACID LEACHING 51
Concentrate Type 54
Concentrate Composition 57
Operating Variables 62
Fate of Impurities During Leaching 70
Process Optimization 76
ECONOMIC ANALYSIS OF PROCESS 78
NEUTRAL ROASTING SYSTEM 84
LEACHING OPERATION 87
ACID RECOVERY SYSTEM 90
S02 RECOVERY SYSTEM 93
-------
-m-
TABLE OF CONTENTS (contd)
CONVERTER OPERATION 94
TOTAL PROCESSING COSTS 97
CONCLUSIONS 102
RECOMMENDATIONS 106
REFERENCES 107
-------
-IV-
LIST OF FIGURES
1 Schematic Flow Diagram for Acid Leaching Copper Concentrates. . 15
2 Equipment for Leaching Neutral-Roasted Concentrates
in Hydrochloric Acid 23
3 Typical Sulfide Titration Curve Using Sulfide Electrode ... 26
4 Mineral Compositions in the Fe-Cu-S System 35
5 Thermogravimetric Analysis of Copper Concentrates in
Helium 40
6 Thermograms for Concentrates in Helium 41
7 Effect of Heating Rate on the Thermal Decomposition of
Morenci Concentrate 43
8 Effect of Temperature on Roasting of Morenci Concentrate ... 44
9 Differential Thermal Analysis Curves for Copper Concentrates
Heated in Purified Helium . 46
10 Thermograms for Morenci Concentrate 47
11 Typical Leaching Data for Neutral Roasted Morenci Concentrate . 52
12 Reactivity of Neutral Roasted Morenci Concentrate 59
13 Effect of Initial Acid Concentration on Leaching of Neutral
Roasted Morenci Concentrate 64
14 Effect of Initial Acid Concentration on Dissolution of
Neutral Roasted Morenci Concentrate 65
15 Effect of Initial Acid Concentration on Leaching of
Neutral Roasted Pima Concentrate 66
16 Effect of Initial Acid Concentration on the Leaching of
Neutral Roasted Anaconda Concentrate 67
17 Effect of Excess Acid on Leaching of Neutral Roasted
Morenci Concentrate 68
18 Effect of Excess Acid on Leaching of Neutral Roasted
Pima Concentrate 69
19 Effect of Temperature on Leaching of Neutral Roasted
Morenci Concentrate 71
20 Effect of Temperature on Leaching of Neutral Roasted Pima
Concentrate 72
21 Effect of Temperature on Leaching of Neutral Roasted
Anaconda Concentrate 73
-------
-V-
LIST OF FIGURES (contd)
22 Effect of Reaction Temperature on the Leaching of Neutral
Roasted Morenci Concentrate 74
23 Process Flow Diagram for 300 T/D Copper Smelter Using
Morenci Concentrate Plus Pyrite Concentrate 80
24 Process Flow Diagram for a 300 T/D Copper Smelter Using
Pima Concentrate Plus Pyrite Concentrate 81
-------
-VI-
LIST OF TABLES
I Composition of Concentrates Evaluated 19
II Composition of Neutral Roasted Concentrates 29
III Effect of Temperature on Composition of Neutral
Roasted Morenci Concentrate 31
IV Composition of Neutral Roasted Pima and Anaconda
Concentrates Prepared at Various Temperatures 32
V Compounds in the Fe-Cu-S System that are Stable
at Low Temperatures 34
VI The X-Ray Diffraction Patterns for Samples of Morenci
Concentrates which were Neutral Roasted at 800°C in
Flowing Argon on the Thermobalance 37
VII X-Ray Diffraction Patterns for Neutral Roasted Pima
Concentrate Prepared Under Various Conditions 38
VIII Effect of Neutral Roasting at 800°C on the Impurities
in Anaconda Concentrate 48
IX Material Balance Data for a Typical Leaching Experiment
Using Neutral Roasted Morenci Concentrate 53
X Leaching Data for Neutral Roasted Concentrates 56
XI Effect of Concentrate Composition on Leaching , 58
XII Effect of Composition on Leaching of Neutral Roasted
Pima and Anaconda Concentrates 61
XIII Leaching of Neutral Roasted Pyrite-Copper Concentrate
Mixtures 61
XIV Fate of Impurities During Leaching of Anaconda
Concentrate 75
XV Effect HpS Treatment on Impurities in the Leach Solution. . 76
XVI Bases Used in Preparing Flow Diagrams 82
XVII Basis for Estimating Plant Operating Costs 83
XVIII Capital Cost for Neutral Roasting Operation 85
XIX Operating Cost for Neutral Roasting Operation 86
XX Capital Cost for the Leaching Operation 88
XXI Operating Costs for the Leaching Operation 89
XXII Capital Costs for Acid Recovery System 91
XXIII Operating Costs for Acid Recovery System 92
-------
-V11-
LIST OF TABLES (contd)
XXIV Capital Costs for S02 Recovery System 95
XXV Operating Costs for S0? Recovery System 96
XXVI Capital Costs for Converter Operation 98
XXVII Operating Costs for Converter Operation 99
XXVIII Capital Costs for a 300 Ton/Day Copper Smelter 100
XXIX Yearly Operating Costs for a 300 Ton/Day Copper
Smelter 101
-------
-1-
CONTROL OF SULFUR DIOXIDE EMISSIONS FROM COPPER SMELTERS:
VOLUME II - HYDROGEN SULFIDE PRODUCTION FROM COPPER CONCENTRATES
ABSTRACT
A laboratory investigation has been made of a modified copper smelting
process which provides a solution to the sulfur dioxide air pollution prob-
lem and produces blister copper, elemental sulfur, and iron oxide without
loss of the precious metals. Preliminary economic evaluation of the process
appears favorable with good prospects for further improvements when com-
pared with conventional processes provided with equivalent air pollution
abatement capabilities.
The process would involve (1) neutral roasting of pyritic copper con-
centrates to convert the contained iron into an acid-soluble form with
evolution of some elemental sulfur in this step, (2) hydrochloric acid
leaching of the roasted concentrate to dissolve the iron with simultaneous
hydrogen sulfide generation and production of an enriched copper sulfide
residue, (3) converting the copper sulfide residue to blister copper by
conventional means, (4) reducing the sulfur dioxide formed in the convert-
ing step to elemental sulfur with hydrogen sulfide from the leaching step,
and (5) processing the iron chloride leach solution to regenerate hydro-
chloric acid and to yield a marketable iron oxide. Step 1 (for pyrite con-
version), as well as steps 3, 4 and 5, are regarded as present industrial
state-of-the-art. Step 2, based on the results from the principal focus of this
laboratory study, would involve relatively straightforward chemical engineering
development for scale-up to pilot-plant scale to demonstrate feasibility.
Principal advantages of the process include: (1) hydrogen sulfide
production without requiring the utilization of a specific chemical reduc-
tant such as costly natural gas or coke; (2) elimination of the costly
investment and operations of reverberatory furnaces with their accompanying
high fuel demand, dilute sulfur dioxide flue gas releases and concurrent
-------
-2-
formation of massive amounts of high-temperature solid wastes; (3) major
reductions in the requirements for slag-forming minerals; and (4) probable
increased overall recovery of copper and precious metals from a specific
ore body.
-------
-3-
CONTROL OF SULFUR DIOXIDE EMISSIONS FROM COPPER SMELTERS:
VOLUME II - HYDROGEN SULFIDE PRODUCTION FROM COPPER CONCENTRATES
SUMMARY
As a result of the laboratory work done under this project, it has
been determined that: (1) by neutral roasting of pyritic copper concen-
trates at 800°C or higher a large fraction of the iron in such concentrates
is converted to an acid-soluble sulfide form; (2) by dissolution of the
solubilized iron in 4-5N. hydrochloric acid at 80-85°C large quantities of
concentrated hydrogen sulfide can be readily obtained; (3) an enriched
copper sulfide still containing the precious metals is readily separated
from the acidic iron chloride solution; and (4) the small amounts of copper
which are also dissolved can be readily recovered by treatment of the
cooled acidic-iron chloride solution with hydrogen sulfide. This treatment
also assures the retention of the precious metals, principally silver, with
the copper sulfide residues.
The regeneration and recovery of the hydrochloric acid from the
ferrous chloride solution as produced above is regarded as a present state-
of-the-art commercial process applied on a substantial and increasing scale
in the steel pickling industry.
The conversion of the moderately-concentrated sulfur dioxide evolved
from conventional copper smelter converters to elemental sulfur by reaction
with the hydrogen sulfide produced as indicated above is also regarded as a
present state-of-the-art process, the so-called Claus process. However, an
alternative means utilizing the U.S. Bureau of Mines Citrate process, which
is in a fairly advanced state of development, may also be employed for
elemental sulfur production.
The production of blister copper from the enriched copper sulfide con-
centrate obtainable by the above treatment is regarded as a present state-
of-the-art process which can be accommodated in copper smelter converters
of the existing design.
-------
-4-
Thus the technical, or at least the chemical, feasibility of the pro-
cesses necessary for the recovery of elemental sulfur in high yield and
quality from copper smelter flue gases without need for a specific and
increasingly more costly chemical reductant such as natural gas or coke
has been demonstrated on the laboratory scale. This is regarded as the
positive answer to the principal original objective of this investigation.
In reviewing the composition of representative pyritic copper concen-
trates and the results of this project, essentially all such concentrates
are and can be expected to be deficient in pyrite. However, it is also
understood that all but a very few copper sulfide ore bodies are charac-
terized by having an excess of pyrite present. Suitable copper sulfide-
iron sulfide concentrations for matte production are almost universally
obtained by depressing the flotation of the excess pyrite and thus reject-
ing it to the tailings. It is thus accepted that sufficient pyrite for
the conduct of the proposed process already exists in almost all copper
sulfide ore bodies and that by moderate, tolerable and probably desirable
changes in flotation process control, the required amounts of pyrite could
be concurrently or separately recovered without difficulty or significant
economic penalty. For chemical reasons as developed in this laboratory
study, separate recovery and processing of the required additional pyrite
appears preferable.
The overall investment and operating costs as determined by a prelimi-
nary economic study of the proposed copper smelting process (through anode
production) with a capacity of 300 tons of copper per day shows it to
involve an investment of about $52,000,000 and an overall smelting cost
for copper of about 12 cents per pound without credit for recovered
elemental sulfur and iron oxide. Both of these resources are regarded as
marketable and at values that are regarded as conservative under today's
conditions (at $20 and $10 per short ton for sulfur and iron oxide,
respectively) the overall smelting cost for copper drops to about 9 cents.
Although a detailed estimate and comparison with a conventional smelter
was not made, it is understood that such smelting costs (through anode
-------
-5-
copper production) are 7 to 8 cents per pound, but without equivalent
sulfur emissions control. However, with such controls another 5 to 6 cents
per pound would be added. It is felt that a detailed study of the proposed
process could identify substantial economies which would reduce the overall
production cost significantly.
The overall assessment of the chemical, technical and economic feasi-
bility of the proposed process is definitely regarded as favorable at this
stage.
Continued work at the bench-scale level to obtain data suitable for
pilot plant design, particularly work involving the continuous rather than
batch operation of the two principal processes of neutral roasting and acid
dissolution, is recommended.
In view of the results of this study and the near and actual state-of-
the-art technology which may be employed for the principal processing
operations, it is concluded that the overall proposed process has favorable
prospects for achieving an adequate and economically acceptable control
(97%) of sulfur dioxide emissions for conventional smelting processes
integrated with the proposed process.
-------
-6-
INTRODUCTION
Although in the U.S. four copper producers* have announced plans for
non-smelting type processes to augment normal production or for large pilot
plant operations, a review of worldwide copper smelting practices involving
about 100 major plants shows no trends to indicate a significant shift away
from conventional roaster-reverberatory-converter type overall operations.
Roasters, when required, are generally used to adjust the sulfur content of
the charge to the reverberatory furnaces. Reverberatory furnaces are used
to adjust the iron content and to produce molten matte (a fluid homogeneous
mixture of iron and copper sulfides). Converters are used to produce
blister copper from matte. Blister copper is a relatively impure metal
containing some copper oxide, the precious metals and other impurities such
as nickel and selenium. However, in order to meet the increasingly more
stringent regulations on sulfur dioxide emissions, there is a very definite
trend in the U.S. and elsewhere to initiate major changes in the smelting
practices, the objective being mainly to assure the production of higher
strength sulfur dioxide waste gases from which principally sulfuric acid
can be produced most efficiently and thus realize a major reduction in
sulfur dioxide emissions. In many cases this acid is finding use nearby
in expanded leaching operations for low-grade oxidized copper ores. The
changes in smelter practices do, however, involve major changes in smelter
equipment such as for flash smelting and continuous converting. One plant
in the U.S. (Tennessee Copper Co.) converts part of its waste gas to liquid
sulfur dioxide. The American Smelting and Refining Co. plant at Tacoma,
Washington, is about to begin very large-scale production of liquid sulfur
dioxide from a portion of the concentrated converter gas that is not used
for sulfuric acid manufacture. Although consideration is being given at
some plants to convert concentrated gas to elemental sulfur or to dispose
* Cyprus Mines: The Cymet Process
Duval Corp.: A Chloride Leach Process
Anaconda: A Low-Pressure Ammonia Leach Process
Hecla: A Roast-Leach-Electrowin Process
-------
-7-
of excess sulfuric acid by conversion to waste gypsum, no plants are known
to be operating such processes on a large scale today in the U.S. One new
copper producer (Hecla Mining Co.) is to operate roasters to produce oxide
copper and sulfuric acid for a leaching and electrowinning process.
In general, the conversion of waste sulfur dioxide to elemental sulfur
has been regarded as an ideal solution to the problem. Activity in this
direction has generally been concerned with the chemical reduction of the
sulfur dioxide utilizing natural gas or coke as the reductant. At some
point in the chemistry of natural gas reduction, hydrogen sulfide is
regarded as being formed, then the usual reaction between hydrogen sulfide
and unconverted sulfur dioxide proceeds to form elemental sulfur. Thus
hydrogen sulfide appears as a logical reagent for effective conversion of
sulfur dioxide to elemental sulfur. Consideration of processes for hydro-
gen sulfide production without the use of a chemical reductant such as
natural gas have not received much attention. However, this was the objec-
tive of the earlier efforts on this present project where H2S production
was accomplished by reaction of high-temperature water vapor with the
residual iron sulfide obtained in an initial step of neutral roasting of
the pyritic copper concentrate. This process was shown to be unfavorable
because of unsatisfactory equilibrium conditions from which only very dilute
(less than 1%) hydrogen sulfide gas was obtained. Accordingly, such pro-
duction and effective conversion of the iron sulfide was achievable only by
treatment with very large volumes of high-temperature water vapor. Pre-
liminary economic studies showed this process to have unlikely prospects
for economic viability.
An alternative process was subsequently scoped. In this process
(1) neutral roasting would still be required (including in many cases pro-
vision of additional pyrite) followed by (2) acid treatment for hydrogen
sulfide production, (3) acid regeneration and recovery, and (4) enriched
copper sulfide recovery for delivery to conventional converters. The
sulfur dioxide from the converting step would be (5) reacted with an ade-
quate amount of hydrogen sulfide from the acid treatment step to produce
-------
-8-
elemental sulfur. This sulfur, together with that evolved from the neutral-
roasting step, would comprise essentially all of the sulfur present in the
original pyritic copper concentrates. Thereby overall production of ele-
mental sulfur would be achieved by thermal means without the need for a
specific chemical reductant. This present report covers the details of
the laboratory efforts directed at assessing the chemical feasibility of
such a process and its economic potential. The principal areas for study
include: (1) the neutral-roasting step and (2) the acid-leaching or
hydrogen-sulfide production step. Related matters concerned with feed,
residue and gas composition were studied concurrently.
Most copper sulfide resources in the U.S. and throughout the world
are associated with pyrite. For the recovery of concentrates from these
ores flotation treatment is generally used. In this treatment excess
pyrite which is generally present in substantial amounts is discarded
along with the gangue. This is accomplished by simple chemical adjustments
in the concentrator to preferentially depress the flotation of the iron
sulfide (pyrite) and thus reject it from the circuit. Since in most cases
additional iron sulfide is essential to the process to provide the required
amounts of hydrogen sulfide, it is a requirement of the suggested process
concept that sufficient pyrite be retained with the copper sulfide concen-
trates (or separately recovered) for subsequent use. This appears to be
easy to accommodate in existing concentrators.
-------
-9-
BACKGROUND
THE OVERALL CONCEPT
Although the individual steps of the proposed process concept comprise
well-known and fairly well-understood chemistry and technology, the combi-
nation of these steps or processes to achieve the desired objective specif-
ically in copper smelting appears not to have been explored. The patents
/I ?\
by S. I. Levyv ' ' employ similar steps specifically to treat pyrite for
recovery of iron, sulfur and other metals such as copper, lead and zinc.
These patents disclose the combination of steps of neutral roasting, the
recovery of sulfur from neutral roasting, hydrogen sulfide conversion to
sulfur, hydrochloric acid treatment to preferentially dissolve iron and
liberate hydrogen sulfide, the separation of a copper-rich residue, the
regeneration of hydrochloric acid with iron oxide recovery and, in addition,
the recovery of zinc and lead by separate electrolytic processes. The
technology employed for iron and hydrochloric acid processing was different
from more recent processes by involving crystallization of the chlorides as
the primary means of separation from other metal chlorides which may be
present in substantial amounts. The inventions do not include the further
processing of insoluble copper residue or the utilization of hydrogen sul-
fide to react with any sulfur dioxide derived therefrom.
NEUTRAL ROASTING
The literature on the chemistry and technology of neutral roasting
wherein pyrite is thermally decomposed to yield elemental sulfur and
pyrrhotite is extensive and goes back many years; however, much of it is
not of sufficient relevance to be useful in the present investigation.
The most recent and useful compilation of background information has been
reviewed in a report by Parsons and Ingraham. ' The thermochemical
mineralogy of the copper-iron-sulfur system has been intensively studied
by many investigators.^ ' This work has included consideration of the
extremely complex and still imperfectly defined central portion of the
-------
-10-
system. Although simple chemical equations can be written to indicate the
possible decomposition of the many copper-iron-sulfide compounds which
may be present in practical pyritic copper concentrates to yield FeS, Cu~S
and sulfur, it is clear that such simplifications do not actually occur
under conditions which are encountered in the present process concept.
As to the application of neutral roasting on a practical scale in
recent years, perhaps the most noteworthy activity has been that of the
Outokumpu Company in Finland relative to their flash smelting developments.
This process has been indicated to utilize a neutral-roasting process on
(?6}
pyrite to produce elemental sulfur and molten iron sulfides (FeS).
This process has sufficient flexibility so that either neutral or almost
any degree of oxidative roasting can be utilized. It is uncertain at
present whether or not elemental sulfur production via neutral roasting is
actually continuing today.
Kunda and Mehta and coworkers^ ' ' have disclosed the details of a
proposed very large-scale process for pyrite neutral roasting in Canada.
These references are most useful for their engineering and economic details.
Watkinson and Germain^ ' of Noranda, also in Canada, have reported useful
laboratory data on neutral roasting of pyrite intended for large-scale
applications. No references to work specifically related to practical
copper concentrate neutral roasting were uncovered although Tkachenko, et
al., and Isakova, et al., °~32' reported on studies of the thermal decom-
position of pyrite, chalcopyrite and bornite.
Although specific references to the neutral roasting of pyritic copper
concentrates were not uncovered, the related work including actual large-
scale applications to pyrite alone, which is a significant fraction of the
neutral roasting processing as intended in the proposed process concept,
leads to the conclusion that neutral roasting as proposed here is largely
a present state-of-the-art technology.
-------
-11-
LEACHING
The leaching of processed pyritic minerals with acids for preferential
dissolution of iron comprises a very old concept. Parsons and Ingrahanr
have also extensively reviewed this. Their review indicates many concepts
which recognize the relative insolubility of copper in such treatment pro-
cesses. Van Weert, Ingraham and Thornhill and co-workers^ " ' recently
reported on the leaching of nickeliferous pyrrhotites and matte with hydro-
chloric acid. In their discussion of this work, regeneration of hydrochloric
acid from hydrated nickel chloride crystals was indicated as a commercial
process. It was also mentioned that such technology was being extended to
ferrous chloride. However, references to operating processes which utilize
the chemistry of leaching were not uncovered. Leaching is, therefore,
perhaps the major part of the proposed overall process which is definitely
not present state-of-the-art technology. However, it appears to be a
technology which is easy to develop to a practical scale and has been
worked on by others on a laboratory scale at least.
ACID REGENERATION
Hydrochloric acid regeneration from corresponding acid-ferrous chloride
solutions is also an old concept and recently, within the last ten years,
has assumed considerable importance as the preferred and alternative means
of providing a regenerative process for use in the steel pickling industry
in the U.S. This process regenerates the pickling acid (hydrochloric acid)
and produces concurrently iron oxide suitable for recycle in steel making.
Reeve and Ingrahanr ' have extensively reviewed this technology which
derives mainly from European developed practices. Installations in the
U.S. include the Ruthner-Dravo process and more recently the Lurgi process.
The Turbulator Process, used in Europe and licensed in the U.S. by Haveg
Industries, has apparently not yet been utilized in the U.S. The latter
process is reported to be able to accommodate solutions as low as 0.5% and
as high as 11% by weight in acid and up to 35% in dissolved ferrous chloride.
This is support for the expectation that such processes have very wide ranges
-------
-12-
of flexibility for processing solutions of differing composition with
essentially no practical limits. Acid regeneration and concurrent iron
oxide production are clearly regarded as present state-of-the-art processes.
The only problem here appears to be one of scale-up. Although pickle
liquor processing as practiced is of substantial size, mainstream ferrous
chloride decomposition and hydrochloric acid regeneration as contemplated
in the proposed concept may be on a scale at least ten times as large as
in a single pickle liquor regeneration plant. The selection of which
existing process of those indicated above to use would be determined by
ease in scale-up and ultimate detailed construction and operating costs.
CONVERTING
Converting of matte to blister copper is a well-established state-of-
the-art process in essentially all sulfide copper smelters throughout the
world. In the proposed process concept, a feed richer in copper sulfide
would replace the usual matte composition. It is understood that such a
change can be accommodated by adjustments in fuel and air or oxygen delivery
to the converter to compensate for the absence of the large amounts of iron
sulfide which are present in conventional mattes. The lowered iron in the
feed will also result in substantial reductions in the amount of slag which
will be produced by the converter. However, special (yet conventional)
means would probably be needed to prepare this relatively small amount of
slag for delivery back to the concentrator for flotation treatment for
residual copper recovery.
SULFUR PRODUCTION
Sulfur production achieved by catalytically reacting sulfur dioxide
and hydrogen sulfide is an old state-of-the-art process. Its principal
application today is the process by which "sour gas" (natural gas contain-
ing hydrogen sulfide, normally as a major contaminant) is purified for
widespread distribution and use. The reaction is also employed commer-
cially in the petroleum refining industry as the means of cleaning the
-------
-13-
vent gases which may have intolerable hydrogen sulfide concentrations. In
these processes (which are modifications of the so-called Claus process),
the extracted hydrogen sulfide is partially burned to form the required
ratio of hydrogen sulfide to sulfur dioxide. This mixture of gases is
then reacted on a bauxite catalyst to yield elemental sulfur as either a
gas, liquid or solid, depending on the temperature employed.
The Claus process, though effective and economical as normally
operated, does not achieve complete reaction. The vent gas may require
further treatment before release. Alternative processes have therefore
been receiving increasing attention.
Another process which is fairly advanced in development is the U.S.
(•}£}
Bureau of Mines Citrate Process/ ; It is primarily being considered as
a means for removing sulfur dioxide from smelter flue gas (and has also
been proposed for power plant flue gases) by absorption of the sulfur
dioxide in buffered aqueous sodium sulfite-sodium citrate solutions
followed by contacting with gaseous hydrogen sulfide (separately prepared)
to yield elemental sulfur. The recovered sulfur is intended to be par-
tially converted back to hydrogen sulfide by chemical reduction to provide
the required amounts of hydrogen sulfide. This process is being installed
by USBM on a demonstration scale using a sulfur dioxide-bearing flue gas
stream from the lead smelter of the Bunker Hill Company at Kellogg, Idaho.
This process is of special interest because of its expected capability for
effective removal of sulfur dioxide from very dilute gas streams with
accompanying high yields and effective conversion to elemental sulfur. In
these features it may have significant advantages over the conventional
Claus process for application in the proposed process concept. In any case,
the high concentrations of sulfur dioxide expected from the converter gas
stream and also the very high concentrations of hydrogen sulfide produced
in the leaching step suggest that either a Claus process or a Citrate pro-
cess may be utilized. If the Citrate process demonstrates substantially
higher overall recoveries of sulfur from dilute flue gases, it may be the
recommended choice. Its full demonstration, which is expected to be
initiated this year, is awaited with interest.
-------
-14-
PROCESS DESCRIPTION
The objective of the proposed process is to produce hydrogen sulfide
by leaching neutral roasted copper concentrate containing iron sulfide
with hydrochloric acid. The hydrogen sulfide is formed by the reaction of
the acid with iron sulfide. Assuming the iron is present as FeS, the H,,S
is formed according to the reaction:
The hydrogen sulfide can then be used to convert the sulfur dioxide in the
smelter waste gas streams, particularly the converter gas, to elemental
sulfur via the reaction
2H2S(g) + S% , - ' 3S(s) + 2H2°(g)
(36 }
Either a modified Claus process or the Bureau of Mines Citrate process v
can be used to produce the elemental sulfur from the H^S and SC^ .
The schematic flow diagram for the proposed process is presented in
Figure 1. The principal operations in the process are as follows:
1. Run-of-mill flotation concentrate is dried at 150-200°C to remove
water.
2. The dried concentrate is fired at 800-1 000°C in a neutral or reducing
atmosphere to remove the labile sulfur. The sulfur is condensed and
collected. The roasting operation converts the bulk of the iron in
the concentrate to a form that is soluble in hydrochloric acid.
3. The neutral roasted concentrate is leached in hydrochloric acid
(4-5N) to dissolve the iron and convert an equivalent amount of the
sulfur to hydrogen sulfide. In this step the copper sul fides are
relatively insoluble.
4. The slurry from the leach tank is filtered and the solid residue
washed with water.
5. The solids, containing the bulk of the copper and a small amount of
iron, are sent to the converter for copper recovery.
-------
SULFUR
t
s~
-^ i
FLOTATION
CONCENTRATE
HC1 '
HC1 RE
Fe203
1
« j
EACH TAN< |
1
WASH
IATER j 1,
V m
-. LEACH SOLUTION rT| -,
f
WASH
FWATER
PER • A
COPPER-BEARING SOLIDS
ER
COPPER-BEARING
^ KhblUUL " ^
TO CONVERTER
FIGURE 1. Schematic Flow Diagram for Acid Leaching Copper Concentrates
-------
-16-
6. The cooled leach solution, plus wash, which contains the bulk of the
iron and small amounts of copper, is treated with a fraction of the
H?S. This precipitates the copper and some trace impurities, such as
silver, as the sulfides. The precipitate is separated by filtration,
washed, and sent to the converter for copper recovery.
7. The leach solution, which now consists of hydrochloric acid and iron
chlorides (principally as ferrous chloride), is treated to convert
the iron chlorides to ferric oxide and recover the hydrochloric acid
for recycle. Commercially available processes used for treating
hydrochloric acid pickle liquor can be used to process the leach
solution.
8. Make-up hydrochloric acid is added to the process as necessary, to
maintain the acid inventory.
9. In the converter the copper sulfide is reacted to form copper and SO,,.
10. The H?S from the leaching step is collected and used to reduce the
SCL from the converter to elemental sulfur.
11. The by-product sulfur and ferric oxide can be stockpiled or sold
depending on market conditions.
In order for the process to achieve maximum utility it is necessary to
produce enough H2S in the leaching operation to react with all of the S02
produced in the copper smelter, primarily from the converting step. This
means that about two-thirds of the sulfur in the neutral-roasted concen-
trate must be converted to H2S. In actual practice more than two-thirds
conversion must be obtained. There are two reasons for this: (1) some
H?S is needed to precipitate the dissolved copper and trace impurities in
the leach solution; and (2) some S02 is formed in the neutral-roasting
operation. It may be required that this S02 also be converted to elemental
sulfur. In a typical commercial operation it would probably be necessary
to convert about 70% of the sulfur in the neutral-roasted concentrate to
H?S in order to convert all of the S02 produced in the plant to elemental
sulfur. For concentrates which are deficient in iron sulfide, additional
amounts of iron sulfide (pyrite) would have to be provided. This is not
regarded as a particularly critical problem because most sulfide copper
-------
-17-
ore bodies contain excess pyrite which is rejected by conventional control
in the concentrator's flotation operations. Such excess pyrite could
therefore be separately recovered or allowed to accumulate with the copper
sulfide concentrate. It is probable that only a very few major sulfide
ore bodies in the U.S. would be sufficiently deficient in pyrite to impair
the likelihood of realizing a practical overall process.*
* The White Pine, Michigan, operation of the Copper Range Co. is the
typical example.
-------
-18-
EXPERIMENTAL PROGRAM
The principal operations in the proposed process were demonstrated on
a laboratory scale using glass and Vycor reactors. To simplify the experi-
mental procedures the operations were carried in batch reactors rather
than continuously.
COPPER CONCENTRATES STUDIED
In order for the proposed process to be applicable to a smelter opera-
tion, enough H?S must be formed in the leaching reaction to combine with
all of the S(L generated in other parts of the smelter (2 moles of H«S
required per mole of S(L). The composition of the neutral-roasted concen-
trate can have a significant effect on the leaching reaction and on the
HpS formed. To evaluate the effect of composition, seven different copper
concentrates, as well as pyrite, were studied in this program. The con-
centrates were obtained from various domestic copper producers as typical
run-of-mi11 flotation concentrates. The chemical and mineralogical com-
positions of the concentrates studied are given in Table I. (The mineral-
ogical compositions are those reported by the suppliers of the concentrates.)
Three of the concentrates (Morenci, Pima, and Anaconda) were considered to
be typical of the general types of copper concentrates produced in the
United States and were studied in considerable detail. The Morenci con-
centrate was obtained from the Phelps Dodge Corp. and is principally a
mixture of chalcocite and pyrite. The Pima concentrate was also obtained
from the Phelps Dodge Corp. and is a mixture of chalcopyrite and pyrite.
The Anaconda concentrate was obtained from the Anaconda Company and is a
complex mixture of iron and copper sulfides as well as zinc, arsenic and
antimony sulfides.
All of the concentrates contain varying levels of minor constituents
which can affect their use in the proposed process. These include those
materials which can affect the safety aspects of the process such as
arsenic, antimony, mercury, etc., and the precious metals, gold and silver,
-------
-19-
TABLE I. Composition of Concentrates Evaluated
(Chemical Composition is in Weight
Percent)
1
Element
Cu
Fe
S
Insol.
As
Sb
Zn
Pb
Mo
Bi
Cd
Se
Te
Hg
Ag*
Au*
>P_
y
C
O 3 T-
0 r- S-
i— 0. >,
i «3 d.
i .C
: O
(0
E
O-
27.14
27.3
29.6
10.16
0.06
0.025
0.30
0.04
0.02
0.016
0.002
0.038
0.0042
0.028
2.8
0.01
Ol
• r-
t- Ol
>-> 1/1 +J
D. 3 T-
Or- 1-
U 0- >i
i— Q.
fO
JC
o
•" -*~ — •
IO
x>
C
o
u
C
26.93
19.4
34.7
6.64
1.86
0.037
6.2
0.05
0.05
0.03
0.12
0.029
0.025
0.024
10.42
0.07
3"
_Q
"I/I I/)
01 01
Li_ ""O
C -r-
X M 4-
QJ r—
r— -3
o. tn in
E <
o
o
01
C
o
>-,
1—
20.31
30.1
40.0
6.48
0.39
0.05
0.3
0.04
0.085
0.012
0.001
0.028
0.0058
0.031
1.54
0.01
Ol
^J
•r- Q)
O 1/1 4J
O 3 -r-
U f— l-
1— 0. >,
1X3 Q.
JC
O
i.
0>
-o
c +->
Ol -r-
> CI-
IO
_ J
12.65
32.0
39.7
8.88
0.36
0.03
1.8
0.20
0.025
0.024
0.004
0.036
0.011
0.024
1.94
0.07
CJ
•* t/1
O> Ol
u. -o
X 1-
01 •—
CL tn
E
o
C
• f—
CU fO
4J C
-(-> 3
to O
CQ s:
25.0
31.3
13.12
0.06
0.063
0.70
0.03
0.018
0.020
0.003
0.051
0.0064
0.038
10.79
0.31
. "
O) *r- OJ
4J &~ 1 *
•r- >,-r-
0 0. i-
0 0 >,
U U O-
r— i —
IT5 ID T3
-C .C c:
O U IO
)
Ol
3
C
IO
C
<0
CO
"
29.58
27.28
29.49 '
01
'r-
i.
>^
D.
O
u
i —
•0
.C
IO
u
01
1C.
36.96
44.23
Ol
+J
•r"
L.
O.
In troy ounces per ton
-------
-20-
which must be recovered for their economic value. Therefore, it is neces-
sary to know the fate of the minor constituents as the concentrates are
processed.
Pima and San Manuel materials are typical of chalcopyrite concentrates.
Morenci and Tyrone materials are typical of chalcocite concentrates.
Anaconda and Lavender Pit materials are typical of complex sulfides charac-
teristically high in zinc and arsenic. Battle Mt. and Anaconda concentrates
are characteristically high in precious metals, gold and silver.
If one ignores the minor constituents, the concentrates have the
following compositions:
Morenci Fe0.87Cu0.58S2 (M0.73S)
Pima Fei.o6Cu0.92Ss (M0.99S)
Anaconda Feo.64Cu0.78S2 (M0.72S)
Tyrone* Fe0.86Cu0.51S2 (M0.69S)
Lavender Pit* Fe0.93Cu0.32S2 (M0.63S)
Battle Mt.** Fe0.99CV85S2 (M0.92S)
San Manuel*** Fei.Q6Cul,01S2 ^M1.04S)
Hecla**** FeS2.09
* Also obtained from Phelps Dodge Corporation representative of their
Tyrone, New Mexico, and Bisbee, Arizona, operations.
** Obtained from the Duval Corporation's Nevada operation.
*** Obtained from Newmont Mining Company's Magma Copper, Arizona,
operations.
**** pyrite concentrate representative of material from Utah, obtained
from Hecla Mining Co.
-------
-21-
NEUTRAL ROASTING OF CONCENTRATES
The neutral roasting of the various concentrates was carried out in an
atmosphere of purified argon or helium, or under a vacuum (<100 microns
absolute pressure). The concentrates were roasted at temperatures from
800 to 1020°C. A Vycor reactor was used to contain the concentrates at
roasting temperatures up to 900°C; alumina was used for temperatures
above 900°C. The general procedure used for neutral roasting the concen-
trates was as follows.
A weighed amount of previously dried concentrate was placed
in a horizontal Vycor reactor tube; the tube was purged with
argon or helium. The reactor was heated in a tube furnace to the
desired temperature and maintained at temperature for 6-24 hours.
The reactor was then cooled to room temperature. The inert gas
flow was maintained throughout the cycle. After cooling, the
neutral-roasted concentrate was removed from the tube and ball-
milled for one hour in a porcelain mill using alundum balls. The
milled concentrate was screened through a 70-mesh screen and
stored in a sealed glass jar. Oversize material was crushed
and milled again.
When the neutral roasting was carried out in an alumina crucible, the
procedure was modified as follows.
A weighed amount of concentrate was placed in an impervious
alumina crucible and the crucible was placed in a closed-end
Vycor tube. The tube was placed in a pot furnace and purged
with argon. The concentrate was then roasted in the manner
described above.
When the roasting was carried out in a vacuum, a two-stage vacuum pump
was used to maintain the vacuum. Two traps were placed between the reactor
and the pump to prevent SO,, and sulfur from entering the pump.
The bulk of the concentrates were roasted at 800°C under argon in
one-kilogram batches. In the case of the Morenci, Pima, and Anaconda
-------
-22-
concentrates, several one-kilogram batches of each were neutral-roasted at
800°C. The various one-kilogram batches of each neutral-roasted product
were combined and blended by ball milling for one hour. The composite
blend of each concentrate was used for the bulk of the leaching studies.
Smaller batches of each of the three concentrates were neutral-roasted
at temperatures above 800°C to determine the effect of temperature on the
product and on the leaching reaction. At the maximum roasting temperature
HOOO°C) the three concentrates were partially melted.
ACID LEACHING OF NEUTRAL-ROASTED CONCENTRATES
Leaching of the neutral-roasted concentrates in hydrochloric acid was
carried out in the laboratory employing equipment as illustrated in Fig-
ure 2. The reaction vessel is a one-liter three-necked flask. The pro-
cedure used in the leach tests was as follows.
1. The reaction flask was placed in a thermostatically controlled water
bath.
2. The required volume of hydrochloric acid was added to the reaction
flask.
3. The water in the bath was heated to the required temperature using a
cartridge heater and thermoregulator. The temperature of the hydro-
chloric acid was monitored with a thermometer. The bath temperature
and acid temperature, at equilibrium, never varied by more than 1°C.
4. Water in the bath and acid in the reaction flask were stirred using
magnetic stirring bars and a stirring hot plate.
5. While the acid was being heated the reactor was purged with argon.
Off-gas from the reactor passed through a water-cooled condenser and
two traps. The bulk of the water in the off-gas was condensed in the
condenser and returned to the reaction vessel. The first trap con-
tained a dilute (^.BH) hydrochloric acid solution and served to
collect any hydrogen chloride in the off-gas. The second trap con-
tained a known volume (usually 1 liter) of 1.5N. sodium hydroxide and
was used to absorb the H^S in the off-gas.
-------
CONDENSER
COOLING WATER
ARGON
TO POWER
SUPPLY
THERMOREGULATOR —•
HEATER
GLASS TANK
WATER
I
WATER
ACID *.
SCRUBBER
0.5N HC1-
THERMOMETER
1-LITER 3-NECK
FLASK
HYDROCHLORIC ACID
+ CONCENTRATE
MAGNETIC
STIRRERS
STIRRING
HOT PLATE
H?S ABSORPTION
.5M NaOH
MAGNETIC
STIRRER
STIRRING
HOT PLATE
FIGURE 2. Equipment for Leaching Neutral-Roasted Concentrates in Hydrochloric Acid
-------
-24-
6. When the acid reached the desired temperature the concentrate was
added.
7. The course of the reaction was followed by absorbing the H^S in the
caustic solution (designated as scrubber solution or off-gas absorber
solution) and analyzing the solution periodically for its sulfide
content.
8. In some runs the solution in the reaction vessel was sampled periodi-
cally, filtered, and analyzed for iron and copper.
9. The run was normally continued until the sulfide concentration in the
scrubber solution (step 7 above) reached a constant level indicating
dissolution of the iron in the acid was complete.
10. When the run was complete, the reaction slurry was filtered and the
solid residue washed with water.
11. The solid residue was dried at 110°C, weighed and analyzed for iron,
copper and sulfur. (In certain runs the residue was also analyzed
for minor constituents.)
12. The filtrate and wash solutions were combined, diluted to a known
volume and analyzed for iron and copper. (As in 11 above in certain
runs the solution was analyzed for minor constituents.)
13. Material balances for the iron, copper and sulfur were then calculated
based on the feed, solid residue, leach solution, and scrubber solu-
tion analyses.
After several runs it became apparent that the amount of H^S liberated
was a good measure of the iron dissolved (normally very little copper dis-
solved) and was the most reliable method of following the leaching reaction.
For this reason in most of the runs no attempt was made to sample and
analyze the leach solution during the course of the run. Only the final
solution was analyzed for iron and copper.
Some runs were made at the boiling point of the hydrochloric acid. For
4N HC1 this was about 106°C. Since temperature control was not critical,
as long as the solution was boiling, the water bath was not needed. Instead
the reaction vessel was heated directly using the stirring hot plate.
-------
-25-
ANALYTICAL PROCEDURES
A variety of procedures was used to analyze the concentrates, neutral -
roasted concentrates, leach solutions, leach residues and scrubber solu-
tions. In addition, considerable use was made of a commercial assayer to
analyze the concentrates, neutral-roasted concentrates and leach residues.
Results obtained from the assayer were checked periodically by in-house
analysis of duplicate samples. The assayer was employed for economic
reasons. In-house analytical facilities could not compete on a cost basis
with the assayer on routine analyses such as iron, copper and sulfur in
concentrates.
The H~S evolved during the leaching was determined by absorbing the
H^S in sodium hydroxide solution and analyzing the solution for sulfide
ion.
H2S + 2NaOH > 2Na+ + S= + 2H20
The sulfide content of the caustic solution was determined using a sulfide
specific ion electrode (Orion Model 94-16A) and calomel reference electrode
(Beckman Perma-Probe solid state reference electrode Model 39407). The
following procedure was used.
A known volume of the sulfide-caustic solution (usually
2-5 mis) was added to about 150 ml of 1M_ sodium hydroxide solu-
tion. The solution was thoroughly mixed using a magnetic
stirrer. The sulfide sample was then titrated with 0.1^ silver
nitrate solution. The titration end point was determined using
the sulfide electrode, reference electrode and a digital milli-
volt meter.
A typical titration curve is shown in Figure 3. The point of greatest
inflection is taken as the end point. The procedure precision was found
to be ±2.0%.
Iron in the leach solutions was determined using the procedure
developed by Harvey, et al/37' and a Gary Model 14H recording spectro-
photometer. The procedure allows the simultaneous determination of both
-------
-26-
-1000
-800
-600
-400
^ -200
200
400
600
END POINT
5ml SAMPLE-NaOH SOLUTION
I I I I
8 12 16
0.1M AqN03 ADDED,
20
24
FIGURE 3. Typical Sulfide Titration Curve Using Sulfide Electrode
-------
-27-
iron (II) and total iron in the solution. Iron (III) is obtained by
difference. Impurities present in the leach solution, such as copper and
chloride, do not interfere at the levels present. The procedure used is
as follows:
A known volume of leach solution (50-100 lambda) is placed in a
50 ml volumetric flask. Ten ml of 0.3% 1,10 phenanthroline and 5 ml
of 0.2M potassium acid phthalate are added. Distilled water is added
to the mark and the solution stirred. The absorbance of the solution
is read (within 30 minutes) on the Gary at 396 my and 512 my. The
precision of the procedure is about ±1.5%.
Preparation of the standard concentration curves is rather complex as is
calculating the iron (II) and iron (III) from the curves. The reader is
referred to the original reference^ ' for complete details on the
procedure.
Copper in the leach solutions was determined by atomic adsorption
using standard procedures.
The procedures described above were also used to determine iron and
copper in the concentrates, neutral-roasted concentrates and leach residues
after they had been solubilized by fusion techniques. Sulfur in the solids
was determined using the Leco S0~ analyzer.
The various concentrates and neutral-roasted concentrates were analyzed
by thermal techniques using the DuPont Model 990 Thermal Analyzer and
Model 951 Thermogravimetric Analyzer.
-------
-28-
RESULTS AND DISCUSSION
The basic steps in the proposed process were demonstrated on a labora-
tory scale. These included the neutral roasting, acid leaching, and copper
recovery from the leach solution. Conversion of the ferrous chloride to
ferric oxide and recovery of the hydrochloric acid was not studied in detail
because commercial processes are available for this operation. To simplify
the equipment required batch reactors were used for the neutral roasting
and leaching operations. On a commercial scale, however, both operations
would expect to be carried out in continuous reaction systems.
NEUTRAL ROASTING
Neutral roasting of the copper concentrates is necessary in order to
carry out the acid leaching operation. When run-of-mill flotation concen-
trate is leached with hydrochloric acid there is very little reaction and
no production of H2S. When the concentrate (or pyrite) is heated to an
elevated temperature (>500°C) in an inert atmosphere part of the sulfur is
evolved and the residue obtained is partially soluble in hydrochloric acid.
Almost all of the iron in the neutral-roasted concentrate dissolves forming
ferrous chloride with the evolution of H2S. The neutral-roasting operation
is, therefore, a vital part of the overall process.
Substantial volumes of each concentrate were roasted under various
conditions for use in the leaching studies. In addition, thermal analysis
(DTA and TGA) was used to determine the optimum conditions for the neutral-
roasting operation. Chemical and X-ray analysis were used to assess the
composition of the neutral-roasted concentrates.
Composition of Neutral-Roasted Concentrates
When the concentrates are heated in an inert atmosphere (or vacuum)
water and elemental sulfur (plus small amounts of H2S and S02) are evolved.
In a static system the amount of sulfur released will depend on the min-
eralogical composition, system temperature and the sulfur partial pressure
-------
-29-
(assuming equilibrium is attained). In a flowing system, where the sulfur
is removed as it is formed, the amount of sulfur evolved will depend on the
mineralogical composition, the reaction temperature and the time at tempera-
ture. The geometry of the concentrate bed can affect the efficiency of
sulfur removal and thus affect the time required to achieve a given sulfur
loss. This will be discussed in a later section.
For the leaching studies, the bulk of the concentrates were neutral-
roasted at 800°C in flowing argon or helium (no differences in product
composition were detected as a result of interchanging the inert gases).
They were processed in one-kilogram batches. Where more than one kilogram
of a given concentrate had to be treated, several batches were processed
and the products combined and blended by ball milling. The compositions
of the neutral-roasted concentrates, prepared at 800°C, are presented in
Table II. It is impossible to assign specific formulas to the neutral-
roasted concentrates. However, ignoring the minor constituents, the
neutral-roasted products have composition ratios as listed following Table II
TABLE II. Composition of Neutral Roasted Concentrates
Concentrate
i
Morenci
Pi ma
Anaconda
Tyrone
Battle Mtn
1 Lavender Pit
I
San Manuel
Hecla
I (pyrite)
Temp.
TO
800
800
800
800
800
800
800
800
Atmosphere
Argon
Argon
Argon
Helium
He 1 i urn
Helium
He 1 i urn
Helium
Time at
Temp.
(hrs)
24
24
24
16
16
16
24
24
Product Composition, wt%
Fe
33.4
29.4
23.2
36.4
28.1
39.1
28.5
52.9
Cu.
24.7
29.4
34.9
23.9
28.1
15.4
32.5
—
S
28.8
27.8
26.5
28.6
24.9
29.0
26.7
33.9
-------
-30-
Concentrate Composition Fe+Cu/S
Morenci Fel.33Cu0.86S (M1.10S)
Pima Fe-j 21Cu1.07S2 ^M1.14S^
Anaconda Fe-, n-|Cu-| ^2 ^1 17^
Tyrone Fel.46Cu0.84S2 (M1.15S)
Battle Mtn Fe1 2gCu1 ]4$2 (M1.22S^
Lavender Pit Fel.55Cu0.54S2 (M1.05S)
San Manuel Fe-, 09^ui o-jS? (^i ?^)
\ » C.L. I . £.3 £ I • ^--J
Hecla FeS1 12
Comparing these compositions with those for the concentrates reported
earlier one can see how the metal-sulfur ratio changes due to sulfur loss
during neutral roasting.
Neutral Roasted
Concentrate Concentrate
Morenci MQ 73S Ml.10S
Pima M0.99S M1.14S
Anaconda Mg y2$ ^1.17^
Tyrone MQ>6gS M]J5S
Battle Mtn MQ g2S M1.22S
Lavender Pit V\Q^S M^^S
San Manuel M^ ^^S ^1.23^
Hecla FeS2.09 FeSl.l2
By neutral roasting at temperatures above 800°C sulfur removal can be
increased. Table III shows the composition of Morenci concentrate which
was neutral roasted at various temperatures. While there are some varia-
tions it is apparent that high temperatures and/or long reaction times
-------
-31-
TABLE III. Effect of Temperature on Composition
of Neutral Roasted Morenci Concentrate
Sample
No.
1
2
3
4*
5**
Temp.
°C
800
800
860
970
1005
Time
(hrs)
24
16
20
3
20
Atmosphere
Product Composition, wt% •
Fe
•
Argon
Argon
Helium
Vacuum
Vacuum
33.4
32.3
32.1
37.3
33.5
Cu
s i
!
24.7 > 28.8 !
!
23.7
24.6
28.5
29.2 i
27.0 1
30.5 j
25.9 j . 25.2 i
I 1
Ignoring the minor constituents the samples have the
following compositions:
Sample No.
1
2
3
4
5
Composition
Fe1.33Cu0.86S2
Fe1.27Cu0.82S2
Fe1.37CY92S2
Fe1.40Cu0.94S2
Fe1.42Cu1.04S2
Fe+Cu/S
M1.10S
M1.05S
M1.15S
M1.17S
M1.23S
* Product consisted of two phases: a solid mass (^ 40 wt%) which
had fused and a sintered mass. The analysis given is for the
fused portion.
** Product had fused.
-------
-32-
increase the metal-to-sulfur ratio of the neutral-roasted concentrates.
Similar data for neutral-roasted Pima and Anaconda concentrates are given
in Table IV.
TABLE IV. Composition of Neutral Roasted Pima and Anaconda
Concentrates Prepared at Various Temperatures
Concentrate
Pima
i
Pima* A
B
i
Pima
Pima
Anaconda
Anaconda
Temp.
(°0
800
860
860
970
1005
800
1000
Time
(hrs)
24
72
' 72
18
6
24
4
Atmosphere
Argon
Vacuum
Vacuum
Vacuum
Vacuum
Argon
Vacuum
Product Composition, wt%
Fe
29.4
28.9
29.1
28.9
29.1
23.2
28.1
Cu
29.4
28.6
30.9
29.7
29.3
34.9
42.3
S
27.8
24,6
25.5
24.7
24.8
26.5
26.9
M/S
»1.14S
M1.25S
M1.27S
M1.28S
M1.27S
M1.17S
M1.39S
* Product consisted of two phases: (A) a partially sintered easily crushed powder,
and (B) a highly sintered hard mass. The phases were separated and analyzed
individually.
The physical characteristics of the neutral roasted concentrates varied
with the reaction temperature and time at temperature. Concentrate processed
at 800°C consisted of partially sintered easily crushed lumps and powder.
Those heated to 1000°C or higher had partially fused and were more difficult
to crush and grind. Concentrates processed at intermediate temperatures
varied from loosely sintered powders to fused lumps depending on the time
and temperature. The Pima concentrate showed a somewhat greater tendency
to sinter than the others.
-------
-33-
The particle size distribution of the Morenci concentrate, which had
been roasted at 800°C for 24 hours in argon and then ball milled, was
determined using U.S. Standard Sieve Series Screens. The distribution was
as shown in the following:
Screen Size Cumulative % Retained
+70 0.08
+100 6.04
+140 21.78
+200 47.84
+230 72.35
+325 89.96
Total 100.00
The material had a tap density of 2.34 g/cm3, and the surface area was
<0.5 sq m/gram.
X-Ray Analysis of Neutral-Roasted Concentrates
The various neutral-roasted concentrates were analyzed by X-ray dif-
fraction in an attempt to identify the minerals present. These attempts
met with only marginal success and it was impossible to identify, with
certainty, the minerals present in the neutral roasted concentrates. A
detailed study of the mineralogical composition of the neutral-roasted
concentrates was beyond the scope of this program.
The iron-copper-sulfur system is extremely complex. Despite decades
of investigation by many different workers the phase diagram for the system
has not yet been completely defined.(5~25' Table V is a list of most of
the minerals (stable at low temperatures) reported in the literature for
the Fe-Cu-S system.^5' It does not include questionable or incompletely
described minerals which have been reported. Figure 4 shows the Fe-Cu-S
system and location in the diagram of the various minerals listed in
Table V. The locations of the neutral-roasted concentrates (ignoring the
minor constituents) are also indicated.
-------
-34-
TABLE V. Compounds in the Fe-Cu-S System that
are Stable at Low Temperatures(5)
Symbol
a-bn
an
bbcv
bn
cc
cf
cv
cb
di
dj
fk
gr
he
h-po
Id
mk
ma
mh
m-po
py
sm
tal
tr
Name
anomalous bornite
an i1i te
blaubleibender covellite
bornite
chalcocite
chalcopyrite
covellite
cubanite
digenite
djurleite
fukuchilite
greigite
haycockite
hexagonal pyrrhotite
Idaite
mackinawite
marcasite
mooihoekite
monoclinic pyrrhotite
pyrite
smythite
talnakhite
troilite
probable new mineral
new mineral
synthetic mineral
Formula
Cu5FeS4
C1.75S
CU1.1S
Cu5FeS4
Cu2S
CuFeS2
CuS
CugS5
CU1.96S
Cu3FeS8
Fe3S4
Fe9S10
Fe1.06S
Fe7S8
FeS2
Fe3S4' Fe3.34S4
Cu9Fe8S16
FeS
CucFeSc
D D
Cu0.12Fe0.94S
Cu3Fe4S6
-------
-35-
Cu
Fe
FIGURE 4. Mineral Compositions in the Fe-Cu-S System
(See Table V for meaning of symbols.)
-------
-36-
The compositions of the neutral roasted concentrates (except for the
(5)
Hecla) fall in the central area of the Fe-Cu-S system. Cabri,v Yund and
Kullerud^24' and others have shown that at elevated temperatures the cen-
tral area of the Fe-Cu-S system is characterized by a large solid solution
field. The composition of the neutral roasted copper concentrates place
them within the solid solution field at temperatures of 700°C and above.
According to Cabri^ "The central area of the Cu-Fe-S system is charac-
terized by a large solid solution field at elevated temperatures which
breaks up into five distinct phases at low temperatures. The picture is
complicated by numerous phase transformations, unquenchable phases, and
phases with closely related crystal chemistry, resulting in their having
similar physical appearance and X-ray diffraction powder patterns. The
slow rate of most of the reactions at low temperatures also makes it diffi-
cult (in some cases impossible) to achieve equilibrium in the laboratory."
This statement helps explain the difficulties encountered in trying to
identify phases present in the neutral roasted concentrates. When the con-
centrates were neutral-roasted they were held at temperature for limited
periods of time (4-24 hours). It is not known if this time was sufficient
for solid solution to occur. In addition, no attempts were made to control
the quench rate when the concentrates were cooled. Therefore samples of a
given concentrate which were neutral-roasted under approximately similar
conditions could have different mineralogical compositions depending on
such factors as variations in cooling rate. For example, four samples of
air dried Morenci concentrate were heated to 800°C in flowing helium on the
thermobalance and then cooled to room temperature. The cooling rate used
was the normal cooling rate of the thermobalance furnace with the power off
(approximately 90 minutes from 800 to 50°C). The samples were then analyzed
by X-ray diffraction. The diffraction patterns for the four samples are
presented in Table VI. It is readily apparent that there are significant
differences between the patterns as well as great similarity. The lines of
highest intensity are very similar for all four samples. Similar results
were observed with the Pima and Anaconda concentrates.
-------
-37-
TABLE VI. The X-Ray Diffraction Patterns for Samples of
Morenci
at 800°C
Sample
No. SM-1
dA°
6.46
3.37
3.28
3.16
3.09
3.01
2.75
2.67
2.51
2.37
2.09
1.97
1.94
1.93
1.89
1.73
1.64
1.61
1.33
1.22
1.11
1.09
1.03
I/Io
15
20
25 .
25
100
20
25
30
15
20
35
20
35
40
70
20
20
35
20
15
15
20
20
Concentrates which were Neutral Roasted
in Flowing Argon on the Thermobalance
Sample
No. SM-2
dA°
6.50
4.56
3.36
3.15
3.09
3.01
2.90
2.78
2.66
2.08
1.93
1.89
1.81
1.73
1.61
1.57
1.33
I/Io
15
15
10
100
55
10
10
10
20
25
20
30
10
15
15
10
5
Sample
No. SM-3
dA°
6.50
3.37
3.30
3.15
3.11
3.00
2.75
2.67
2.54
2.10
1.93
1.89
1.75
1.61
1.33
1.15
1.11
1.09
I/Io
30
20
25
45
100
25
20
30
15
45
50
55
20
30
20
15
15
15
Sample
No. SM-4
dA°
6.42
3.36
3.11
2.68
2.10
1.89
1.61
1.45
1.33
1.23
1.09
1.03
I/Io
10
15
100
10
10
55
20
5
5
10
10
5
-------
-38-
Samples of neutral-roasted Pima which were prepared at different tem-
peratures exhibited completely different X-ray diffraction patterns (see
Table VII). Samples of neutral-roasted Anaconda prepared at different
temperatures also showed a lack of similarity in their diffraction patterns,
In the case of neutral-roasted Morenci the situation was different. The
samples prepared at different temperatures had similar but not identical
diffraction patterns.
TABLE VII. X-Ray Diffraction Patterns for Neutral Roasted Pima
Concentrate Prepared Under Various Conditions
Sample P.-1* Sample P-8**
dA° I/Io
3.40 15
3.10 100
2.67 15
1.89 80
1.61 30
1.55 5
1.34 10
1.33 10
1.31 10
1.22 5
1.09 10
dA°
6.31
3.30
3.15
3.10
3.00
2.72
2.67
2.53
2.10
1.93
1.75
1.73
1.64
1.33
1.09
I/Io
30
30
60
30
40
30
40
10
85
100
20
35
20
5
5
* Concentrate neutral roasted for 24 hours at 800°C
in inert atmosphere.
** Concentrate neutral roasted for 6 hours at 1005°C
under vacuum.
-------
-39-
It was not possible to identify, with complete confidence, specific
mineral phases in any of the neutral roasted concentrates. None of the
X-ray diffraction patterns obtained could be attributed specifically to
known patterns published in the ASTM Powder Diffraction File.
In the case of the neutral-roasted Morenci and Pima concentrates
prepared at 800°C their diffraction patterns were similar, but not identi-
cal, to those reported for mooihoekite (CuQFeQS,c) and talnakhite
(5,7,8,10)
9 C9- 16'
The two materials probably contain either or both
minerals. It is also possible they may contain some cubanite (CuFe2$3).
In the case of neutral roasted Anaconda concentrate prepared at 800°C,
the X-ray diffraction pattern could not be identified with any published
patterns for the Fe-Cu-S system. The material prepared at 1000°C had a
distinctly different pattern. The pattern had a slight similarity to the
patterns for cubanite (CuPe^S.,) and bornite (Cu5FeS.), and the material
could be a mixture of the two minerals.
The bulk of the leaching tests were carried out with concentrates
which were neutral-roasted at 800°C. In the case of the Morenci, Pima and
Anaconda, several batches of each concentrate were roasted at 800°C and
blended. The blended lot of each concentrate was used for the leaching
tests. Therefore, differences in mineral composition from batch to batch
due to slight uncontrolled variations in the neutral roasting operation
would be masked by the blending operation.
Thermal Analysis of Concentrates
The neutral roasting of concentrates was studied using differential
thermal analysis and thermogravimetric analysis. Thermograms were obtained
for the various concentrates in purified helium. Figure 5 shows weight
loss as a function of sample temperature for Morenci, Pima, and Anaconda
concentrates when heated in helium. The thermograms for the other concen-
trates are shown in Figure 6. The initial weight loss (up to a temperature
of about 350°C) is due to the vaporization of water. At higher temperatures
-------
100
96
i—
o
UJ
s 92
—i
<
z
CD
£ 88
U_
O
K-- 84
Z
O
UJ
3
uj 80
_i
(X
s:
76
72
ATMOSPHERE - PURIFIED HELIUM
HEATING RATE - 20°C/MINUTE
SAMPLE PREDRIED AT 110°C
I
100 200 300 400 500 600 700 800 900 1000
SAMPLE TEMPERATURE, °C
o
i
FIGURE 5. Thermogravimetric Analysis of Copper Concentrates in Helium
-------
100
96
92
s ••
o
84
80
76
72
SAN MANUEL
BATTLE MTN.
-•£_
ATMOSPHERE-PURIFIED HELIUM
HEATING RATE-20°C/MINUTE
I I I
I
J I
I
100 200 300 400 500 600 700 800 900 1000
SAMPLE TEMPER/JURE, °C
-p.
i
FIGURE 6. Thermograms for Concentrates in Helium
-------
-42-
the weight loss is due to the evolution of sulfur (and small amounts of
H?S and SCL). In the case of the Anaconda concentrate a fraction of the
weight loss is due to vaporization of arsenic sulfide and possibly zinc
sulfide.
The data presented in Figures 5 and 6 were obtained at a heating rate
of 20°C/minute. To show the effect of heating rate on the decomposition
reaction(s) samples of Morenci concentrate were heated at various rates.
The results obtained are presented in Figure 7.
To obtain a better measure of the kinetics of the decomposition
reaction samples of Morenci concentrate were inserted into a preheated
furnace and the weight loss followed as a function of time. The thermo-
grams obtained are shown in Figure 8. It normally required from 1 to
2 minutes for the sample to reach the reaction temperature. During this
heat-up time, considerable sulfur evolution occurred, especially when the
control temperature was above 500°C. It was impossible, therefore, to
measure precisely the evolution of sulfur as a function of time at tempera-
ture. In obtaining the data presented in Figure 8 the sample temperature
was monitored as a function of time as well as sample weight. The dashed
portions of the thermograms represent the periods in which sample tempera-
tures were increasing to the control temperatures. It can be seen from
the data that a substantial fraction of the weight loss occurred before
the control temperature was reached. At temperatures of 600°C and above
most of the weight loss occurs in the first 25-50 minutes. Thereafter the
loss of weight continues at a slow, almost constant, rate for long periods
of time. One run at 800°C, using Morenci concentrate, was continued for
24 hours; a continual weight loss was observed throughout the run. The
overall weight loss at the end of the 24 hour period was 26.06%. The weight
change observed during the last few hours of the run was so slight, however,
it may have been due to instrument instability.
The thermobalance data were obtained using platinum pans with a thin
layer of concentrate spread over the pan surface. Under these conditions
the sulfur formed was easily removed from the reaction zone by the flowing
-------
TOO
ATMOSPHERE - PURIFIED HELIUM
SAMPLES PREDRIED AT 1 10"C
100 200 300 400 500 600 700 800 900 1000
SAMPLE TEMPERATURE, °C
FIGURE 7. Effect of Heating Rate on the Thermal Decomposition
of Morenci Concentrate
-------
100
96
5 92
88
84
80
55 76
72
I
1000°C
I
I
I
100 200 300
400 500 700 700
REACTION TIME, MINUTES
800 900
1000
FIGURE 8. Effect of Temperature on Roasting of Morenci Concentrate
-------
-45-
helium. When the bulk samples of concentrate were neutral-roasted the
situation was different. In the horizontal tube reactor the depth of con-
centrate in the tube was up to 1.5 inches and the bed was 8-12 inches long.
The argon purge gas tended to flow over rather than through the bed. Under
these conditions sulfur removal from the reaction zone was poor. The same
situation existed in the pot reactor where the depth of concentrate was
several inches. Because removal of the sulfur was so poor with the bulk
samples they had to be heated for many hours to achieve the same weight
loss that was observed on the thermobalance in an hour or less.
Each of the concentrates was analyzed by differential thermal analysis.
Thermograms obtained with the Morenci, Pima, and Anaconda concentrates in
argon are shown in Figure 9. In each case the endothermic reactions up to
about 350°C correspond to loss of water from the sample. The endothermic
reactions between 400 and 700°C correspond to the decomposition reactions
which evolve sulfur (and H^S and SO^). Above 700°C the major endothermic
reactions probably correspond to solid phase changes and/or sample melting.
It is fairly certain that the endotherms at 950-975°C correspond to melt-
ing. The exotherms which occur on cooling at 950-900°C result from freez-
ing: while in the case of the Anaconda the exotherm at 740°C on cooling
is probably a solid phase transition. The thermograms for the other con-
centrates are similar to those obtained with the Morenci, Pima, and
Anaconda concentrates.
A better understanding of the reactions involved can be obtained by
comparing the DTA and TGA data for a given concentrate. DTA and TGA plots
obtained with Morenci concentrate in purified helium at a heating rate of
5°C per minute are shown in Figure 10. The initial endotherms (up to 350°C)
correspond to the initial weight loss due to vaporization of water. The
large endothermic reaction between 400 and 700°C corresponds to the maximum
sample weight loss (evolution of sulfur). The endothermic reactions above
700°C are reversible and indicate a phase change of one or more constit-
uents of the concentrate. Some weight loss occurs above 700°C but this
would not account for the endothermic reactions observed.
-------
-46-
o
o
MORENCI CONCENTRATE
HEAT-
o
o
PIMA CONCENTRATE
HEAT-
o
x
ANACONDA CONCENTRATE
HEAT-
HCATING RATE - 30 C/mln
J I I I
-L I I I I
200
4CO
600
SAMPLE TEMPERATURE, °C
800
1000
FIGURE 9. Differential Thermal Analysis Curves for Copper
Concentrates Heated in Purified Helium
-------
-47-
o
x
LU
O
o
DTA CURVES
HEAT-
100
S 95
UJ
X
_l
<
I 90
u.
o
^-
z
UJ
£ 85
-------
-48-
Trace Impurities
The copper concentrates contain a number of trace impurities which
can play a significant part in a commercial process. Some of the concen-
trates contain significant precious metal values (Anaconda and Battle
Mountain). Others contain impurities which can represent a safety problem
and their fate during the various process operations must be known. The
Anaconda concentrate is one which contains significant levels of impuri-
ties; it was used to determine the fate of impurities during neutral
roasting. The concentrate was neutral-roasted for 24 hours in flowing
argon. The product was then analyzed for impurities. The results are
shown in Table VIII, together with the analyses of the original concentrate.
To account for the weight loss encountered during neutral-roasting the
impurities are reported as grams per 100 grams of original concentrate.
TABLE VIII. Effect of Neutral Roasting at 800°C on the
Impurities in Anaconda Concentrate
Grams per 100 Grams of Original Concentrate
Impurity
As
Sb
Zn
Pb
Mo
Bi
Cd
Se
Te
Hg
Ag*
Au*
Concentrate
1.86
0.037
6.2
0.05
0.05
0.03
0.12
0.029
0.025
0.024
10.42
0.07
Neutral Roasted
Concentrate
0.17
0.06
3.76
0.2
0.064
0.032
0.50
0.018
0.0056
0.00016
10.2
0.051
* In troy ounces per ton of original concentrate.
-------
-49-
From the results it can be seen that the bulk of the arsenic in the
concentrate is volatilized during roasting and should end up in the sulfur.
A fraction of the zinc also appears to have volatilized, probably as the
sulfide. Conflicting results were obtained with some of the trace level
impurities. With some elements (i.e., Sb, Pb, Cd) the analytical results
show a greater amount in the neutral-roasted concentrate than in the feed,
clearly an impossibility. This is due to lack of precision of the ana-
lytical method at very low concentrations. All that can be said is that
the neutral-roasted concentrate contains a certain amount of the impurity
but the exact level cannot be defined. In the case of gold and silver the
analyses agree, within the procedure precision, and essentially all of the
gold and silver in the concentrate end up in the neutral-roasted concen-
trate. This is a desirable situation because it assures that the precious
metals will be recovered from the concentrate by the usual techniques in
the subsequent conventional processing steps of converting and electrolytic
refining.
Sulfur Dioxide Formation
When the copper concentrate is neutral roasted, some HpS and S(L are
formed. When the gas stream from the reactor cools the ^S and SCL combine
to form elemental sulfur. However, the off-gas always contains more S02
than is required to react with the H^S; thus cold off-gas always contains
some SOp.
The exact mechanisms by which the S(L and H-S are formed have not been
determined, but they probably result from the reaction of water in the con-
centrate with the concentrate and/or elemental sulfur. This hypothesis is
strengthened by the fact that reducing the water content of the concentrate
prior to neutral roasting reduces the amount of S02 formed during roasting.
No attempt was made to determine the amount of S02 and H2S formed
during neutral-roasting, but the excess SCL in the off-gas was measured for
Morenci concentrate. When the concentrate is neutral-roasted at 800°C in
flowing helium, approximately 1.2% of the sulfur in the concentrate shows
-------
-SO-
up as free SCL. The SCL yield was approximately the same when the concen-
trate was neutral-roasted at 1000°C. The concentrate used in these tests
contained about 3.5% water. When the concentrate was dried under vacuum
at 170°C, the water content was reduced to about 1.5%. When this material
was neutral-roasted at 800°C the free S(L formed amounted to about 0.8% of
the sulfur in the concentrate.
To determine the effect of a reducing atmosphere on the decomposition
reaction and SO,, formation, Morenci concentrate was roasted in an atmosphere
of Argon-4% Hydrogen at 800°C. Use of a reducing atmosphere increased the
excess SO^ in the off-gas to about 2.1% of the sulfur in the concentrate.
The composition of the neutral-roasted concentrate was not changed signifi-
cantly from that obtained in argon.
Two samples of Morenci concentrate were also roasted at 800°C in
He-2% O^ The amount of S0« formed was substantially increased, and the
final products were magnetic. The composition of the two products were:
Sample No. 1 Sample No. 2
Fe 31.8 wt% 32.86 wt%
. Cu 24.2 24.6
S 26.4 26.7
Fe1.38Cu0.93S2 Fe1.42Cu0.93S2
M1.15S M1.17S
The sulfur content of the samples is lower than usual and it is probable
that some oxidation of the samples occurred, possibly with some sulfate
formation. The H?S evolved when these materials were treated with HC1 was
also greatly reduced. This is convincing evidence that the iron sulfide
had been oxidized.
When pyrite was neutral-roasted, about 1.5% of the sulfur in the
(79)
pyrite was evolved as SOp. Work reported by othersv ' indicated much
higher S0? formation as well as some COS formation when pyrite was neutral-
roasted. Reasons for the discrepancies between the studies have not been
developed.
-------
-51-
ACID LEACHING
The leaching of neutral roasted concentrates in hydrochloric acid was
studied in considerable detail using a batch reaction system. The princi-
pal reaction involved is the dissolution of iron sulfide to form ferrous
chloride and hydrogen sulfide. The amount of ferric chloride formed was
usually very low. The effects of the following variables on the reaction
were evaluated:
1. concentrate type (i.e., Morenci, Pima, etc.)
2. concentrate composition variations (due to temperature and time for
roasting)
3. reaction temperature
4. acid concentration
5. acid-concentrate ratio
Two variables which were not studied were stirring rate and concentrate
surface area. An attempt was made to use the same stirring rate for each
experiment, but the variations in H2S yield and iron dissolution observed
in duplicate runs may reflect variations in stirring rate. Because the
reaction .involved is a liquid-solid reaction, stirring would have less
effect on the kinetics than if a gas-solid reaction were occurring as in
the case of oxidative leaching reactions involving the use of oxygen. No
attempt was made to study the effect of concentrate surface area on the
leaching reaction. The only control used was to insure the neutral roasted
concentrates used in the leach tests were all -70 mesh in particle size.
The progress of a given leaching experiment was followed by measuring
the HLS evolved. In some'runs the leach solution was sampled periodically
and analyzed for iron and copper. Figure 11 shows typical data obtained
with neutral-roasted Morenci concentrate and a reaction temperature of
106°C (boiling point of 4f[ hydrochloric acid). By analyzing the feed
material, leach solution, solid residue and off-gas absorber solution
material balances for iron, copper and sulfur were calculated for the
leaching reaction. Material balance data for a typical leaching experi-
ment are shown in Table IX.
-------
-52-
100
o
Ul
s
1
z
^H
5!
t—t
OL
£
=>
60
50
40
30
20
10
IRON
SULFUR
NEUTRAL ROASTED MORENCI CONCENTRATE
r.
4N HYDROCHLORIC ACID
ARGON COVER GAS
REACTION TEMPERATURE~ 106°C
20
COPPER
—^-
40 60 80 100
REACTION TIME, MINUTES
120 140
FIGURE 11. Typical Leaching Data for Neutral Roasted Morenci Concentrate
-------
-53-
TABLE IX. Material Balance Data for a Typical Leaching Experiment
Using Neutral Roasted Morenci Concentrate
Reaction Conditions:
50.0 grams neutral roasted concentrate
300 ml 4j^ hydrochloric acid
Reaction temperature - 80°C
Reaction time - 150 minutes
Feed Analysis:
Fe = 33.36% (16.68 grams)
Cu = 24.66% (12.33 grams)
S = 28.8% (14.40 grams)
Residue Analysis:
Amount of residue = 21.6 grams
Fe = 0.58% ( 0.13 grams)
Cu = 56.4% (12.18 grams)
S = 20.03% ( 4.33 grams)
Solution Analysis: (Leach solution plus wash = 1000 ml)
' Fe = 16.4 g/1 (16,4 grams)
Cu = 0.46 g/1 ( 0.46 grams)
Scrubber Solution Analysis:
S = 0.311 moles/1 (9.97 grams)
Material Balances:
Feed = Residue + Solution + Scrubber Solution
Fe: 16.68 g = 0.13 + 16.4 = 16.53 g
Cu: 12.33 g = 12.18 + 0.46 = 12.64 g
S: 14.40 g = 4.33 + 9.97 = 14.30 g
-------
-54-
After a few experiments it became evident that the progress of a
leaching experiment could be followed simply by measuring the H,,S evolved.
When H?S evolution stopped or was very slow, the dissolution of iron was
complete. The dissolution of copper continued, but at a slow rate com-
pared to iron dissolution. Copper dissolution did not contribute signifi-
cantly to the generation of HpS. The rate of H2$ evolution was, therefore,
a good measure of the rate of iron dissolution; it was not necessary to
analyze the leach solution during the course of the experiment. It was
only necessary to analyze the final leach solution for iron and copper
(for material balance purposes).
In most of the leaching experiments, argon or helium was used as the
purge gas. It was felt that this would prevent the oxidation, by air, of
iron (II) in the leach solutions to iron (III). If iron (III) is present
in the solution, it could increase the copper dissolution. The analytical
data showed that in most runs the iron in solution was present as iron (II).
An exact measure of the iron (III) was difficult because the iron (II) and
total iron usually agreed within the precision of the analytical procedure.
In some runs air was used as the cover gas. In these runs the iron (II)
and total iron still were the same (within the analytical precision) and
no increase in iron (III) was observed. In addition, no significant
increase in copper dissolution was observed. Therefore, use of an inert
cover gas during leaching does not appear to be necessary.
Concentrate Type
To determine how different concentrates would respond to acid leach-
ing, the eight neutral-roasted concentrates were leached under identical
conditions. The concentrates used had all been neutral-roasted in an inert
atmosphere at 800°C for 16-24 hours.
The conditions used in the tests were not the optimum for leaching.
Time did not permit determining the optimum leach condition for each
neutral-roasted concentrate. Instead, it was necessary to select a stan-
dard set of (less than optimum) leach conditions for use in comparing the
eight concentrates.
-------
-55-
The procedure used was as follows. Fifty grams of neutral-roasted
concentrate were placed in the reaction vessel and the vessel purged with
argon. Three hundred ml of 4N^ hydrochloric acid (M00% excess) were
added, and the system heated to the boiling point of the acid (^106°C).
It took approximately 10 minutes for the acid to reach the boiling point.
Each run lasted 90 minutes. H2$ evolution was essentially complete after
90 minutes with each concentrate.
The results obtained with the various concentrates are given in
Table X. It is readily apparent from the data that since at least 67% of
the sulfur in the original concentrate needs to be converted to H,,S that
none of the neutral-roasted copper concentrates produced, under the leach-
ing conditions used, the amount of H2S needed for a workable process. The
copper concentrates which are high in pyrite (Morenci and Tyrone) come the
closest to meeting process requirements. The San Manuel, Pima and Anaconda
concentrates, which are low in pyrite, produce the least HgS. The Hecla
concentrate (which is principally pyrite) after neutral-roasting reacted
almost completely with the acid. The residue remaining after leaching con-
tained only trace levels of iron and sulfur.
From the results presented, it is apparent that in order to obtain a
workable process more H^S must be produced from the copper concentrates.
There are several ways in which this might be done.
1. Optimizing the leaching conditions.
2. Varying the neutral roasting conditions to change the composition of
the neutral-roasted concentrates.
3. Add neutral-roasted pyrite to the neutral-roasted concentrate to
increase the H2S formation.
All three approaches were investigated using neutral-roasted Morenci, Pima
and Anaconda concentrates.
In the laboratory studies the leach solution and solid residue were
separated by filtrations using a #42 paper. In a commercial operation
filtration would probably also be used for the separation. In the plant
-------
-56-
TABLE X. Leaching Data for Neutral Roasted Concentrates
Reaction Conditions:
.• •.•^T-'-i
Temperature
Reaction Time
Concentrate
Hydrochloric Acid -
:Wl06°C
90 minutes
50.0 grams
300 ml 4N acid
Neutral Roasted
Concentrate*
Morenci
Pi ma
Anaconda
Tyrone
Lavender Pit
Battle Mountain
San Manuel
Heel a**
Amount of Material in Concentrate
Reacted, %***
Fe
87.7
62.5
74.9
83.4
65.0
85.3
64.6
100.0
Cu
1.80
1.90
1.10
2.90
2.70
4.10
4.13
S '
60.6
41.0
47.9
61.2
52.3
57.1
40.1
100.0
* All concentrates neutral roasted at 800°C for 16-24 hours in
helium or argon. See Table II for analyses.
** Only 25.0 grams of neutral 'roasted concentrate used.
***
Average of two or more runs: iron and copper dissolved in
leach solution and sulfur evolved as H«S.
-------
-57-
the residue would have to be thoroughly washed to remove chloride effec-
tively before further processing to recover the copper. The filtration
characteristics of the leach residue are, therefore, an important factor
to be considered. It was found that, with the exception of the neutral -
roasted Pima, the leach residues were easily filtered and washed. Con-
flicting results were obtained with the Pima concentrate. In most runs the
residue was as easy to filter and wash as the other concentrates. In a few
runs, however, the residue was extremely difficult to filter. The cause of
the filtration difficulties has not been identified.
Concentrate Composition
The composition of a neutral-roasted concentrate varies depending on
the conditions of roasting. This variation in composition will in turn
affect the results obtained during acid leaching. To determing how composi-
tion would affect the H2$ yield, samples of neutral-roasted Morenci, pre-
pared under various conditions, were leached under identical conditions.
The leach procedure was identical to that described in the previous section
(fifty grams of neutral-roasted concentrate were leached in 300 ml of boil-
ing 4N^ HC1 for 90 minutes).
The results obtained with neutral roasted Morenci concentrate are
given in Table XI. The results show that the H2S yield (and iron dissolu-
tion) is increased by increasing the metal-sulfur ratio of the neutral-
roasted feed. This ratio is best increased by increasing the neutral-
roasting temperature, lengthening the roasting time, or both. Maximum h^S
yield was obtained when the neutral-roasting temperature was high enough to
melt the concentrate (^1000°C).
Differences in the mineralogical composition of concentrate roasted at
different temperatures may also help explain the differences in the H2S
yield. Since the mineralogical composition of the neutral-roasted concen-
trate could not be adequately determined, however, it was impossible to
evaluate this effect.
-------
-58-
TABLE XI. Effect of Concentrate Composition on Leaching
Leaching Condition:
Concentrate - 50 g neutral roasted Morenci
Acid - 300 ml 4N_ HC1
Temperature - ^ 106°C
Reaction Time - 90 minutes
Sample
No.
1
2
3
4
Roastim
Temperati
800°C
800°C
970°C
1005°C
Results
Sample
No.
1
2
3
4
3 Composition, wt%
jre Fe Cu S
32.3 23.7 29.2
33.4 24.7 28.8
37.3 28.5 30.5
33.5 25.9 25.2
of Leaching Tests:
Amount of Material in
Concentrate Reacted, %*
Fe Cu S
75.9 2.9 52.3
87.7 1.8 60.6
88.6 0.97 62.2
95.6 4.0 71.1
M/S
M1.05S
M1.10S
M1.17S
M1.23S
* Fe and Cu dissolved in leach solution, S evolved as H?S.
It was found that Morenci concentrate neutral-roasted at 1005°C was
much more reactive than that produced at lower temperatures. Data presented
in Figure 12 shows that the concentrate prepared at 1005°C reacts more
rapidly with acid at room temperature than material prepared at 800°C does
at 60 an.d 106°C. Another surprising factor is that the temperature of
leaching has only a slight effect on the reaction rate for the material
prepared at 1005°C.
-------
MORENCI ROASTED AT 80CTC
MORENCI ROASTED AT 1005"C
50g CONCENTRATE
300 ml 4N HCl
40 50 60 70
REACTION TIME, MINUTES
100
i
en
FIGURE 12. Reactivity of Neutral-Roasted Morenci Concentrate
-------
-60-
Neutral-roasted Pima and Anaconda concentrates prepared at various
temperatures were also leached to determine the effect of composition on
H2$ generation and iron dissolution. Leaching conditions were identical
for each experiment and were the same as those used with the Morenci con-
centrate. The results obtained are presented in Table XII. Again it was
found that, in general, the higher the metal-sulfur ratio in the neutral-
roasted concentrate (of a given type) the higher the HLS yield and iron
dissolution. Increased roasting temperature also increased the reactivity.
Copper dissolution during leaching of the various concentrates was
erratic. In every run in which periodic samples were taken copper dissolu-
tion increased with time. However, there was considerable variation in the
amount of copper dissolved in similar or duplicate runs. Reasons for the
variability in copper dissolution have not been determined.
Tests with neutral-roasted Heel a pyrite showed the material dissolved
almost completely in hydrochloric acid with essentially complete dissolution
of the iron and conversion of sulfur to H^S. One way of increasing the HpS
yield from neutral-roasted copper concentrates would be to leach a mixture
of neutral-roasted pyrite and neutral-roasted copper concentrate (pyrite is
usually available as a waste stream from the flotation plant in most copper
operations). Roasting the pyrite and concentrate separately would insure
sufficient H^S production. A second possibility would be to combine the
pyrite and concentrate prior to roasting. To see if this procedure could
be used, mixtures of Heel a pyrite with Morenci, Pima and Anaconda concen-
trate were neutral-roasted in helium at 820-860°C for 16 hours and then
leached. The results obtained are presented in Table XIII. The addition
of pyrite to the Pima concentrate increases the H^S yield on leaching to an
acceptable level. With the Anaconda concentrate there was a small but
significant increase in the hLS yield. In the case of the Morenci, the H?S
yield decreased slightly but was within the reproducibility of the experi-
mental procedure. One factor noted was that the reaction rate for neutral
roasted concentrate-pyrite mixtures decreased compared to straight neutral
-------
-61-
TABLE XII. Effect of Composition on Leaching of Neutral
Roasted Pima and Anaconda Concentrates
Leaching Conditions:
50.0 g neutral roasted concentrate
300 ml 4J1 HC1
Temperature -v. 106°C
Reaction Tine 90 minutes
Concentrate
Pima
P1ma* A
B
Pima
P1ma
Anaconda
Anaconda
Roasting
Temp.
°C
800
860
860
970
1005
800
1000
Comp. of Neutral Roasted Cone.
w«
Fe
29.4
29.1
28.9
28.9
29.1
23.2
28.1
wt%
Cu
29.4
30.9
28.6
29.7
29.3
34.9
42.3
wt%
L S
27.8
25.5
24.6
24.7
24.8
26.5
26.9
M/S
M1.14S
H1.25S
M1.27S
M1.28S
M1.27S
M1.17S
M1.39S
Amount of Material
in Cone. Reacted ,_*,*
fe
62.5
92.2
86.3
88.2
92.4
74.9
85.6
Cu S
1.90
1.3
0.6
0.6
1.6
1.1
4.8
41.0
64.8
56.3
59.6
59.7
47.9
51.8
* Fe and Cu dissolved in leach solution and S evolved as H?S.
, TABLE XIII. Leaching of Neutral Roasted Pyrite-Copper
Concentrate Mixtures
Leaching CunuiL'iu.is:
50.0 g neutral roasted concentrate
400 ml 4H HC1
Temperature ^ 106°C
Reaction time up to 120 minutes
Leaching Results:
Feed to Roaster
80Z Morenc1-20« Pyrite
57. 5X Pima-42.5% Pyrite
50% Anaconda-SOS Pyrite
Roasting
Temp.
(°C)
860
820
860
Roasting
Time
(hrs)
16
16
16
Comp. of Neutral
Roasted Cone. ,
Fe
35.4
39.5
35.0
Cu
21.1
30.0
20.5
wtS
S
28.2
28.8
28.0
Amount Reacted 1n
Leaching, %*
Fe
84.2
79.1
71.4
Cu
1.1
2.5
2.8
S
59.3
68.0
54.1
* Fe and Cu dissolved in leach solution and S evolved as H2S.
-------
-62-
roasted concentrates. Normally at 106°C the leaching reaction was essen-
tially complete in 60 minutes or less. With the neutral-roasted mixtures
it required 90-120 minutes for completion.
Based on the results of the leach tests with concentrate-pyrite mix-
tures, it appears that separate roasting of the concentrates and pyrite
would be preferred to combined roasting.
Samples of Morenci concentrate which had been roasted at 800°C in
He-2% 02 were leached in 4N. HC1 at 106°C. The H2S yield was greatly
decreased as was the iron dissolution. The maximum KLS yield obtained was
only 35%. In addition the reaction rate was greatly decreased. It
required approximately 10 hours before the evolution of H2S stopped. Iron
dissolution was less than 60% and copper dissolution less than 1%. These
results are difficult to explain. If oxidation of the samples occurred
during roasting, the iron dissolution should be much higher. Similarly if
sulfation occurred during roasting, both iron and copper dissolution should
be higher. The only conclusion that can be made is that the presence of
oxygen in the cover gas during roasting can adversely affect the process by
reducing the HpS yield obtained on roasting.
When samples roasted in Ar - 4% H2 were leached, the yields obtained
were similar to those obtained with concentrate roasted in argon. There-
fore the use of a reducing atmosphere during roasting does not appear to
adversely affect the process.
Operating Variables
The principal operating variables which affect the leaching reaction
are:
1. acid concentration,
2. acid-concentrate ratio,
3. reaction temperature.
The effect of each variable on the leaching of roasted Morenci, Pima, and
Anaconda concentrates was studied. The concentrates used were all neutral
roasted at 800°C in flowing argon for 24 hours (see Table II for analyses).
-------
-63-
It was found that the acid concentration used had a significant effect
on the leaching of neutral roasted Morenci concentrate (the maximum acid
concentration tested was that of the azeotrope 'vSM). Maximum iron dissolu-
tion and h^S formation was obtained with an initial acid concentration of
3-4 molar (see Figure 13). A 100% excess of acid (based on iron content
of the feed) was used in each run. Both iron dissolution and H?S produc-
tion dropped sharply as the initial acid concentration was varied from the
optimum. The dissolution of copper increased rapidly with increasing acid
concentration and was especially severe when the initial acid concentration
was greater than 4M. The amount of undissolved solid residue from the
leaching reaction varied with acid concentration and was a minimum, as
expected, when H^S production and iron dissolution were a maximum (see
Figure 14).
The initial acid concentration has far less effect on the leaching of
neutral-roasted Pima concentrate as far as H^S production and iron dissolu-
tion is concerned (see Figure 15). Both H?S formation and iron dissolution
increased slightly with increased acid concentration. As was the case with
Morenci concentrate, copper dissolution increased rapidly with increasing
acid concentration.
With neutral-roasted Anaconda concentrate maximum iron dissolution and
HpS production were obtained with an initial acid concentration of about 4M^
(see Figure 16). Changes from the optimum initial acid concentration caused
only a slight decrease in H^S formation. Copper dissolution increased with
increasing acid concentration but not to the same degree as with Morenci
and Pima concentrates.
The ratio of acid to concentrate used can affect the H^S yield obtained,
With neutral-roasted Morenci, a 100% excess of acid (in excess of that
needed to react with the iron in the feed) was required to obtain the maxi-
mum HpS yield and iron dissolution (see Figure 17). With neutral roasted
Pima, H2S formation and iron dissolution were relatively independent of the
excess acid used (see Figure 18). The neutral-roasted Anaconda concentrate
gave a slight increase in HpS production with increasing acid-concentrate
ratio.
-------
-64-
o
QC
100
90
80
70
60
50
£ 40
<
LU
oc
o
o
-------
-65-
75
oc.
UJ
D-
ft
LU
O
l—i
LO
UJ
o:
:r
o
O
to
HH
o
70
65
60
55
50
45
40
INITIAL ACID CONCENTRATION, M
FIGURE 14. Effect of Initial Acid Concentration on Dissolution
of Neutral-Roasted Morenci Concentrate
-------
-66-
CJ
Of
o
-------
-67-
o
OL
Q.
•>
O
o
o
o
z:
ct
80
70
60
50
40
30
20
10
TEMPERATURE- 106°C
TIME - 90 MINUTES
25g CONCENTRATF
.1002 EXCESS HC1
COPPER
123456
INITIAL ACID CONCENTRATE, M
FIGURE 16. Effect of Initial Acid Concentration on the Leaching
of Neutral-Roasted Anaconda Concentrate
-------
-68-
a:
UJ
a.
-------
-69-
o
ce
a
ui
UI
Of
o
o
50
40
30
E 20
10
IRON
TEMPERATURE ioe°c
TIME - 90 MINUTES
25
50
75
SULFUR
O
'COPPER
100
125
150
175
fXCESS ACID USED, PERCENT
FIGURE 18. Effect of Excess Acid on Leaching of Neutral-Roasted
Pima Concentrate
-------
-70-
The reaction temperature has a significant effect on the leaching
reaction. The effect observed was unexpected in that maximum hLS production
and iron dissolution were obtained at temperatures below the boiling point
of the acid solution. The results obtained with the three concentrates are
shown in Figures 19-21. The tests were carried out by bringing the 4f1 HC1
to the reaction temperature and then adding the concentrate. The runs were
continued until the evolution of H2$ was essentially complete. Maximum H?S
production was obtained with each concentrate when the reaction temperature
was between 80-90°C.
With neutral roasted Morenci the rate of H2$ formation increases with
increasing reaction temperature as shown in Figure 22. At a temperature of
106°C, H2S formation and iron dissolution is complete in 30 minutes or
less. At 70°C and below the reaction rate decreases rapidly and reaction
times of 4 hours or longer are required. At 80°C, where maximum H S forma-
tion is obtained (with Morenci), the reaction is complete in about 60-70
minutes. The reactivity of neutral-roasted Anaconda is similar to that of
the neutral-roasted Morenci, while reactivity of the neutral-roasted Pima
was somewhat less. With each concentrate maximum H_S yield is obtained at
a reduced reaction temperature at the expense of increased reaction time.
One additional factor was noted with regard to reaction temperature.
If the acid and concentrate were combined and heated to the reaction tem-
perature, the H2S yield was higher than when the concentrate was added to
the acid at the reaction temperature. This was especially true at reaction
temperatures of 90°C and above. The reason for this phenomenon has not
been resolved.
Fate of Impurities During Leaching
Samples of solution and residue from a leaching experiment with neutral
roasted Anaconda concentrate were analyzed to determine the fate of impuri-
ties during leaching. Table XIV shows how the impurities divide between
the leach solution and solid residue. Data for the original concentrate
and neutral-roasted concentrate are also shown (to put the data on a uniform
-------
-71-
100
90
80
o
oc
o
LLJ
70
60
50
O
O
30
20
10
COPPER
60
70
80
90
100
110
REACTION TEMPERATURE, °C
FIGURE 19. Effect of Temperature on Leaching of Neutral -
Roasted Morenci Concentrate
-------
-72-
CJ
ce.
-------
-73-
o
cc
O
-------
-74-
I/O
CM
o:
o 40 -
o:
ID
t/0
100% EXCESS 4N HC1
50.Og NEUTRAL ROASTED MORENCI
50 TOO 150 200 250
REACTION TIME, MINUTES
300 350
FIGURE 22. Effect of Reaction Temperature on the Leaching
of Neutral-Roasted Morenci Concentrate
-------
-75-
TABLE XIV. Fate of Impurities During Leaching of Anaconda Concentrate
Impurity
As
Sb
Zn
Pb
Mo
Bi
Cd
Se
Te
Hg
Ag*
Au*
Grams Per 100 Grams of Ori<
Concentrate
1.86
0.037
6.2
0.05
0.05
0.03
0.12
0.029
0.025
0.024
10.42
0.07
Neutral Roasted
Concentrate
0.17
0.06
3.76
0.20
0.064
0.032
0.50
0.016
0.0051
0.00016
10.2
0.051
inal Concentrate
Leach
Residue
0.074
0.05
0.20
0.20
0.005
0.006
0.0025
0.0094
0.0035
0.0002
2.24
0.06
Leach
Solution
0.084
0.10
3.12
0.32
0,005
0.008
0.0016
0.0064
0.0024
0.000016
7.70
0.012
*In troy ounces per ton of original concentrate.
basis the impurity levels are reported as grams per 100 grams of original
concentrate). The overall material balances for the trace level impurities
are poor due to analytical problems. It is apparent, however, that the
trace impurities divide between the leach solution and the solid residue.
The bulk of the zinc ends up in the leach solution as does the silver,
while the bulk of the gold remains in the residue.
When the leach solution is treated with H2$ most of the copper pre-
cipitates as does the silver and whatever gold is present (see Table XV).
The effect on the other impurities in the leach solution is minimal.
-------
-76-
TABLE XV. Effect H2S Treatment on Impurities
in the Leach Solution
Impurity
Cu
As
Sb
Zn
Pb
Mo
Bi
Cd
Se
Te
Hg
Ag
Au
Original
Leach Solution
0.92
0.105
0.125
3.90
0.40
0.006
0.010
0.002
0.008
0.003
0.00002
0.033
0.00005
Treated
Leach Solution
0.05
0.06
0.10
4.20
0.40
0.005
0.002
0.008
0.009
0.002
0.00001
0.002
Trace
Process Optimization
The laboratory studies have shown that it will be possible to achieve
sufficient H2S production with most types of copper concentrates to make
the proposed process viable. It will require, however, optimization of
the neutral-roasting and acid leaching operations and pyrite addition to
insure adequate H2S production from low iron concentrates.
To maximize H2S production with all types of concentrates, the follow-
ing process conditions should be met.
1. The neutral roasting operation should be carried out at the highest
temperature possible (>800°C). Melting of the concentrate during
roasting produces the optimum product from a reactivity and H?S
generation standpoint.
-------
-77-
2. Care must be taken to exclude oxygen from the cover gas during roast-
ing. A neutral or reducing atmosphere must be used to obtain maximum
H2$ production upon leaching.
3. If possible, a mixture of neutral-roasted copper concentrate and
neutral-roasted pyrite should be used for leaching. This will insure
adequate H2$ production. Since a typically sized copper smelter would
require more than one roaster, the copper concentrate and pyrite can
and should be roasted separately.
4. The leaching should be carried out with an initial acid concentration
of about 4N.at a temperature of 80-90°C. An excess of acid should be
used. The excess required will depend on the concentrate but will
probably be at least 50% and possibly as much as 100%.
5. The residence time of the concentrate in the leach tank should be the
shortest possible time consistent with the required H2$ production.
Increasing the residence time will increase copper dissolution.
If the process operation conforms to the conditions set forth above, the
H^S production will be sufficient for a viable process.
-------
-78-
ECONOMIC ANALYSIS OF PROCESS
The cost of utilizing the proposed process for a 300 tons per day copper
smelting operation was estimated based on the data developed in the laboratory
studies. Two estimates were prepared: one assumed the use of a chalcocite
(Morenci) concentrate feed and the second a chalcopyrite (Pima) concentrate
feed. In each case pyrite addition was used to increase the availability
of H^S. The pyrite concentrate was assumed to be available as a by-product
stream from the flotation mill. There may be some question concerning the
real need for additional pyrite for Morenci concentrate processing, however
pyrite addition is required with the Pima concentrate. It was used in the
Morenci case simply to insure adequate KLS formation. The pyrite would be
roasted separately from the copper concentrate, and combined with the neutral
roasted copper concentrate prior to leaching.
The following assumptions were also made in making the cost estimates:
1. The copper-bearing residue from the leaching operation would be fed
directly to a converter. This eliminates the need for a reverberatory
furnace.
2. HC1 recovery and conversion of ferrous chloride to ferric oxide would
be carried out in equipment similar to that currently being used to
process hydrochloric acid pickle liquor ' in the steel industry.
3. The reaction of H-S and SO- to form elemental sulfur would be carried
out using the "citrate" process developed by the U.S. Bureau of
(36}
Mines. If some SO^ release could be tolerated, a modified Claus
process could be used for the reaction at a somewhat lower cost.
4. The precious metals in the concentrates would end up in the blister
copper from the converter. They would be recovered when the copper
is refined by electrolysis.
5. The impurities in the concentrate, other than the precious metals,
were ignored in preparing the cost estimates.
-------
-79-
Process flow diagrams were prepared for the two cases under considera-
tion and are shown in Figures 23 and 24. Stream flows were calculated using
the data developed in the laboratory studies. The bases used in the calcu-
lations are shown in Table XVI. Impurities in the concentrates were not
considered in the preparation of the flow diagrams.
For the purpose of the cost estimates, the overall smelter process was
broken down into five major operations. Capital and operating costs for
each major operation were estimated separately (the various subsidiary
operations were included in the first estimates for the principal operations),
The five major operations are:
!• Neutral Roasting System - This includes storage, drying and neutral
roasting of the concentrates; particulate removal from the roaster
off-gas; condensation, collection and stockpiling of the elemental
sulfur; and quenching and granulation of the neutral roasted
concentrates.
2. Acid Leaching System - This operation includes blending of the neutral
roasted concentrates; leaching of the concentrates; filtration of the
leach slurry and washing of the solid residue; treatment of the leach
solution with H,,S to precipitate the dissolved copper and precious
metals; filtration of the leach solution to collect the precipitated
copper and precious metals; washing of the precipitate; and collection
of the H^S from the leach circuit.
3. Acid Recovery System - This includes evaporation of the filtered leach
solution to remove the excess water; conversion of the ferrous chloride
to ferric oxide and hydrogen chloride; collection of the hydrogen
chloride as 4 to 5N^ hydrochloric acid; and handling and stockpiling
of the ferric oxide.
4. Sulfur Dioxide Collection and Sulfur Production - This operation
includes collection of the SO^-containing gases from the converter;
handling of the hydrogen sulfide gas stream; reaction of the S0? and
H2$ in an aqueous citrate system to form elemental sulfur; stockpiling
of the sulfur; and stack-venting of the SOp-bearing off-gas (approxi-
mately 0.4% of the sulfur in the total plant feed ends up in the
stack gas from the SCL recovery unit).
-------
TO STACK
S : 1986 1/D
TO STACK
H70 : 83 T/D
MORENCI CONCENTRATE
1365 TONS/DAY
Fe
Cu
S •
26% • 354.9 T/D
• 212% • 303.3 T/D
35% • 477.8 T/D
H20 • 8* • 109.2 T/D
S2 = 24.1 TH
Slas Sty : a9 T/D
S2 i 174.5 T/0
SOs SO?). 7.2 T/D
1074.1
TO STACK"
137 T/D
NATURAL CAS .
5.430 MSCF 10
1175.1 T/D
Fe = 404.2 T/D
Cu . 303.3 T/D
S = 328.6 T/D
H,S
kATURAL
CAS 1
410MSCF/D
PYRITE CONCENTRATE
137 TONS/DAY
Fe • 36% • 49.3 T/D
S • 42% • 57.5 T/D
H?0 • 8% • 11.0 T/0
-MAKEUPHCI
TO STACK
TO STACK
SOS
2.2 T/D
SULFUR
S- 325.8 T/D
TO STACK '
SlasSC^I. CL6T/D
COPPER ANODES'^
Slas H2SI. 218 T/D
H2S
4N HCI
CIRCUIT'!* .2 i
T ' 1.63x106 CAL/D
Slas H2S>= as T/D
SOs H2SI. 217.2 T/D
Fe2
372.4 T/0
Cu . 0.1 T/D
2.1x105 Cal/D
1.87x106 CAL/0
1.79N HCI
FeCl2
FILTER
LEACH RESIDUE
StasSO?) » 110.8 T/D
Cu - 300 T/D
Fe = 31.8 T/0
S : 110.6 T/D
[
R
WASH WATER
2x.05 GAL/D
1.82X106 GAL/D
Fe = 372.4 T/l>
1
COPPER
RECOVERY
NATURAS
MSCf-/
WASH
SxlO4
CAS
D
WATER
CAL/D if
3.2 T/D
as T/o
SLAG
Cu i 300 T/D
t
HOLDING
FURNACE
i
t
FLUX
50 T/
•*cIT
Fe.
. 3.2 T/0
31.8 T/0
oo
o
i
FIGURE 23. Process Flow Diagram for 300 T/D Copper Smelter Using Morenci
Concentrate Plus Pyrite Concentrate
-------
TO STACK (
Slas S02*z7.4
, SULF
10 CONOEr
SULFUR
S i 141.7 T/D
„ TO STACK
PIMA CONCENTRATE
122 TONS/DAY
Fe = 25.0% - 300.5 T/D
Cu * 255% • 306.1 T/D
S = 28.0% - 336.6 T/D
H20 = 8.0% = 96.1 T/D
H20 = 80 T/D
NATURAL GAS_^
4.530MSCF/D
TO STACK
Slas SC^fc 2.3 T/D
$• 340.8 T/0
TO STACK
SCISSOR. 0.6
COPPER ANODES
Cu = 300 T/D
SI
S(a$H2S
SULFUR
RECOVERY
Slas SO?)
/U *
REP
FUfl
UR
ISER
Sla
-I H22 I NEUTRAL
-1 T/D | ROASTER
t
H2S
as H2S). 228 T/D
H2S
S? = 65.1 T/0 '
Slas SOz) : 2.4 T/0
76.6 T/D
• SO?) . 5.0 T/D
1024.3
T/D
1299 T/D
Fe . 434.4 T/D
S = 343.7 T/D
:u . 306.1 T/D
i
las H2S) = as T/0
) . 227.2 T/0
H20
23.8 T/D
„ 274.7 NEUTRAL 4 3187 , ' u 372 T/D
*~iyb~RnASTER*lF~l°RIER r*-^ —
NATURAL
1.130MSCF/D
r MAKEUP HCI
.TO STACK
H.n. 7 i.
*~m. SNHCI 4 HCI .
1 — — , ' i ru.iM r.Ai in RECOVERY
| FILT
Cu -302.8 T/D
Fe - 57.9 T/0
S i 115.7 T/D
= 115 9 T/D f^^
T T
INING. HOLDING
MACE FURNACE
'
• > I i-e2 03
1 . Fe - 376.5 T/D
Cu . n i i/n
WASH WATER HCI NATURAL CAS
2xl01) GAL/0 MSCF/O
, WASH WATER
Jf-| WLIU. COPPER 5xl(^CAL/D r^-
r— ' Fe . 375.6
Cu : 3.3 1
T/D' RECOVERY 'U^
/D
1
imFI k SIAG
:!iHJ *" Cu • 6.0 T/D
T F« . 57.9 T/D
\ FLUX
90 T/0
PYRiTE CONCENTRATE
372 TONS /DAY
Fe . 36% . 133.9 T/D
S . 4?% . 156.2 T/D
H?0 - 8* : 29.8 T/D
ID^Cal/D
1.23X106 Cal/0;
1.56N. HCI
1.26\1 F«CI2
TER~I
Cu . 3.2 T/0
S- 0.8 T/D
I
CO
FIGURE 24. Process Flow Diagram for a 300 T/D Copper Smelter
Using Pima Concentrate Plus Pyrite Concentrate
-------
-82-
TABLE XVI. Bases Used in Preparing Flow Diagrams
Morencl Pima
Smelter Smelter
1. Copper Production (tons/day) 300 300
2. Feed-Concentrate (tons/day) 1365 1202
3. Feed-Pyrite (tons/day 207 429
4. Drier Temperature (°C) 200 200
5. Concentrate Roaster Temperature (°C) 900 900
6. Pyrite Roaster Temperature (°C) 800 800
7. S 1n Concentrate Converted to SO, in 15 15
Roaster (%) i
8. S In Pyrite Converted to S02 in Roaster (%} 1.6 1.6
9. Leach Acid Concentration (Nj 4 5
10. Excess Acid Used (?) 88 40
11. S in Neutral Roasted Concentrate 62.6 55
Converted to H2S (S)
12. S in Neutral Roasted Pyrite Converted 100 100
to t^S (%)
13. S02 Conversion to Elemental Sulfur (%) 98 98
14. S Content of Blister Copper (wtJ) 0.2 0.2
5. Converter Operation - This includes operation of the converter to form
blister copper, fire refining of the blister copper, casting into anodes,
handling and disposal of the converter slag, collection of the SCL-bearing
gases, cooling and cleaning of the gases in a cyclone and waste heat
boiler.
The products from the process are copper anodes containing the precious
metals; ferric oxide of fairly high purity; pure elemental sulfur from the
S02 recovery unit; and impure elemental sulfur from the neutral roasters.
If the elemental sulfur is to be marketed, the sulfur from the neutral
roasters would require additional purification. Add-on equipment could be
used for the purification at a nominal cost. The ferric oxide should be of
suitable purity for direct marketing since the residue impurities known to
-------
-83-
be present in the original concentrates which may carry over into the iron
oxide are of little significance in iron oxide intended for use in iron
making. The principal residual elements such as zinc, lead, arsenic and
antimony are expected to be of no consequence in iron making which would
be the intended market for by-product iron oxide.
Wherever possible, equipment sizing, utility requirements, capital
costs and labor requirements are based on published data for similar opera-
tions. Where published data are not available, capital costs were estimated
by standard cost estimating procedures. Operating costs were estimated
using the data presented in Table XVII. In general a conscious effort was
made to be conservative in estimating both capital and operating costs.
It was felt that this was necessary because of the lack of pilot and plant-
scale data for some of the major operations. With additional data, it
would probably be possible to reduce both capital and operating costs
significantly.
TABLE XVII. Basis for Estimating Plant Operating Costs
(330 day per year operation)
1. Direct Labor (Including fringe benefits)
A. Operating . $5.00/hr
8. Maintenance - 31 of fixed capital costs
C. Supervision - 181 of operating and maintenance labor
t. General Plant Overhead - 2/3 of direct labor
3. Utilities
A. Power - $0.01/h*
B. Natural Gas - $0.75/1000 ft3
C. Cooling Water - $0.03/1000 gal
0. Process Water - $0.20/1000 gal
E. Boiler Water - $0.75/1000 gal
F. Steam . $0.50/1000 pounds
4. Maintenance Supplies and Parts - 1.5S of fixed capital costs
5. Operating Supplies - 10S of direct labor
«. Pyrite - $Z/ton
7. Taxes and Insurance - 21 of plant cost
B. Depreciation - 15 year-straight line
-------
-84-
NEUTRAL ROASTING SYSTEM
Storage and handling facilities would be required for the copper con-
centrate and the pyrite concentrate. Although drying may not be needed, it
is included as a step in the process in order to minimize-reactions which
may increase the sulfur dioxide content of the neutral roasted flue gas.
Drying of the concentrates would be accomplished in separate direct fired
fluid bed driers. Each drier would be fired with natural gas. Maximum
drier temperature would be 200°C. The off-gas from each drier would pass
through a cyclone for particulate removal and then be vented to the stack.
The dried concentrates would be fed to gas-fired fluid-bed roasters. The
pyrite roaster would operate at 800°C. Residence time in the roaster would
be 30 minutes. Two copper concentrate roasters would be used, each operat-
ing at 900°C. Residence time in the roasters would be one hour. The
reactors would be fired with a rich fuel mixture to reduce the oxygen in
the product gas. The off-gas and elemental sulfur from the roasters would
pass through cyclones, for particulate removal, and then be combined. The
combined gas stream would pass through a waste heat boiler, electrostatic
precipitator and into a sulfur condenser. The liquid sulfur would be col-
lected and granulated. A waste heat boiler for low pressure steam would
be operated in conjunction with the sulfur condenser. The clean off-gas
would be vented to the stack. The SO,, in the off-gas would be approximately
1.5% of the sulfur in the combined feed. The neutral roasted products would
pass from the roasters into a quench tank where they would be cooled and
granulated (if necessary). Quenching could be in water.
Capital and operating costs for the neutral roasting operation are
presented in Tables XVIII and XIX. Equipment, labor and utility require-
ments are based primarily on a paper published in 1968 by Mehta and Okane^28'
entitled "Economies of Iron and Sulfur Recovery from Pyrite." In the esti-
mates, allowances were made for inflation since 1968, higher operating
temperatures of the roasters, and duplicate equipment requirements. Operat-
ing costs were calculated using the figures presented in Table XVII.
-------
-85-
TABLE XVIII. Capital Cost for Neutral Roasting Operation
Cost SlOOO's
Morenci Pima
1. Purchased Equipment Cost - Total* $2,260 $2,160
2. Equipment Installation (25% of Item #1) 565 540
3. Piping Cost (20% of Item #1) 452 432
4. Instrumentation (15% of Item #1) 339 324
5. Electrical (10% of Item #1) 226 216
6. Buildings (20% of Item #1) 452 432
7. Site Preparation (10% of Item #1) 226 216
8. Plant Design and Engineering (25% of Item #1) 565 540
9. Auxiliaries (30% of Item #1) . 678 648
PLANT COST $5,673 $5,508
10. Contingency (15% of Plant Cost) 864 826
TOTAL PLANT COST $6,627 $6,334
11. Plant Start-up Costs 125 125
12. Interest During Construction (10%) 497 475
TOTAL FIXED CAPITAL COST $7,249 $6,934
13. Working Capital (10% of fixed total cost) 725 693
TOTAL CAPITAL COST $7,974 $7,627
*Based on data from Mehta and 0'
-------
-86-
TABLE XIX. Operating Cost for Neutral Roasting Operation*
Cost $1000's/yr
MorenciPima
1. Operating Labor (6 men/shift) $ 250 $ 250
2. Maintenance Labor (3% of fixed capital costs) 217 208
3. Supervision (18% of Items 1 & 2) 84 82
4. General Plant Overhead (2/3 of Items 1-3) 367 360
5. Power (6 kwh/ton feed) 30 30
6. Natural Gas Ij46o 1,400
7. Boiler Feed Water 25 26
8. Maintenance Supplies and Parts 109 104
9. Operating Supplies 55 54
10. Technical Services (lab, etc.) 35 35
11. Pyrite ($2/ton) ^0 246
12. Taxes and Insurance (2% of plant cost) 132 127
TOTAL $2,854 $2,922
13. Contingency (10% of Items 1-12) 285 296
14. Depreciation (15 years) 532 508
TOTAL OPERATING COST $3,671 $3,726
Cost Per Pound Copper Produced $0.0185 $0.0188
*Based on data from Mehta and
-------
-87-
Capital costs for Morenci and Pima concentrates were estimated to be
$7,974,000 and $7,627,000, respectively. Yearly operating costs were
$3,671,000, and $3,726,000, which is equivalent to $0.0185 and $0.0188 per
pound of copper produced.
LEACHING OPERATION
In the leaching circuit the neutral roasted concentrates are reacted
with hydrochloric acid in a series of closed leach tanks to produce H?S
and dissolve the iron. The leaching temperature is 80 to 85°C. The solids
residence time in the leaching tanks is assumed to be three hours. Reaction
temperatures are maintained using low pressure steam from the waste heat
boilers. Mechanical agitation is provided in each leach tank. Liquid hold
up in the leach tanks is about 200,000 gallons for the Morenci system and
130,000 gallons for the Pima system. The hydrogen sulfide produced would
be scrubbed with dilute hydrochloric acid to remove any traces of hydrogen
chloride, cooled to about 50°C and sent to the SO- recovery system (a portion
of the H2S would be consumed for copper recovery). A nitrogen or neutral
flue gas sweep would be used on each leach tank to facilitate H?S handling
in the off-gas system. Nitrogen would be obtained from the oxygen plant
(required for converter operation) and would amount to about one-half the
HpS on a volume basis.
The leach slurry would be filtered through a rotary drum filter. The
washed solids would pass to the converter. The leach solution (plus wash)
would flow to a copper recovery tank where it would be sparged with H?S
to precipitate any dissolved copper and precious metals. Liquid residence
in the tank would be about 30 minutes. The resulting slurry would be
filtered through a primary rotary drum filter and a polishing filter. The
washed solids would be sent to the converter and the clear leach solution
(plus wash) to the acid recovery system.
Estimating the capital and operating costs for the leaching operation
is more subject to question than other sections of the process because the
leaching operation has not been demonstrated on a pilot or plant scale.
-------
TABLE XX. Capital Cost for the Leaching Operation
Cost $1000's
MorenciPimT
1. Purchased Equipment Cost - Total $ 2,700 $2,000
2. Equipment Installation (25% of Item #1) 675 500
3. Piping Cost (25% of Item #1) 675 500
4. Instrumentation (15% of Item #1) 405 300
5. Electrical (10% of Item #1 270 200
6. Buildings (20% of Item #1) 540 400
7. Site Preparation (10% of Item #1) 270 200
8. Plant Design and Engineering (25% of Item #1) 675 500
9. Auxiliaries (30% of Item #1) 810 600
PLANT COST $ 7,020 $5,200
10. Contingency (20% of Plant Cost) 1,404 1,040
TOTAL PLANT COST $ 8,424 $6,240
11. Plant Startup Cost 150 120
12. Interest During Construction (10%) 632 468
TOTAL FIXED CAPITAL COST $ 9,206 $6,828
13. Working Capital (10% of fixed capital cost) 921 .683
TOTAL CAPITAL COST $10,127 $7,511
-------
-89-
TABLE XXI. Operating Costs for the Leaching Operation
Cost $1000's/yr
Morenci Piiria
1. Operating Labor (5 men/shift) 208 208
2. Maintenance Labor 276 205
3. Supervision (18% of items 1 and 2) 87 74
4. General Plant Overhead (2/3 of Items 1-3) 381 325
5. Power 200 140
6. Process Water 30 30
7. Cooling Water 40 49
8. Steam (available from waste heat boilers) —
9. Hydrochloric Acid (0.1% of inventory lost/day) 40 27
10. Maintenance Supplies and Parts 138 102
11. Operating Supplies 57 49
12. Technical Services 100 100
13. Taxes and Insurance 168 125
TOTAL $1,725 $1,425
14. Contingency (15% of Items 1-12) 259 214
15. Depreciation (15 years) 675 501
TOTAL YEARLY OPERATING COST $2,659 $2,140
COST PER POUND OF COPPER $0.0134 $0 0108
PRODUCED
-------
-90-
Equipment costs and labor requirements were estimated from data for various
mineral leaching operations reported in the literature, allowing for inflation
and a chloride-containing system. Capital and yearly operating costs for the
Morenci and Pima systems are summarized in Tables XX and XXI. The capital
costs are estimated to be $10,130,000 and $7,510,000 for the two cases.
The lower capital cost for the Pima system results from the higher acid
concentration used and the reduced volume of liquid which is handled. The
yearly operating costs for the two cases are estimated to be $2,660,000
and $2,140,000, which corresponds to a cost $0.0134/lb of copper for the
Morenci system and $0.0108/lb of copper for the Pima system.
ACID RECOVERY SYSTEM
In the acid recovery system the leach solution (plus wash) is treated
to regenerate the hydrochloric acid and convert the ferrous chloride to
ferric oxide. The system proposed for use is the Lurgi System,'4' which
is used for the regeneration of hydrocloric acid pickle liquor. In the
Lurgi-type system the leach solution is first concentrated in an evaporator.
The concentrated liquor is then fed to a fluid bed reactor where the HC1 and
water are vaporized and the ferrous chloride converted to ferric oxide.
2FeCl2 + 2H20 + 1/202 > 4HC1 + Fe^
The reactor is operated at 800°C and is gas fired. The hot off-gas from the
reactor passes through a cyclone to remove Fe^O- particulates and then to
the evaporator for heat recovery. The cooled gases from the evaporator
pass to an absorber where the HC1 is absorbed to form 4 to 5t± hydrochloric
acid, which is recycled to the leach circuit. The HC1 free gases are then
vented to the stack. The ferric oxide is obtained as chloride-free, free-
flowing, coarse, granular solid having a bulk density of about 150 lb/ft3.
Capital and operating costs for the acid recovery system are summarized
in Tables XXII and XXIII. Operating costs were calculated using the data
from Table XVI. Capital costs for the Morenci and Pima systems are estimated
-------
-91-
TABLE XXII. Capital Costs for Acid Recovery System
Cost $1000's
Morenci n'ma
1. Purchased Equipment Cost - Total* $ 4,100 $ 3,100
2. Equipment Installation (25% of Item#l) 1,025 775
3. Piping Cost (25% of Item #1) 1,025 775
4. Instrumentation (15% of Item #1 615 465
5. Electrical (10% of Item #1) 410 310
6. Buildings (20% of Item #1) 820 620
7. Site Preparation (10% of Item #1) 410 310
8. Plant Design and Engineering (25% of Item #1) 1,025 775
9. Auxiliaries (30% of Item #1) 1,230 . 930
PLANT COST $10,660 $ 8,060
10. Contingency (15% of plant cost) 1,600 1,210
TOTAL PLANT COST $12,260 $ 9,270
11. Plant Startup Costs 200 175
12. Interest During Construction (10%) 920 695
TOTAL FIXED CAPITAL COST $13,380 $10,140
13. Working Capital (10% of fixed capital cost) 134 101
TOTAL CAPITAL COST $13,514 $10,241
-------
-92-
TABLE XXIII. Operating Costs for Acid Recovery System
Cost $1000's/yr
MorenciPima
1. Operating Labor 210 210
2. Maintenance Labor (3% of fixed capital costs) 401 304
3. Supervision (18% of Items 1 and 2) 110 93
4. General Plant Overhead (2/3 of Items 1-3) 483 405
5. Power 50 35
6. Natural Gas 4,950 3,300
7. Cooling Water 60 40
8. Process Water 100 65
9. Hydrochloric Acid (22° Be1) 80 60
10. Maintenance Supplies and Parts 200 152
11. Operating Supplies 72 61
12. Technical Services 40 40
13. Taxes and Insurance 245 185
TOTAL $7,001 $4,950
14. Contingency (10% of Items 1-13) 700 495
15. Depreciation 901 683
TOTAL OPERATING COST $8,602 $6,128
COST PER POUND COPPER PRODUCED $0.0434 $0.0309
-------
-93-
to be $13,500,000 and $10,200,000, respectively. The yearly operating costs
are estimated at $8,600,000 and $6,100,000, corresponding to $0.043 and
$0.031 per pound of blister copper produced.
Acid recovery systems of the size required to treat the plant leach
liquor have not been built to date. There is, therefore, some question as
to scale-up from existing units. In addition, there is only limited operat-
ing data available on the Lurgi system. For these reasons, conservative
estimates were used for both capital costs and operating costs. By a more
detailed study of the Lurgi and other systems such as that of Haveg,'4' it
should be possible to reduce both capital and operating costs significantly.
Natural gas for heating the reactors and evaporators accounts for more
than 60% of the yearly operating costs. By improved design, additional heat
recovery (multiple effect evaporators) etc., it should be possible to reduce
fuel requirements by a significant amount.
S00 RECOVERY SYSTEM
£
The citrate process developed by the U.S. Bureau of Mines for recovery
of S09 involves the reaction of S00 and H0S in an aqueous solution to form
(34) ^
elemental sulfur. '
S02 + 2H2S -> 3S + 2H20
The basic steps in the process are:
1. Washing of the cooled S02-bearing gas to remove particulates and SO.,,
2. absorption of the S02 from the cleaned gas by a solution of citric
acid, sodium citrate and sodium thiosulfate,
3. reaction of the absorbed S0? with H?S in a closed vessel, precipitating
elemental sulfur and regenerating the absorption solution, and
4. recovery of the sulfur by oil melting.
S02 recoveries of 90 to 99% have been reported.^ ' In this study a
recovery of 98% was assumed. A small amount of the S02 is converted to
sulfate in the process. For this study it was assumed that enough fkS
would be required to react with all of the SOp absorbed, ignoring any
possible sulfate formation.
-------
-94-
The citrate process has only been demonstrated on a small pilot-plant
scale. Scale up to the size required for this study may be questionable
until the plant scale operation, which is underway at the Bunker Hill
Smelter, Kellogg, Idaho, has been demonstrated. If a lower S0? recovery
could be tolerated (90 to 95%) then a modified Claus process might be used.
The Claus process has been demonstrated .in large-scale applications and
would have lower capital and operating costs than the citrate process.
Capital and operating costs for S02 recovery using the citrate process
were estimated using Bureau of Mines data reported in 1973^36^ with an
allowance for inflation. It was assumed that the gas flow to the S0?
recovery unit would be approximately 30,000 SCFM with an average S0? con-
centration of about 6%. To account for fluctuations in gas flow the system
was sized to handle 38,000 CFM. Capital and yearly operation costs for
each case under consideration are summarized in Tables XXIV and XXV. Esti-
mated capital costs for the Morenci and Pima systems are $8,270,000 and
$8,620,000, respectively, while yearly operating costs are estimated to be
$2,100,000 and $2,180,000. The corresponding costs per pound of copper
produced are $0.0106 and $0.0110.
CONVERTER OPERATION
The copper-bearing solids from the leach circuit are fed directly to
a converter. This feed material is high in copper (^50 wt%) and low in
iron (510 wt% Fe). The sulfur content (^20 wt%) is also lower than in the
normal converter feed. For this reason, much of the heat required for
converted operation must be supplied by natural gas or other fuel. A portion
of the oxygen required for reaction with the sulfur and burning of the
natural gas will be supplied from an oxygen plant (^20%). The rest will be
supplied by air. The off-gas from the converter will pass through a waste
heat boiler and then through cleaning equipment to remove the bulk of the
particulates and cool the gas to about 300°C. The gases then pass to the
S0? recovery unit.
-------
-95-
TABLE XXIV. Capital Costs for S02 Recovery System*
Cost $1000's
*Based on data from J. B. Rosenbaum, et.
Morenci Pima
1. Purchased Equipment Cost - Total $2,300 $2,400
2. Equipment Installation (25% of Item #1) 575 600
3. Piping Cost (25% of Item #1) 575 600
4. Instrumentation (15% of Item #1) 345 350
5. Electrical (10% of Item #1) 230 240
6. Buildings (20% of Item #1) 460 ' 480
7. Site Preparation (10% of Item #1) 230 240
8. Plant Design and Engineering (25% of Item #1) 575 600
9. Auxiliaries (30% of Item #1) 690 720
PLANT COST $5,980 $6,240
10. Contingency (15% of plant cost) 897 936
TOTAL PLANT COST $6,877 $7,176
11. Plant Startup Costs 125 125
12. Interest During Construction (10%) 516 538
TOTAL FIXED CAPITAL COST $7,518 $7,839
13. Working Capital (10% of fixed capital cost) 752 784
TOTAL CAPITAL COST $8,270 $8,623
-------
-96-
TABLE XXV. Operating Costs for S02 Recovery System
Cost $1000's/yr
MorenciPima
1. Operating Labor (4 men/shift) $ 167 $ 167
2. Maintenance Labor 226 235
3. Supervision (18% of Items 1 and 2) 71 73
4. General Plant Overhead (2/3 of Items 1-3) 309 317
5. Power 20 21
6. Natural Gas 50 52
7. Steam (waste steam available at no charge) — —
8. Process Water 15 16
9. Cooling Water 35 36
10. Chemicals 162 169
11. Maintenance Supplies and Parts 113 118
12. Operating Supplies 46 46
13. Technical Services 60 60
14. Taxes and Insurance 138 144
TOTAL $1,412 $1,454
15. Contingency (10% of Items 1-14) 141 145
16. Depreciation (15 year) 551 575
TOTAL OPERATING COST $2,104 $2,176
COST PER POUND COPPER PRODUCED $0.0106 $0.0110
-------
-97-
The copper leaves the converter at a purity of about 98.8% and contains
about 0.2% sulfur. The blister copper is transferred to a holding furnace
and then to a refining furnace. The copper from the refining furnace is
cast into anodes for further refining. The off-gases from the holding and
refining furnaces are vented to the stack. Approximately 0.1% of the
sulfur in the original concentrates would be present in the off-gas as S0?.
The cost estimates include all operations through the casting of anodes but
does not include subsequent operations.
The slag from the converter contains a small but significant amount of
copper. This slag would be recycled into the concentrator-flotation plant
where the copper would be recovered as part of the copper concentrate.
The steam produced in the waste heat boiler meets, most of the steam
requirements of the entire plant operation. Additional steam capacity
required is minimal.
The capital and operating costs for the converter operation are sum-
marized in Tables XXVI and XXVII. Equipment and operating cost data were
obtained from a 1973 Bureau of Mines publication by Bennett, et al.'38'
Capital costs for the Morenci and Pima systems are $12,200,000 and
$12,700,000, respectively. Yearly operating costs are estimated to be
$6,120,000 and $6,260,000, which is equivalent to $0.0309 and $0.0316 per
pound of anode copper.
TOTAL PROCESSING COSTS
The total estimated capital and yearly operating costs for a 300 ton/day
copper smelter are summarized in Tables XXVIII and XXIX. The capital costs
for the Morenci and Pima cases are $52,000,000 and $46,700,000, respectively.
These estimates include all of the facilities required to convert copper
concentrate into copper anodes.
The yearly operating costs for the two cases are estimated at
$23,150,000 and $20,430,000, which corresponds to a cost of $0.117 per
pound of anode copper for the Morenci case and $0.103 for the Pima case.
-------
-98-
TABLE XXVI. Capital Costs for Converter Operation
Cost $1000's
Morenci Pima
1. Purchased Equipment Cost - Total* $.3,400 $ 3,500
2. Equipment Installation (25% of Item #1) 850 875
3. Piping Cost (20% of Item #1) 680 700
4. Instrumentation (15% of Item #1) 510 525
5. Electrical (10% of Item #1) 340 350
6. Buildings (25% of Item #1) 850 875
7. Site Preparation (10% of Item #1) 340 350
8. Plant Design and Engineering (25% of Item #1) 850 875
9. Auxiliaries (30% of Item #1) 1,020 1,050
PLANT COST $ 8,840 $ 9,100
10. Contingency (15% of plant cost) 1,326 1,365
TOTAL PLANT COST $10,166 $10,465
11. Plant Startup Costs 175 175
12. Interest During Construction (10%) 762 785
TOTAL FIXED CAPITAL COST $11,103 $11,425
13. Working Capital (10% of fixed capital cost) 1,110 1,143
TOTAL CAPITAL COST $12,213 $12,668
*Based on data from Bennett, et.al.^38^
-------
-99-
TABLE XXVII. Operating Costs for Converter Operation*
Cost $1000's/yr
MorenciPiiria
1. Operating Labor (28 men/shift) $1,165 $1,165
2. Maintenance Labor 340 345
3. Supervision (18% of Items 1 and 2) 271 272
4. General Plant Overhead (2/3 of Items 1-3) 1,184 1,188
5. Power 160 162
6. Natural Gas 750 755
7. Cooling Water 30 30
8. Boiler Water 30 30
9. Flux-Converter ($5.50/ton) 91 163
10. Refractory 150 150
11. Maintenance Supplies and Parts 167 171
12. Operating Supplies 178 178
13. Technical Services 100 100
14. Taxes and Insurance 203 . 209
TOTAL $4,819 $4,918
15. Contingency (10% of Items 1-141 482 492
16. Depreciation (15 years) 814 845
TOTAL YEARLY OPERATING COSTS $6,115 $6,255
COST PER POUND OF COPPER ANODE $0.0309 $0.0316
*Based on data from Bennett, et. al.
(38)
-------
-100-
TABLE XXVIII. Capital Costs for a 300 Ton/Day Copper Smelter
Cost $1000's
MorenciPi ma
1. Neutral Roasting Operation $ 7,974 $ 7 627
2. Leaching Operation 10,127 7 511
3. Acid Recovery System 13,514 10 241
4. S02 Recovery Unit 8>270 8>623
5. Converter Operation 12 213 12 668
TOTAL $52,098 $46,670
TABLE XXIX. Yearly Operating Costs for a 300 Ton/Day Copper Smelter
Cost $1000's/yr
Morenci Pima
1. Neutral Roastino Operation $ 3^71 $ 3 726
2. Leaching Operation 2,659 2 140
3. Acid Recovery System 8,602 6 128
4. S02 Recovery Unit 2,104 2 176
5. Converter Operation 6,115 6 255
TOTAL OPERATING COSTS $23,151 $20,425
COST PER POUND OF ANODE COPPER $0.117 $0.103
6. Credit for Sulfur ($20/ton) 3546i 3^35
7. Credit for Fe203 ($10/ton) 1,757 -\ j776
COST PER POUND OF ANODE COPPER $0.091 $0 078
(with credit for S and Fe0
-------
-101-
If credit is allowed for the sulfur ($20/ton) and ferric oxide ($10/ton),
the operating costs are reduced to $0.091 and $0.078 for the Morenci and
Pima cases, respectively.
Although for comparison it would be most convenient to have an operat-
ing, elemental sulfur-producing, state-of-the-art S0? air pollution abate-
ment system integrated with a conventional copper smelter, such a combina-
tion does not, to our knowledge, exist at this time. The extensive study
made by the Fluor Utah Corporation and as authoritatively reviewed by
Swan* ' ' is therefore taken as the basis for reference for this compari-
son. From that study which covers the entire U.S. copper smelting industry,
the installation cost for 90% emissions control for an add-on system result-
ing in elemental sulfur production would amount to $415,000,000 (1970 costs).
The annual operating cost for these additional facilities without credit for
by-product sulfur was reported to be $130,000,000, which would amount to an
additional cost for the production of copper of 4-1/3 cents per pound, based
on this 1970 study estimate. A reasonable addition to this cost of 1 to
1.5 cents per pound should be added to account for inflationary factors
since the 1970 estimate was made. It is likely then that at present and
near-future costs the addition should be in the range of 5 to 6 cents per
pound of copper for emissions control as provided for in the Fluor study.
A similar situation should be projected for overall smelting costs. A 1973
f 38)
U.S. Bureau of Mines reportv ' estimated the overall smelting cost to be
6.4 cents per pound of copper. With anticipated near-future increases in
fuel and labor costs, this figure should probably also be adjusted upward to
a range of 7 to 8 cents. Thus, by combining these cost figures, a figure
of 12 to 14 cents is obtained for present and near-future costs for copper
smelting with equivalent sulfur dioxide air pollution abatement including
recovery of some portion of the sulfur in elemental form. This range can
be compared with the 10 to 12 cents range for the proposed process concept
with a tendency to favor the lower figure. These results indicate that the
proposed process concept has a favorable economic outlook when viewed under
circumstances which appear to be conservative.
-------
-102-
CONCLUSIONS
ADEQUATE PRODUCTION OF HYDROGEN SULFIDE FOR ELEMENTAL SULFUR RECOVERY
IS ACHIEVABLE UITHOUT NEED FOR ANY SPECIFIC CHEMICAL REDUCTANT
The proposed process concept can result in the production of sufficient
hydrogen sulfide from pyritic copper concentrates for reaction with sulfur
dioxide emissions from converters for elemental sulfur recovery without
need for any specific chemical reductant. The process may be able to use
any type of fuel.
THE ECONOMICS OF AN OVERALL SMELTING PROCESS UTILIZING THE PROCESS
APPEARS FAVORABLE
The preliminary economic assessment indicates that an overall smelting
process employing the concept has favorable costs for achieving copper
production with adequate sulfur emissions control. Credit for the iron
oxide and elemental sulfur provide an even more favorable economic picture
at conservative values. These preliminary evaluations suggest an adequate
basis for continued support on the development of the process through the
pilot plant design phase by EPA and industry.
NEUTRAL ROASTING IS A FEASIBLE MEANS OF CONVERTING THE IRON SULFIDE TO THE
ACID-SOLUBLE FORM
The results of this study show that neutral roasting is a chemically
feasible means of assuring the maximum selective solubilization of the iron
sulfide in pyritic copper concentrates without major dissolution of copper.
Neutral-roasting as studied by others for pyrite alone supports the proposed
process as a feasible and practical method for also treating the pyritic
copper concentrates for preliminary removal of a significant fraction of the
sulfur initially as elemental sulfur as well as for converting the remain-
ing iron sulfide into a soluble form.
-------
-103-
HYDROCHLORIC ACID LEACHING IS A PRACTICAL METHOD FOR HYDROGEN SULFIDE
PRODUCTION
Hydrochloric acid leaching of neutral-roasted concentrates is a direct
and practical means for selective dissolution of iron and release of
hydrogen sulfide. Copper, which may be dissolved in minor amounts, is
recoverable to a major extent along with any solubilized precious metals
(silver) by simple treatment of the cooled solution with the hydrogen sulfide,
Hydrogen sulfide can be produced by this means in amounts sufficient for
reaction with all of the sulfur dioxide released in the step of converting
the enriched copper sulfide to blister copper. Known present state-of-the-
art technology is applicable to conduct the hydrogen sulfide-sulfur dioxide
reaction for high-purity elemental sulfur production via the so-called Claus
process or by the U.S. Bureau of Mines Citrate process.
HYDROCHLORIC ACID REGENERATION IS A PRESENT STATE-OF-THE-ART PROCESS
Hydrochloric acid regeneration from acid-ferrous chloride solutions
is regarded as a present state-of-the-art process which does not need demon-
stration at the laboratory scale. Existing processes which are already
employed on a large scale in the steel pickling industry can accommodate
ferrous chloride solutions of any concentration and acidity.
THE BLISTER COPPER WILL BE RECOVERED IN SATISFACTORY YIELD AND QUALITY
The process is essentially a closed one insofar as copper processing is
concerned. Solids are processed by present state-of-the-art methods in
conventional converters, liquids are specially treated for very effective
copper recovery. Dusts would be retained by the usual methods. If anything,
a higher yield of copper would be anticipated because of the avoidance of
the usual reverberatory slag losses. Quality should be essentially no dif-
ferent from the usual blister copper. The small amounts of converter slag
would be recycled to the concentrator for additional cop'per recovery. In
conventional smelting this slag is added instead to the reverberatory
furnaces.
-------
-104-
THE IRON OXIDE AND HYDROGEN SULFIDE DERIVED SULFUR WILL BE OF HIGH QUALITY
The iron oxide by-product produced during hydrochloric acid regeneration
from ferrous chloride solutions is expected to be of high quality and suit-
able for marketing for steel-making. Small amounts of other metal chlorides
which may also be oxidized, such as zinc, principally, and possibly lead,
are expected to be of no consequence in determining the suitability of the
iron oxide. The sulfur obtained via the hydrogen sulfide-sulfur dioxide
is expected to be of the very high purity as represented by that typically
produced via the Glaus process. The sulfur produced in the neutral roasting
operation will be relatively impure but can be purified to suitable quality
by present state-of-the-art processes/26^
SIGNIFICANT ECONOMIES APPEAR TO BE LIKELY FOR THE PROCESS
Although it is felt that the preliminary economic evaluation of the
process concept is adequate to encourage continued consideration and support
for development of an overall process, significant economies appear to be
probable by more intensive consideration of alternative lower-cost fuels and
engineering refinements to assure increased heat economy. The estimates,
for convenience and simplicity at this stage of the investigation, assumed
the use of high-cost natural gas as the fuel.
THE PROPOSED PROCESS IS APPLICABLE IN ANY SULFIDE SMELTER HAVING SUFFICIENT
PYRITE
Given a sufficient supply of pyrite, any operator of a conventional
smelter for copper sulfides concentrates could consider the installation of
the proposed process. Such installation may be more readily incorporated
into a new smelter than in an old one. Space and physical arrangement would
be the principal problems in modifying old smelters to accommodate the
revised process.
-------
-105-
INCREASED COPPER AND PRECIOUS METAL RECOVERY FROM THE ORE BODY SHOULD BE
POSSIBLE
By use of large amounts of pyrite which is normally discarded, the
small but significant amounts of copper and accompanying precious metals
in such discard would now be almost completely recovered at essentially
very low incremental cost. For exploitation of some ore bodies the
increased recoveries may be quite significant.
OTHER BENEFITS IN RECOVERY OF METAL VALUES, RESOURCE CONSERVATION AND
ECONOMICS MAY BE REALIZED
Because of the need for pyrite, a lower-grade ore may be beneficially
utilized. The use of the major slag-forming minerals such as silica and
limestone would be greatly reduced. The handling and disposal of the large
amounts of useless slag would be greatly reduced. The two other major
components of the concentrates, iron and sulfur, would be recovered in
suitable purity such that beneficial utilization could then be anticipated
with significant economic credit to the overall process. The large-volume,
low-strength sulfur dioxide flue gases characteristic of reverberatory
furnace operations would be avoided.
-------
-106-
RECOMMENDATIONS
In view of the demonstrated chemical feasibility, favorable economics
and indicated other significant benefits which appear to be probable by the
operation of the proposed process concept, the following recommendations in
support of continued development of the integrated process appear appropriate:
1. Large bench-scale experimental work should be continued to obtain
needed data on dynamic systems suitable for the design of a pilot plant.
Such work would be done on the steps of neutral roasting, hydrochloric
acid leaching, residual copper precipitation and separation of solid,
enriched copper sulfide and residues from the ferrous chloride leach
liquors.
2. Work jointly with an appropriate architect-engineer-constructor to
assure that information suitable for the design and estimating of the
acid regeneration system is obtained in the continued laboratory studies.
3. Conduct engineering studies to determine the most economical fuel to be
used and equipment refinements which may be justified to assure most
beneficial heat economy.
4. Refine the economic assessment of the overall process.
5. Conduct process engineering and economic comparisons between this
process and other copper smelting processes which exist in the U.S. or
are being considered for the production of elemental sulfur.
-------
-107-
REFERENCES
1. Levy, S. I. and G. W. Gray, British Patent 309,269 (1928), "Treating
Pyrite Residues," and French Patent 668,093 (1928), "Recovering
Copper."
2. Levy, S. I., U. S. Patent 1,746,313 (1930), "Treatment of Copper-Rich
Material," and also U.S. Patent 1,980,809 (1934), "Production of
Ferric Oxide and Other Metal Values from Pyrites."
3. Parsons, H. W. and T. R. Ingraham, "The Hydrogen Sulfide Route to
Sulfur Recovery from Base Metal Sulfides, Part I, The Generation of
H^S from Base Metal Sulfides," June 1970.
4. Reeve, D. A. and T. R. Ingraham, "The Hydrogen Sulfide Route to Sulfur
Recovery from Base Metal Sulfides, Part III, Recovery of Iron Products
from Ferrous Chloride Solution," Canada, Department of Energy, Mines
and Resources, Mines Branch, Ottawa, Mines Branch Program on Environ-
mental Improvement.
5. Cabri, L. J., "New Data on Phase Relations in the Cu-Fe-S System,"
Econ. Geol., 68, pp. 443-54, 1973.
6. Cabri, L. J., et al., Lau. Mineral, ]2_, Pt 1, 1973.
7. Cabri, L. J. and S. R. Hall, "Mooihoekite and Haycockite, Two New
Copper-Iron Sulfides and Their Relationship to Chalcopyrite and
Talnakhite," Am. Mineral, Etf, pp. 689-708, 1972.
8. Cabri, L. J. and D. C. Harris, "New Compositional Data on Talnakhite,"
Econ. Geol., 66, pp. 673-75, 1971.
9. Clark, A. H., "An Unusual Copper-Iron Sulfide," ibid, 65, pp. 590-91,
1970. —
10. Hall, S. R. and E. J. Gabe, "The Crystal Structure of Talnakhite,"
Am. Mineral, 57_, pp. 368-80, 1972.
11. MacLean, W. H., et al., "Exsolution Products in Heated Chalcopyrite,"
Can. I Ea. Sci., 9_, pp. 1305-17, 1972.
12. Schlegel, H. and A. Schuller, Met Kunde. 4£, pp. 421-29, 1952.
13. Merwin, H. E. and R. H. Lombard, "The System Cu-Fe-S," Econ. Geol.,
32, pp. 203-84, 1937.
14. Donnay, G and G. Kullerud, "High Temperature Chalcopyrite," Carnegie
Inst. Washington Year Book, 57^ p. 246, 1958.
15. Roseboom, E. H., Jr. and G. Kullerud, "The Solidus in the System
Cu-Fe-S Between 400° and 800°C," ibid, pp. 222-27, 1958.
16. Brett, P. R., "The Cu-Fe-S System," Carnegie Inst. Washington Year,
62, pp. 193-96, 1963.
-------
-108-
17. Brett, P. R., "Experimental Data from the System Cu-Fe-S and Their
Bearing on Exsolution Textures in Ores," Econ. Geol., 59, pp. 1241-69,
18. Yund, R. A. and G. Kullerud, "Thermal Stability of Assemblages in the
Cu-Fe-S System," J. Petrology, 7_, pp. 454-488, 1966.
19. Von Gehlen, K. and G. Kullerud, "Pyrrhotite-Pyrite-Chalcopyrite
Relations," Carnegie Inst. Washington Year Rnnk. fil. nn IZA.^
20.
21.
22.
23.
1962.
Kullerud, G.
Kullerud, G.
Kullerud, G.
Morimoto, N.
Acta Cryst. ,
, "The CugS5-Cu5FeS4 Join," ibid, 59_, pp. 114-16, 1960.
, "The Cu-S System," ibid, 59, pp. 110-11, 1960.
, "The Cu-Fe-S System," ibid, 63, pp. 200-202, 1964.
, "Structures of Two Polymorphic Forms of Cu^eS.,"
17, pp. 351-60, 1964. 5 4
24. Yund, R. A. and G. Kullerud, "The Cu-Fe-S System," Carnegie Inst.
Washington Year Book, 59, pp. 111-14, 1960.
25. Yund, R. A. and G. Kullerud, "The System Cu-Fe-S," ibid, 60, pp. 180-
81, 1961.
26- "Outokumpu Process for the Production of Elemental Sulfur
from Pyrite," Sulfur, No. 50, 1964.
27. Kunda, W., V. N. Mackiw and B. Rudyk, "Iron and Sulfur from Sulfidic
Iron Ores," The Canadian Mining and Metallurgical (CIM) Bulletin for
July, 1968, pp. 819-835.
28. Mehta, B. R. and P. T. O'Kane, "Economics of Iron and Sulfur Recovery
from Pyrites," ibid, pp. 836-845.
29. Watkinson, A. P. and C. Germain, "Thermal Decomposition of Pyrite in
Fluid Beds," Noranda Research Centre, Pointe Claire, Quebec, Canada,
A paper presented at the 100th AIME Annual Meeting, New York City, NY
March 1-4, 1971.
30. Tkachenko, 0. B., A. L. Tseft, and 0. Kunseitov, "Thermal Decomposition
of Pyrite in Vacuum and Subsequent Treatment of the Residue," Tr. Inst.
Met. Obogashch., Akad. Nauk Kaz., SSR, 19, pp. 53-58, 1966.
31. Tkachenko, 0. B., A. L. Tseft, I. G. Dem'yanikov, and V. Ya. Slashchinina,
"Thermal Decomposition of Chalcopyrite in Vacuum and Subsequent Treatment
of the Residue," Tr. Inst. Met. Obogashch., Akad. Nauk Kaz. SSR, 19,
pp. 41-52, 1966. —
32. Isakova, R. A., M. I. Usanovich, N. A. Potanina and L. E. Ugryumova,
(Inst. Met. Obogashch, Alma-Ata, USSR), Izv. Akad. Nauk Kaz.. SSR,
Ser. Khim, ]£, (5), pp. 78-81, 1969.
33. Van Weert, G., K. Mah and N. L. Piret, "Hydrochloric Acid Leaching of
Nickeliferous Pyrrohotites from the Sudbury District," CIM Bulletin,
No. 1, pp. 97-103, 1974.
-------
-109-
34. Ingraham, T. R., H. W. Parsons and L. J. Cabri, "Leaching of
Pyrrhotite with Hydrochloric Acid," Can. Metal. Quart., 11, pp 407-11,
1972. 3 —
35. Thornhill, P. G., E. Wigstal and G. Van Weert, "The Falconbridge Matte
Leach Process," J. Metals. 23, No. 7, pp. 15-18, 1971.
36. Rosenbaum, J. B., W. A. McKinney, H. R. Beard, L. Crocker and
W. I. Nissen, "Sulfur Dioxide Emission Control by Hydrogen Sulfide
Reaction in Aqueous Solution—The Citrate System," Bureau of Mines,
U.S. Department of the Interior, Report of Investigations; 1973,
RI-7774.
37. Harvey, A. E., et al., "Simultaneous Spectrophotometric Determination
of Iron (II) and Total Iron with 1,10-Phenanthroline," Anal. Chem.,
27, pp. 26-29, 1955.
38. Bennett, H. J., L. Moore, L. E. Uelborn and J. E. Toland, IC-8598,
"An Economic Appraisal of the Supply of Copper from Primary Domestic
Sources," U.S. Bureau of Mines Information Circular/1973.
39. PB-208-293, "The Impact of Air Pollution Abatement on the Copper
Industry," by Fluor Utah Engineers and Contractors, Inc., for the
Kennicott Copper Corporation, April 20, 1971.
40. Swan, D., "Study of Costs for Complying with Standards for Control of
Sulfur Oxide Emissions from Smelters," Mining Congress Journal, vol. 57,
No. 4, pp. 76-84, April 1971.
-------
-nn-
TECHNICAL REPORT DATA
(Please read luu/uctiont on the reverse before completing)
1. REPORT NO.
EPA-650/2-74-085-b
2.
3. RECIPIENT'S ACCESSION>NO.
4. TITLE ANOSUBTITLE
Control of Sulfur Dioxide Emissions from Copper
Smelters; Volume II--Hydrogen Sulfide Production
from Copper Concentrates
5. REPORT DATE
September 1974
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
C. A. Rohrmann and H. T. Fullam
9. PERFORMING ORG MMIZATION NAME AND ADDRESS
Battelle Pacific Northwest Laboratories
Battelle Boulevard
Richland, Washington 99352
10. PROGRAM ELEMENT NO.
1AB013: ROAP 21ADC-056
11. CONTRACT/GRANT NO.
68-02-0025
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AAIO PERIOD COVERED
Final; 6/73-4/74
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report gives results of a laboratory study of the control of SO2 emis-
sions from copper smelters by H2S production from copper concentrates. Digestion
of neutral roasted pyritic copper concentrates with HC1 was studied as a means of
producing sufficient H2S for reaction with the SO2 from converter gas to produce
elemental sulfur and then minimize SO2 emissions from copper smelters. In this
step the copper sulfides are maintained in insoluble form. A large fraction of the
iron was shown to be solubilized with equivalent production of concentrated H2S at
relatively low temperatures. In such a process a pure form of iron oxide would be
produced as a by-product during acid regeneration. In the integrated scheme, diges-
tion would likely be the only step which is not a state-of-the-art process. An econ-
omical overall process appears probable to yield elemental sulfur and also pure iron
oxide as useful by-products. Costly reverberatory furnaces would be eliminated.
Needs for slag forming minerals would be greatly reduced. With efficient utilization
of excess pyrite, increased recovery of copper and precious metals from the ore
body should also be realized. Further laboratory studies are recommended in pre-
paration for pilot plant investigations.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Air Pollution
Sulfur Dioxide
Copper Ores
Smelters
Smelting
Hydrogen Sulfide
Hydrochloric Acid
Air Pollution Control
Stationary Sources
Pyritic Copper
13B
07B
11F
13H
8. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
117
Unlimited
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |