EPA-650/2-74-085-a
SEPTEMBER 1974
Environmental  Protection  Technology Series
                                                   x't'
                                                  ,-''.-'.

-------
                                                          )0

                                       EPA-650/2-74-085-0
CONTROL  OF SULFUR DIOXIDE  EMISSIONS
          FROM COPPER  SMELTERS:
      VOLUME  I  - STEAM  OXIDATION
    OF PYRITIC COPPER CONCENTRATES
                         by

           C.A. Rohrmann, H.T. Fullam, and P.P. Roberts

              Battelle Pacific Northwest Laboratories
                    Battelle Boulevard,
                 Richland, Washington 99352

                  Contract No. 68-02-0025
                 Program Element No. 1AB013
                   ROAP No.  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

-------
CONTROL OF SULFUR DIOXIDE EMISSIONS FROM
   COPPER SMELTERS BY STEAM OXIDATION
     OF PYRITIC COPPER CONCENTRATES
              Prepared for
   THE ENVIRONMENTAL PROTECTION AGENCY
       CONTROL SYSTEMS LABORATORY
     DEVELOPMENT ENGINEERING BRANCH
                   and


      THE PHELPS DODGE CORPORATION



    Under Contract Number 68-02-0025
                June 1972
                BATTELLE
     Pacific Northwest Laboratories
     Chemical Technology Department
     .  RICHLAND, WASHINGTON  99352

-------
                                  CONTENTS
                                                               Page No.
ABSTRACT 	  1
INTRODUCTION 	  2
   1.  Copper-Bearing Minerals and Their Relationship to
       Study	3
   2.  Previous Work	4
   3.  Objective of the Study	4
PROJECT TASK SUMMARIES 	  4
   Task I   - Background Information 	  4
   Task II   - Laboratory Investigations	5
   Task III - Economic Assessment of the Process	6
   Task IV   - Identification of Additional  Work	6
SECTION 1	6
   Background Information and Evaluation 	  6
SECTION II	11
   Experimental Laboratory Investigations	11
      1.  Copper Concentrates	11
      2.  Equipment	12
             a.  Steam Oxidation	12
             b.  Desulfurization	18
      3.  Analytical Methods	19
      4.  Experimental Results	21
             a.  Differential Thermal Analysis of
                 Concentrates	21
             b.  Labile Sulfur Removal	23
             c.  Hydrogen Sulfide Yields	27
             d.  Reaction Kinetics	41
             e.  Fate of Concentrate Impurities	46
      5.  Interpretation of Experimental Results	47
             a.  Process Limiting Factors	47
SECTION III	49
   Economic Assessment of the Process Concept	49
      1.  Basis	49
      2.  Material Balances	49
      3.  Major Equipment Sizing	50
      4.  Capital Cost    Comparison	52

-------
CONTENTS (contd)
                                                               Page No.
      5.  Comparison of Fuel and Depreciation Costs	53
      6.  Estimate of Lower Limit of Hydrogen Sulfide
          Concentration in Steam Leaving Oxidizer	53
SECTION IV	54
   Recommendations	54
SECTION V	55
   Identification of Additional Work Required	55
      1.  Suggested Additional Work	56
      2.  Impact on Conventional Smelting Practices	57
      3.  Advantages	58
REFERENCES	60
APPENDIX	63
   World Primary Copper Smelters and Smelting Capacity	64

-------
                           LIST OF FIGURES
FIGURE 1.


FIGURE 2.

FIGURE 3.


FIGURE 4.


FIGURE 5.

FIGURE 6.

FIGURE 7.


FIGURE 8.



FIGURE 9.


FIGURE 10.


FIGURE 11.


FIGURE 12.



FIGURE 13.
                                                  Page No.

Diagram of Experimental Setup Used in Steam
Oxidation Tests	15

Steam Oxidation Reactor and Furnace	16

Differential  Thermal Analysis of Air-Dried
Copper Concentrates in Helium	22

Differential  Thermal Analysis of Air-Dried
Concentrates  Heated in Air	24

The Iron-Sulfur System (Rosenqvist) 	25

H2S Yield as  a Function of Reaction Time	29

Effect of Reaction Temperature on the
Equilibrium H~S Yield (Experimental)	30

Comparison of Experimental and Calculated
(Thermodynamic) H^S Yield at Various
Reaction Temperatures	35

Calculated Equilibrium Gas Composition as
a Function of Reaction Temperature	40

H?S Yield as  a Function of Reaction Time
for a Batch Reactor	43

H?S Yield as  a Function of the Pyrrhotite
Reacted for Various Water Flow Rates	44
Contact Time Required to Reach the Equilibrium
Hydrogen Sulfide Concentration as a Function of
the Pyrrhotite Reacted - at 800C 	
.45
Steam Oxidation Process Material Balance Flow-
sheet 	51

-------
                           LIST OF TABLES
                                                               Page No.

TABLE 1.   Description of Copper Concentrates Used
           in the Steam Oxidation Studies	13

TABLE 2.   Weight Loss of Concentrates on Vacuum
           Drying at 110C	13

TABLE 3.   Chemical Analyses of Copper Concentrates
           Used in Steam Oxidation Studies	14

TABLE 4.   Equilibrium Hydrogen Sulfide as a Function
           of Temperature	32

TABLE 5.   Calculated Equilibrium Gas Compositions	34

TABLE 6.  Equilibrium Constants as a Function of
           Reaction Temperatures	39

-------
                       CONTROL OF SULFUR DIOXIDE
                   EMISSIONS FROM COPPER SMELTERS BY
            STEAM OXIDATION OF PYRITIC COPPER CONCENTRATES
                               ABSTRACT

     This project to study the chemistry, engineering and economics of
steam oxidation as a means for the large-scale, low-cost preparation
of hydrogen sulfide for use as an intermediate in the production of
elemental sulfur from the gaseous effluents from conventional copper
smelters was initiated in June of 1971 by means of a contract between
Battelle and the sponsors, the Environmental Protection Agency and the
Phelps Dodge Corporation.  The laboratory studies showed that hydrogen
sulfide was produced in relatively high yield and rapid rate from the
iron sulfide obtained in a first step of neutral  or non-oxidizing roast-
ing of pyritic copper concentrates.  In addition, the copper sulfides
were found to be non-reactive with the high temperature (700-800C)
water vapor.  However, equilibrium conditions limit the concentration
of hydrogen sulfide to less than about 1% in the water vapor.  No prac-
tical means could be foreseen which would significantly increase this
concentration.  At such a concentration the energy requirements needed
for the supply of high-temperature water vapor assures that the pro-
cess could not compete economically with a more conventional process.
A change in scope of this project to study alternate means of hydrogen
sulfide production without need for a reductant is recommended.  An
aqueous process based on the use of regenerable hydrochloric acid which
still maintains the copper in the form of an enriched sulfide concen-
trate is the preferred route proposed for further study.

-------
                                  -2-

                             INTRODUCTION

     The conventional processes for smelting copper sulfide ores and con-
centrates involve successive steps or combinations of roasting in air to
remove part of the sulfur, melting and iron slagging in reverberatory
furnaces to yield fused copper-iron sulfides (matte) and reduction of the
matte to blister copper in  converters.   In all of these steps the sulfur
in the original ores and concentrates is released as sulfur dioxide gas
at various dilution levels from which it is not readily recoverable, par-
ticularly at low concentrations.  Furthermore, once recovered, usually as
sulfuric acid from the waste gases of higher concentration, this low-valued
compound has limited use and marketability.
     For many years it has been known that water vapor at high temperatures
can also oxidize sulfides to yield a large fraction of the sulfur in the
form of hydrogen sulfide.  However, in contrast with sulfur dioxide, hydro-
gen sulfide is very readily concentrated and easily converted to elemental
sulfur by means of existing, coirmon, large-scale technology.  Furthermore,
with hydrogen sulfide at hand, it can also be made to react very readily
with sulfur dioxide to produce elemental sulfur.
     In the past few years Battelle supported preliminary laboratory inves-
tigations of the steam oxidation process as it might be applied to a number
of mineral sulfides.  Its application to representative and typical  pyritic
copper concentrates appeared to be unique.  Although the removal  of sulfur
via conversion to hydrogen sulfide was not found to be complete,  it was suf-
ficiently high to justify further intensive study.
     In the conventional (Claus) process for the conversion of hydrogen
sulfide to elemental sulfur, about one-third of the hydrogen sulfide is
burned to form sulfur dioxide, or the latter may be provided from other
sources.  These two gases are then reacted catalytically to yield elemental
sulfur.  The Battelle laboratory work indicated that the steam oxidation
reaction was probably complete enough to provide sufficient hydrogen sul-
fide for reaction with the residual sulfur dioxide which would be expelled
in subsequent smelter operations.

-------
                                  -3-

     In June of 1971 Battelle was awarded a contract to conduct research on
the steam oxidation concept for the Environmental Protection Agency, Control
Systems Laboratory, Development Engineering Branch.  The Phelps Dodge Corpo-
ration, the next to the largest producer of copper from domestic sources in
the U.S., was a joint sponsor of this project.
     Although other means could be considered for the production of the de-
sirable hydrogen sulfide, such processes generally require the use of a re-
ductant such as natural gas, hydrogen or coke for direct reaction with sulfur
or with sulfur dioxide.  The steam oxidation concept is unique in that hydro-
gen sulfide may be produced directly from processed concentrates without the
need for a reductant.  However, the energy requirements for production of
the high-temperature water vapor were recognized as the principal unknown
which would have an important bearing on the realization of a practical pro-
cess.
1.  Copper-Bearing Minerals and Their Relationship to this Study
         Although there are sources of copper existing as so-called oxidized
    copper in the form of oxides, silicates, carbonates and sulfates, by far
    the most important copper minerals are the sulfides.  Two of the principal
    copper sulfide mineral resources are chalcopyrite and chalcocite.  These
    are usually recovered as concentrates which are mixtures with varying
    amounts of pyrite, iron disulfide.  However, some of the chalcopyrite
    concentrates (the most important copper mineral) may be quite low in
    pyrite (10% or less).  There are also a few very important commercial ore
    bodies that are extremely complex in their copper sulfide mineralization
    and comprise, in addition to those minerals mentioned above, significant
    fractions of bornite, enargite, tetrahedrite, covellite, digenite and
    tennantite.  In addition, economically significant amounts of precious
    metals (gold and silver) and molybdenum are also present in some pyritic
    copper sulfide concentrates.  By way of clarification, the identifica-
    tion "pyritic copper sulfides" is meant to cover all copper sulfide con-
    centrates in which iron sulfide exists in significant amounts.  This
    iron sulfide content would include pyrite as well as any iron sulfide
    which is bound up in other important sulfide mineral forms such as in
    calcopyrite and bornite.  On the other hand, if a concentrate were pure
    chalcocite, it could not be classed as a pyritic copper sulfide because

-------
                                  -4-

    pure chalcocite would contain no iron sulfide.  However, most chalcocite
    concentrates contain rather high levels of pyrite with some of this being
    in very close mineralogical association.  The only exception in the U.S.
    is the White Pine concentrate of the Copper Range Company which is char-
    acteristically very low in pyrite.
         To provide reasonably good coverage of the major pyritic copper con-
    centrates and their behavior in the steam oxidation process requires that
    most of the work be concentrated on representative chalcopyrite and high
    pyrite chalcocite concentrates.  In addition, a few confirmatory tests
    were conducted on concentrates of more complex mineralization and on con-
    centrates in which precious metals  (silver) and molybdenum are significant
    by-products to be certain that the  process will not adversely influence
    the ultimate recovery of these values.
2.  Previous Work
         Although extensive previous work on the steam oxidation of pyrite and
    neutral-roasted pyrite (pyrrhotite) was noted, essentially no such work
    directed at similar treatment of pyritic copper concentrates was uncovered.
3.  Objective of the Study
         It was the objective of this project to investigate the chemistry and
    to appraise the kinetics and engineering and economic feasibility of the
    steam oxidation process as a means of air pollution abatement for copper
    smelters.

                        PROJECT TASK SUMMARIES
Task I - Background Information
1.  The concept of oxidation of iron sulfide with high-temperature water
    vapor is old as evidenced by literature references dating back to as early
    as 1836.  Many expired patent references to processes based on the reaction
    exist.
2.  The references do not provide an adequate basis for assessing the real
    technical and economic feasibility  of a practical application of the
    reaction, particularly to typical pyritic copper concentrates.
3.  The use of hydrogen sulfide to react with sulfur dioxide is a well-known
    and extensively applied means for the large-scale production of elemental
    sulfur.

-------
                                  -5-

Task II -Laboratory Investigations
1.  This study has shown that hydrogen sulfide can be produced from the
    iron sulfide in copper concentrates by exposure to water vapor, pref-
    erably in the range of 700 to 800C.
2.  Essentially all of the iron sulfide can thus be converted to magnetite
    without attack on the copper sulfide.
3.  It was observed, however, from the laboratory studies backed up by
    thermodynamic considerations that the  concentration of hydrogen sulfide
    in the off-gas will never exceed about 1% by volume.  Although slightly
    higher concentrations would be expected at temperatures higher than 800C,
    sintering of the concentrate mixtures  is encountered at which time a
    rapid fall off of reactivity was observed.
4.  Although the equilibrium yields would  be expected to be similar for any
    composition involving water and pyrrhotite, significant differences in
    yields for concentrates from various  sources were shown in the results.
    It is concluded that the reactions are more complex than anticipated.
    At this time the exact nature of these reactions is not known.
5.  On the basis of the present laboratory and thermodynamic studies, no
    means are apparent to assure the achievement of an improvement in reaction
    gas concentration required to assure  a competitive steam oxidation process.
6.  The reaction rate appears to be quite  rapid (even after 90% of the
    pyrrhotite has been reacted, a contact time of only 3 seconds at 800C
    is required to achieve the equilibrium ^S concentration);  however,
    the low equilibrium composition makes  it essentially impractical to take
    advantage of this rapid rate since a  reactor of very large cross-sectional
    area with only a small depth of concentrate would appear to be required.
7.  Although some concentrates appear to  have too low an iron content to
    provide sufficient hydrogen sulfide,  we do not regard this as a general
    problem because pyritic copper concentrates are prepared from ores
    generally high in pyrite.  This excess pyrite, over that required for the
    production of a preferred matte composition, is rejected in the flota-
    tion circuit.  Most of such pyrite could be retained if there were a
    need for itsuch as to assure sufficient hydrogen sulfide production.

-------
                                  -6-

Task III - Economic Assessment of the Process
    Preliminary economic studies made on the basis of the laboratory results
show that construction and operating costs are not competitive with a conven-
tional natural gas-SCL reduction process without about a sixty percent in-
crease in hLS concentration in the reaction off-gas.
Ta^sk IV - Identification of Additional Work
    Steam oxidation as studied here appears to hold little or no prospect as
the means of economic hydrogen sulfide production on which to base a practi-
cal air pollution control process for copper smelters.  No additional work
on the steam oxidation concept is suggested.  However, the study has provided
a firm basis on which more attractive alternative concepts for hydrogen sul-
fide production may be developed.  Continuation of the work to achieve adequate
hydrogen sulfide production by reaction of neutral-roasted pyritic copper con-
centrates with a non-oxidizing, regenerable acid such as hydrochloric acid is
proposed.  The extensive work already done under this study with representa-
tive concentrates, the understanding of their primary characteristics and the
apparent simplicity and near state-of-the-art technology of most of the process
steps to be involved in this alternative concept supports the suggested con-
tinuation of the work.

                               SECTION I
Background Information and Evaluation
    The reaction of steam with metallic sulfides has been a subject of chemical
as well as industrial interest for many years.  Regnault^ ' about 135 years
ago studied the reaction in a qualitative manner for a number of metal  sul-
fides.  In this work Regnault clearly observed the distinctive difference
in the reactivity of iron sulfide and copper sulfide with the former being
more easily reacted to yield hydrogen sulfide, hydrogen and magnetic iron
oxide.  No equations, rates of reaction or temperatures other than "red heat"
                                       (2}
or "white heat" were reported.  Gautierv ' 70 years later acknowledged the
iron sulfide-water vapor reaction to be as follows after observing that the
ratio of hydrogen sulfide to hydrogen was about 3:1
                3FeS + 4H0   -   Fe0  + 3H$ + H

-------
                                  -7-
He also observed that copper sulfide was reduced to metal  at high tempera-
ture with formation of sulfur dioxide rather than hydrogen sulfide.   Tem-
perature measurements were not specified.
    In the following 40 years (1910-1950) there were some widespread efforts
at gaining a better understanding of the steam oxidation reaction; however,
a clear realization of the thermodynamics, equilibria, and kinetics  suitable
for determining whether or not realistic applications were possible  does not
appear to have resulted from these works.  This period was also a time of
rather intense patent activity relating to the steam oxidation concept.
    Likewise in this period the patents also indicated a great deal  of ac-
tivity aimed at the commercial applications of the steam oxidation reaction
with iron sulfides for the production of hydrogen sulfide and as a means for
the recovery of elemental sulfur.
    In an early patent (U.S. 987,156), McLarty^ ' proposed the conversion of
nickel sulfide ores to oxides by means of pressurized steam.
    The Hall process and patents^ '*' almost exclusively related to the
techniques and devices to produce hydrogen sulfide and sulfur from pyrite
via some combination of what is now called neutral roasting and steam oxida-
tion.  The Bacon patents^ ' emphasized the advantage in hydrogen sulfide
production achievable by the introduction of alkaline material to the ores
being oxidized by steam.  Bacon also includes the reaction between sulfur
                                        fa\
dioxide and pyrite.  The patent by Leesev ; mentions that at the preferred
operating temperature the steam oxidation reaction is complete in a  few
seconds.
                                                    (g\
    As late as 1958 deJahn was granted a U.S. patent^ ' that involved treat-
ment of pyrite with "very hot steam" to yield sulfur and iron oxide.
    Among the patent literature there is also frequent reference to  the use
of steam along with oxidizing, reducing or acidic gases to achieve specific
separations or conversions with metal sulfides generally.   It was felt that
these were not sufficiently relevant to the principal reactions and  objec-
tives being studied in this project and thus were not detailed in this
review.  References of Moltereck,^10^ Arnold,(11^ Hooten,(12) Perkins,(13)
Petersen,(14) Lindemann,(15) Goetz,(16) Horakova,(17) Birkeland,(18) and
Millar^  ' are representative of these.

-------
                                    -8-

      The patents by Hall,  Bacon,  and Leese and others  are all  of interest and
 provided a basis for optimism for the realization  of a process;  however,
 there does not appear to  be any direct commercial  application  of any  of  these
 inventions.   Most of these patents are repeated in patents granted  in  foreign
 countries.   Thus, although the body of patent  literature  is  quite large,
 there is much repetition with the same inventions  covered by patents  in more
 than  one country.
      After about 1940 there  began to appear some fairly detailed theoretical
 and laboratory studies  of  the reaction.(20~29)  Uno^28^ conducted a mathema-
 tical  study  of the  equilibria for the reaction:
                    3 FeS + 4  H20  + Fe304 +  3 H2S +  H2
 The results  indicated that an  H2S concentration of  about  1%  would be expected
 at 800C.   Frey^     also studied  the same  reaction  from the  standpoint of
 thermodynamics  data  and calculated  that 1.8% of the water  as required by  the
 equation  would  be converted to  hydrogen sulfide and hydrogen at  equilibrium
 at a  temperature  of  800C.
      Much earlier Thompson and Tilling^  '  did laboratory  work on the de-
 sulfurization of  pyrite with  steam,  other gases and mixtures of  gases.
 Although  these  authors established  that the reaction of interest began at
 580C and that  it continued at higher  temperatures, they also concluded that
 steam was catalytic  in promoting  the  decomposition of pyrite at  lower tem-
 peratures.  Their results  also indicated that the steam -  FeS reaction was
 rapid.  Quantitative, and  kinetic data were not presented.
     The work of Kirillov  and Makarov^  ' was more extensively experimental.
Their data also indicated  that maximum yield or conversion of sulfur from
the FeS to H2S  required many times  the stoichiometric amount of steam.  Al-
though their methods of analysis were not reviewed in detail their results
 indicated favorable  practical prospects for the reaction.   The  H2S content
of the gas was  high.  The H2S to  S02 ratio was also generally high.   In fact,
most of such data appeared to yield higher concentrations  than  one would
normally expect.  The gas analysis data appeared to be reported on a water-
free basis which would emphasize  these high figures.
                (24)
     Rama Murthiv  ' also did experimental  work but with limited results
reported, such as a one-hour exposure of pyrite to hydrogen mixed with steam

-------
                                    -9-
 at  600C yielded  a  69%  recovery of the sulfur  in the form of both
 and S02.
                    /or  OC\
      Inaba and  Konov   '    did extensive laboratory study and thermodynamic
 calculations on the reactions involved.  They  observed that although FeS
 could be completely desulfurized with water vapor, the thermodynamics were
 not very favorable.  However, it appears that  they also anticipated practi-
 cal applicability,  if improvements could be made in equipment and techniques.
     No quantitative data on the reaction of steam with copper sulfide
 minerals have been reported although Parlr  '  observed very slow reactions
 taking place between steam and various copper  minerals at temperatures in
 the range of 200-460C.
     Studies of the thermodynamics and kinetics of the reaction of steam with
 iron sulfides have been made but no equivalent studies for copper sulfide
minerals at the temperatures of interest were  found.
                  (29)
     Diev, et al.,v  ' using steam and oxygen  in the  converting step for
copper smelting, released up to about one-third of the total  sulfur in the
elemental  form.   They studied the reactions thermodynamically using pub-
lished data to determine the equilibrium composition  of the gases.   Since
fluid silica slags as well  as molten iron and copper  sulfides were  involved
at temperatures  far above those where steam oxidation  is  known to proceed,
these results do not appear to have a direct relationship to  the  steam oxida-
tion concept as  being studied under this  present project.
     In  the application of a steam oxidation process  for  the  conversion of
pyritic  materials  to hydrogen sulfide, it is probable  that a  preliminary
step involving simply the thermal  decomposition of pyrite to  yield  elemental
sulfur and pyrrhotite (FeS) would be adopted.   This step  is generally termed
"neutral  roasting," since it does not involve  oxidation of the  pyrite or
pyrrhotite to yield sulfur  dioxide.   Also,  since it does  involve  essentially
present  state-of-the-art technology,  no exhaustive  effort  was  made  to review
the extensive background literature.   A few references, however,  provide some
useful information and contribute to an understanding  of  this  well-known pro-
cedure.
                  (32)                   (33)
     Kunda, et al.     and  Mehta, et al.      provide extensive.data on the
large-scale thermal  decomposition and economics of  pyrite  processing.  These

-------
                                 -10-

two papers are concerned with the hydrometallurgical approach to the problem
of ultimate iron and sulfur recovery from neutral-roasted pyrite.
     Graves and Heatlv  ' also covered the economics of sulfur recovery from
the large-scale processing of pyrite including the step of neutral roasting.
     It should be noted that all of these latter references were presented
at a time of sulfur shortage and high sulfur prices in the U.S. and Canada.
This situation has decidedly changed since that time.
                                                 (35)
     Another useful reference is that of Kassila.   '  In this paper the
flash smelting process as developed by Finland's Outokumpu Company is de-
scribed.
     Also included is a description of the process as used at the Kokkola
plant to produce elemental sulfur via the thermal decomposition of pyrite.
This process was used on a large scale for at least about five years in the
mid-19601s.  Although similar processes have been used and proposed for use
throughout the world, this plant probably represents the most recent practical
application of the thermal decomposition of pyrite for large-scale elemental
sulfur production.  Guccione^  ' provides additional details on the process
used at the Kokkola plant.
     Very little of the works cited above are strictly relevant to the present
study which is focused on the reaction between steam and the principal copper
sulfide material concentrates of the U.S.  Most of the previous work deals
with reactions of the iron sulfides alone.  While it is true that the expected
reaction to produce hydrogen sulfide from copper concentrates will be with
the pyrite and probably also the copper-iron sulfides present, the composi-
tion of the concentrates vary and the effect of the various components on
the steam oxidation are unknown.
     In prior unpublished work at Battelle-Northwest, tests were carried out
to demonstrate the potential of the steam oxidation reaction to produce hydrogen
sulfide using representative minerals and copper concentrates.  Included in
these tests are pyrite, copper sulfide flotation concentrates, zinc sulfide
(sphalerite or marmatite), tetrahedrite and galena.  The data can be sunmarized
up to the present time on the results of these tests as follows:

-------
                                 -11-

     1.  Pyrite after conversion to pyrrhotite by the usual heating step
         in the absence of air (neutral roasting) reacted readily with water
         vapor at atmospheric pressure and at about 850C.  The reaction
         appears to follow that described by Thompson and Tilling'^0', as
                   3FeS + 4H20 + Fe.,0. + 3H2S + Hp.
         Sulfur and hydrogen sulfide were determined in the off-gas.  Hydro-
         gen was not sought.  Sulfur dioxide was absent.  The pyrrhotite
         (FeS) was not noticeably magnetic.  The solid residue was highly
         magnetic and was assumed to be mainly magnetite, Fe^O^.  Weight
         changes, without resort to chemical analyses, indicated about a
         90% conversion.  Chemical analysis subsequently obtained showed
         that 97.5% of the sulfur in the pyrite sample had been removed.
     2.  Similar results were obtained with the copper concentrate; however,
         it was concluded that only the iron sulfide fraction reacted.  Cop-
         per sulfide is not considered to be reactive with steam at the con-
         ditions of the tests.
     3.  Zinc sulfide was unreactive under similar conditions.
     In the neutral-roasting step, which would probably be used as a pretreat-
ment step to convert the pyrite to pyrrhotite and thereby also remove volatile
contaminants as well as part of the sulfur as elemental sulfur, some rather
direct and uncomplicated reactions can be visualized, such as the following
for the decomposition of chalcopyrite, the most conmon copper mineral:
                   2 CuFeS2 ->  2 FeS + S + Cu2S
     However, in view of the complex thermochemical mineralogy of the copper-
iron-sulfur system, such a desirable result for chalcopyrite or other copper-
iron-sulfur mineral decompositions is probably never obtained.

                              SECTION II

Exjaer i men t a 1 L,a bg_rat ory Investigations
1.  Copper Concentrates
         The steam oxidation studies were carried out using copper concentrates
    recovered from pyritic copper deposits.  Most of the work was concentrated
    on representative chalcopyrite and chalcocite-type copper concentrates
    containing significant amounts of pyrite.  In addition, a few confirmatory
    tests were conducted with concentrates of more complex mineralization and
    on concentrates containing precious metals to be certain that the steam
    oxidation process will not adversely affect the ultimate recovery of these

-------
                                 -12-

    values.  Table 1 gives a description of the six concentrates selected
    for use in this study.
         The six selected concentrates were obtained from the various copper
    producers.  About four pounds of each were vacuum dried 24 hours at 100C.
    The vacuum drying resulted in weight losses varying from ^0.5% to over
    11% for the different concentrates as shown in Table 2.
         The vacuum-dried material was carefully mixed to minimize hetero-
    geneity, sampled and stored in sealed bottles for use in the bench scale
    experiments.
         The samples were analyzed for Cu, S, Fe, insolubles and other metals.
    The results of these analyses are given in Table 3.
2.  Equipment
    a.  Steam Oxidation
            The steam oxidation studies were carried out by passing steam in
        a helium carrier gas through a heated bed of the copper concentrate
        and analyzing the reaction products.  A schematic diagram of the ap-
        paratus used for most of the experiments is shown in Figures 1  and 2.
            High-purity helium was used as a carrier gas and was passed over
        copper metal  turnings to remove traces of oxygen.
            The steam generator consisted of a Vycor vessel  heated  by a small
        furnace to approximately 400C.  Water was injected into the vessel
        by a syringe pump capable of highly reproducible water flow rates.
        The pump was calibrated at all  of the flow rates used in the experi-
        ments by weighing water collected over a measured pumping time  interval,
        In some of the early experiments water vapor was injected into  the
        reactor by bubbling the carrier gas through a series of gas purifier
        bottles containing water at a constant known temperature and calculat-
        ing the water content from vapor pressure data.   The latter technique
        was satisfactory only for Tow steam flow rates.   In  both methods the
        lines leading to the reactor were held at ^200C by  heating tapes.
            The reactor consisted of a  small  Vycor vessel  50 mm I.D.  x  50
        mm high with a porous quartz frit on the bottom to support  the

-------
                                 -13-
           TABLE 1.   Description of Copper Concentrates  Used
                     in the Steam Oxidation Studies
             Source
1.  Pima Mining Co., Arizona,via
    Phelps Dodge Corporation

2.  Morenci, Arizona, Phelps
    Dodge Corp.

3.  Tyrone, New Mexico,
    Phelps Dodge Corp.

4.  Lavender Pit-Bisbee, Arizona,
    Phelps Dodge Corp.
5.  Battle Mountain, Nevada,
    Duval Corp.

6.  Butte, Montana,
    Anaconda, Co.
   Principal  Minerals

Chalcopyrite
Pyrite & Chalcocite
Pyrite & Chalcocite
   Comments
Low in pyrite,
contains silver

Contains Mo
Contains Mo
Complex CuFe sulfides
(bornite, chalcopyrite)
and pyrite

Chalcopyrite,Chalcocite,  High in pyrite
pyrite                    and silver
Very complex CuFe
Sulfide minerals
including enargite
Contains As, Zn,
Cd, Bi and Pb
    TABLE 2.  Weight Loss of Concentrates on Vacuum Drying at 110 C
               Concentrate Source

               Pima
               Morenci

               Tyrone
               Lavender Pit
               Battle Mountain
               Anaconda
     weight Loss, %

           6.7

           2.6

           1.4
           0.4

          11.4 .

           0.8

-------
TABLE 3.  Chemical  Analyses of Copper Concentrates Used in  Steam Oxidation  Studies

Source Cu
Pima i 27.14
Morenci

Tyrone
Lavender
Pit
Battle
Mtn.

20.82

20.31
12.65

24.50


Anaconda j 26.93
i
Weight %
S Fe Insol As Sb Zn
29.6 27.3 10.16 0.06 0.025 0.3
36.2 27.3 9.12 0.06 0.025 1.5

40.0 30.1 6.48 0.39 0.05 0.3
39.7 32.0 8.88 0.36 0.03 1.8

31.3 25.0 13.12 0.06 0.063 0.70


34.7 19.4 6.64 1.86 0.037 6.2


Pb
0.04
0.04

0.04
0.20

0.03


0.05


Mo
0.020
0.030

0.085
0.025

0.018


0.05


Bi
0.016
0.008

0.012
0.024

0.020


0.03

i
Cd
0.002
0.001

0.001
0.004

0.003
Se
380
320

280
360

510

i
0.12
290
i
PPM
Te
42
68

58
110

64


250

.Troy Oz/Ton
Hq ! Ag
280 ! 2.8
280

310
1.82

1.54
240 1.94
i
380

10.79

1
240 10.42
1
Au
00.010
. Q.020
i
0.01
0.07 !

0.31


0.07 :


-------
                   ROTOMETER
                 SHUT-OFF
HELIUM  GAS TANlT)   VALVE
                                                                    SYRINGE WATER
                                                                        PUMP
                                             HEATED COPPER
                                               TURNINGS
                                  MANOMETER
FLASH  BOILER

                                                            C7vr>
  WET TEST METER
                 GAS TRAPS WITH CuSO,
                                            RECORDER
                                                                CONDENSATE
                                                                 RESERVOIR
                          REACTION VESSEL
                             IN FURNACE
                                                                                      WATER TRAP
                                        GAS CHROMATOGRAPH

              FIGURE  1.  Diagram of  Experimental Setup Used in Steam Oxidation Tests

-------
                  -16-
         REACTION
         PRODUCTS
THERMOCOUPLE
TO RECORDER

THERMOCOUPLE
TO CONTROLLER
                                   REACTANT
                                   QUARTZ
                                    FRIT


                                   FURNACE
         STEAM  &
      CARRIER  GAS
FIGURE 2.  Steam Oxidation Reactor and  Furnace

-------
                         -17-

reactant bed and disperse the incoming gas.  The capacity of the
reactor was 100 g of copper concentrate.  A thermocouple well  was
fitted through the top in such a way that the thermocouple junctions
were surrounded by the charge.  The gas inlet tube was 9 rnn I.D. and
the outlet tube was 20 rrni I.D.  The charge was loaded into the
reactor through the outlet tube.
     The reactor was heated in a 2-1/4 in. diameter by 12 in.  long
tube furnace.  The temperature was monitored and controlled by use
of two chromel-alumel thermocouples, a proportioning controller and
a strip chart recorder.  Temperature control was maintained within
 2C of the set point.
     The gas stream leaving the reactor contains helium, hydrogen,
water, hydrogen sulfide, sulfur dioxide and sulfur.  The sulfur and
water were removed by use of a water-cooled condenser.  The remaining
gases passed through two or three traps containing solutions of
copper sulfate-sodium chloride to remove the hydrogen sulfide and
sulfur dioxide, and finally through a wet test meter.  The sampling
valve for the gas chromatograph was between the condenser and the
first trap.
     The helium used as a carrier gas was controlled by a gas flow-
meter and a small needle valve placed at the inlet of the system.
     In the normal operating, procedure a weighed amount of the cop-
per concentrate was placed in the reactor and spread to give the
charge a level surface.  The apparatus was assembled and purged with
high-purity helium until the gas chromatograph indicated the absence
of air.  The reactor furnace was turned on, and when the desired
temperature was reached, the water flow to the steam generator was
started.  The exit gas stream was analyzed for hydrogen sulfide and
sulfur dioxide at frequent intervals throughout the run.  The flow
rate of the helium plus any hydrogen was monitored by periodically
recording the wet test meter reading.  The pressure of the system
and the temperature of the hood housing the gas chromatograph and
wet test meter were noted.

-------
                             -18-

b.   Desulfurization
         In any application of the steam oxidation  concept  it  is  likely
    that the pyritic copper concentrates would first be  neutral-roasted
    to remove primarily the labile sulfur from the  pyrite and  also  to
    remove other major volatile contaminants such as arsenic and  antimony.
    Furthermore, such separation would minimize the significant reaction
    between water vapor and sulfur which also yields hydrogen  sulfide  and
    sulfur dioxide.  In addition,  since pyrite decomposition takes  place
    at temperatures substantially below that at which the steam oxidation
    reaction is optimum, the neutral-roasting operation  should be con-
    ducted separately.  Otherwise, the hydrogen sulfide  produced  con-
    currently would be evolved with the other volatiles  such as sulfur,
    arsenic, antimony and possibly lead, plus major dilution with the
    neutral-roasting gases.  Additional clean-up of the  hydrogen  sulfide
    may then also be required in order to produce high-quality sulfur.
         For other minerals in the copper-iron-sulfur system,  such  as
    chalcopyrite, decomposition into  rather simple  components  may also
    be visualized, such as:
            2 CuFeS2 -v 2 FeS + Cu2$ + S.
    However, the complexity of this system suggests that such  a direct
    and desirable result would probably not be achieved. Instead,  some
    mineral forms of lower sulfur content could be  expected along with
    the products as indicated by the  above simplified reaction concept.
    Although the scope of the study did not provide for  a detailed  study
    of these possible reactions, it is likely that  in the subsequent
    steam oxidation step, oxidation is sufficiently extensive  to  assure
    that all iron-sulfur compounds are converted to iron oxides with the
    sulfur evolved largely as hydrogen sulfide.
         The reactor used for neutral  roasting of the copper concentrates
    consisted of a 2 in. I.D. by 30 in. long Vycor  tube  closed at one  end
    with a rubber stopper at the open end.  A 9 mm  I.D.  tube inserted
    through the rubber stopper extending nearly to  the closed  end served
    as an inlet for the helium used to maintain a non-oxidizing atmosphere.
    A second 9 mm Vycor tube closed at one end was  used  as  a thermocouple

-------
                                 -19-

        well.  The exit line was a 20 mm glass tube leading to a 500 cc
        flask to trap the bulk of the sulfur.
             The reactor was heated in a 2-1/4 in. I.D. by 16 in. long tube
        furnace.  The temperature was controlled manually with a variable
        transformer and monitored with a chromel-alumel thermocouple and a
        strip chart recorder.
             The desulfurization was carried out by placing a weighed amount
        of the vacuum dried copper concentrate in the reactor tube and the tube
        placed in the furnace so the charge was centered in the hot zone and
        the helium flow started.  When all the air was purged from the reactor
        tube, the furnace was adjusted to the desired operating temperature
        (usually 800C) and maintained at that temperature for 16 hours.  The
        system was then cooled, the charge removed, pulverized, weighed and
        stored for future use in capped jars.
3.  Analytical Methods
         The gas produced in the steam oxidation reaction will contain hydro-
    gen, hydrogen sulfide, sulfur dioxide, sulfur, water and carrier gas
    (helium in our experiments).  As the gases cool,  the sulfur and water
    condense.  In the presence of water and a catalytic surface, the follow-
    ing reaction also takes place:
                   2 H2S + S02 - 1.5 S2 + 2 H20.
    Because of this reaction the determination of the composition of the gas
    leaving the reactor is a difficult analytical problem.   From the practi-
    cal standpoint, it is essential  to know only the  net hydrogen sulfide
    and sulfur dioxide resulting from the steam oxidation reaction(s).  The
    net hydrogen sulfide and sulfur dioxide after removal of the condensed
    water and sulfur were determined by gas chromatography or absorption
    followed by a chemical method described below. The total  sulfur used
    up in the reactions was determined by measuring the sulfur content of
    the copper concentrate, both before and after the reactions.
         The chromatograph used in  all  the measurements was an Aerograph
    Autoprep Model  700 with a thermal  conductivity detector.   A 1/4 in. O.D.
    x 5 ft long stainless steel columns of Poropak Q  provided  excellent

-------
                                  -20-
        separation of  hydrogen sulfide and sulfur dioxide.  At  a  column
        operating temperature of  100C the retention time for hydrogen sul-
        fide was 2 min., 45 sec., and for sulfur dioxide 5 min.,  20 sec.  The
        chromatograph  was equipped with a gas  sampling valve which was con-
        nected  in-line with the gas  stream to  be monitored.  A  2-ml sample
        of the  gas stream could be injected  into the chromatograph without
        interrupting the process  flow.
            The chromatograph was calibrated  for hydrogen sulfide and sulfur
        dioxide by injecting samples of known  hydrogen sulfide  or sulfur di-
        oxide concentration using the gas sampling  valve to measure the
        volume.  The chromatograms were recorded on a Hewlett-Packard recorder.
        Calibration curves were obtained by  plotting gas concentration
        versus  peak height times  peak width  at half-maximum.  The minimum
        measurable concentration  was 0.1% hydrogen  sulfide and  0.25% sulfur
        dioxide.
            In some experiments, the total  hydrogen sulfide and  the total
        sulfur  dioxide evolved in the steam  oxidation run were  determined
        by a method based on the  absorption  of these gases in a solution
                                                                        (*)
        containing a known amount of copper  sulfate and sodium  chloride.
        In this method hydrogen sulfide is quantitatively precipitated as
        cupric  sulfide and sulfur dioxide is oxidized to sulfur trioxide.
        In both reactions a stoichiometric amount of hydrogen is  released.
        By measuring the amount of copper precipitated and the  amount of
        hydrogen ion formed after absorption of the gas being analyzed,  the
        hydrogen sulfide and sulfur  dioxide  can be  calculated.
            The analyses for copper, iron,  gold, silver, molybdenum, cad-
        mium, zinc, arsenic, antimony, and sulfur in the copper concentrates
        and  in  the solid reaction products from the steam oxidation process
        were made  by a commercial laboratory using  standard wet-chemistry
        methods.   However, for the analysis  for the trace constituents such
        as selenium, tellurium and mercury,  an atomic absorption  spectro-
        metric  method  was used by the same commercial laboratory.
(*)   Schwartz,  Arthur,  "Determination of Hydrogen  Sulfide,Sulfur Dioxide,
     Carbonyl  Sulfide,Carbon Desulfide and Carbon  Dioxide  in  a  Gas  Mixture,"
     Anal.  Chem.  43,  pp.  389-392  (1971).

-------
                                   -21-
4.  Experimental Results
    a.  Differential Thermal  Analysis of Concentrates
             Differential  thermal  analysis (DTA) was performed on the various
        copper concentrates.   The  analyses were made using a du Pont Model  900
        DTA unit.   Samples of the  air-dried concentrates were heated in helium
        and air at controlled rates to about 1000C.  The samples heated in
        helium each exhibited an endothermic reaction beginning between 450
        and 570C  which is due to  the liberation of labile sulfur from the
        pyrite and chalcopyrite.  Figure 3 shows typical DTA scans obtained
        with Morenci and Lavender  Pit concentrates in helium.  The tempera-
        tures at which sulfur evolution began are as follows for the various
        concentrates:
                       Concentrate          T,C
                       Morenci                458
                       Tyrone                 459
                       Lavender Pit          457
                       Battle Mountain        470
                       Pima                   457
                       Anaconda              570
        The  temperature at which the  endothermic reaction  begins  is  about
        100C  higher for the Anaconda concentrate than  for the  other concen-
        trates.   The reasons for this difference are  not known  but are
        probably  related to the complex mineralogy of the  Anaconda concentrate.
            When the  concentrate samples,  which had  been  heated  in  helium,
        were reheated  in air, each exhibited  a  very large  exothermic reaction
        which  began  between 310 and 538C.  With each concentrate the reaction
        occurred  over  a temperature range of  at least 200C.  In each case
        the  exothermic reaction was followed  by a strong endothermic re-
        action which began between 630 and  750C depending on the concen-
        trate.  The  exothermic  reaction is  due  to oxidation of  the metal
        sulfides.  The endothermic reaction is  probably due to  (1) the de-
        composition  of any copper sulfate formed during the oxidation reaction,
        and  (2) a change in crystal structure of the  metal  oxides present.
        All  of the DTA curves obtained with the various concentrates are
        similar in their general  character  to those reported in the  litera-
        ture for  the sulfides of iron and copper when heated in air.
                                           i

-------
                        -22-
o
X
<

o
                 MORENCI  CONCENTRATE
o
X
o
o
               LAVENDER  PIT CONCENTRATE
L


0
           1
          I
1
1
200     400      600     800


    REACTION TEMPERATURE,  (
                                           1000
                        1200
  FIGURE  3.  Differential Thermal Analysis of Air-Dried

            Copper Cencentrates in Helium

-------
                              -23-
         When air-dried concentrate samples were heated directly in
    air the DTA scans obtained were very similar to those obtained with
    samples which were heated in helium, and then air.  However, for the
    samples heated directly in air, the exothermic reaction  began at a
    lower temperature and was of much greater magnitude.  Typical  DTA
    traces for copper concentrates heated in air are shown in Figure 4.
b.   Labile Sulfur Removal
         When copper concentrates containing pyrite or chalcopyrite are
    heated to an elevated temperature in a  vacuum or inert atmosphere
    the pyrite decomposes with the evolution of sulfur vapor.  In  a
    static system the extent of the decomposition is determined by the
    reaction temperature  and the sulfur partial  pressure.   If the  sulfur
    vapor is removed from the reaction vessel  as it forms, the extent of
    the pyrite decomposition will  depend on  the  reaction temperature and
    time at temperature.   The products of the  pyrite decomposition reaction
    are sulfur vapor and  pyrrhotite.   The composition of the  sulfur vapor
    will  depend on  the vapor temperature and pressure, and will  contain
    a  mixture of $2, 83,  84, 85, Sg,  S7, and Sg  molecules. At extremely
    high temperatures, some dissociation of  $2 molecules to sulfur atoms
    occurs, but this occurs far above the temperature region  of interest
    in this study.
         Both pyrite and  pyrrhotite are non-stoichiometric sulfides.
    They can be approximated by the formulas Fe$2 and FeS. However,  the
    iron-to-sulfur  ratio  for each compound can vary over a substantial
    range, depending on the sulfur partial pressure.   Figure  5 shows  a
                                             (37)
    portion of the  iron-sulfur phase  diagram/   ' and the  range of homo-
    geneity that exists for each compound.   For  example,  the  diagram
    shows the wide  range  of composition which  exists  for the  so-called
    beta form of iron sulfide (FeS-^  ).   Furthermore, this form is
    stable up to very high temperatures.  Thus,  one could  not expect  to
    obtain constant iron-to-sulfur ratios when thermally decomposing
    pyrite.  The ultimate composition  would  be determined  by  the dura-
    tion of exposure and  the partial  pressure  of sulfur over  the material.
    In a flowing gas system, equilibrium conditions could  be  reached
    only after very prolonged periods.

-------
                       -24-
              MORENCI  CONCENTRATE
o
X
o
o
UJ
           LAVENDER  PIT  CONCENTRATE
                                                    J
          200
400
600
800
1000
1200
              REACTION TEMPERATURE,  UC
  FIGURE 4.   Differential Thermal Analysis of Air-Dried

            Copper Concentrates Heated in Air

-------
                              -25-
1600  -
                                 60
                        SULFUR,  ATOM PERCENT
70
         FIGURE 5.   The  Iron-Sulfur System (Rosenqvist)

-------
                           -26-
      If the steam oxidation reaction is carried out with the copper
concentrate as received from the producer, the labile sulfur is
evolved as the concentrate is heated to the reaction temperature.
This  sulfur condenses in the downstream sections of the equipment
and can interfere with the equipment operation.  It was decided,
therefore, that the labile sulfur would be removed from the various
concentrates, and the desulfurized products would be used for the
steam oxidation reactions^  The desulfurization reactions were
carried out in the equipment discussed in the previous section.  The
desulfurization temperature, in each case, was 800C and the concen-
trate was held at temperature for 16 hours in a stream of flowing,
high-purity helium.
      A large quantity of each concentrate was desulfurized.  A
1-2 kg quantity of each desulfurized concentrate was thoroughly
blended and analyzed.  In all  subsequent experiments involving a
specific concentrate, the feed material was taken from the blended
and analyzed batch of desulfurized product.
     The six concentrates showed the following weight losses due
to sulfur evolution.
                Initial Sulfur    Weight      Weight Loss as %
Concentrate          Wt%	    Loss, %     of Initial Sulfur
Morenci              36.2          16.8             46.4
Tyrone               40.0          19.0             47.5
Pima                 29.6           9.6             32.5
Lavender Pit         39.7          18.4             46.4
Battle Mountain      31.3          15.2             48.5
Anaconda             34.7          17.6             50.7
It is impossible to assign a specific formula to the iron sulfide
product from the desulfurization reactions.  However, if one assumes
that the copper is present as  CuS and any other metal  sulfides
(other than iron)  are present  as the normal sulfides, the iron-
sulfur ratios for the various  desulfurized concentrates are approxi-
mately as follows:

-------
                              -27-
              Desulfurized          Iron Sulfide
              Concentrate           Composition
              Moreno i                  FeSi.i6
              Tyrone                  FeS1-OB
              Pi ma                    FeS^Q
              Lavender Pit            FeS^ 12
              Battle Mountain         FeSi.ii
              Anaconda                FeS1>01
    Except for the desulfurized Anaconda concentrate,  the iron sulfide
    compositions are about what one would predict from thermodynamic
    considerations.
         When the concentrates were desulfurized at 800C,  sintering of
    the product did  not appear to be a problem.  Only  the Pima concen-
    trate showed a tendency to sinter.  The  Pima product formed large
    partially-sintered lumps at 800C which  were easily broken up.
    The other products formed loosely-bonded agglomerates which were
    readily broken up  by shaking the product in a glass jar.   Subsequent
    experiments showed that only the desulfurized Pima concentrate
    presented sintering problems during the  steam oxidation reaction at
    temperatures of  800C or less.   Above 800C all of the  concentrates
    showed a tendency  to sinter during steam oxidation, and at 900C
    sintering became a severe problem.
         When the various concentrates were  desulfurized, the off-gas
    always contained some h^S and S02 in addition to elemental sulfur;
    however, the amounts produced were quite small.  The reactions  which
    result in the ^S  and S02 formation are  unknown, but could involve
    bound water which  is not removed during  the drying step.

c.  Hydrogen Sulfide Yields
         In the experiments designed to measure the ^S yield, 80-100
    grams of desulfurized concentrate were placed in the vertical  reactor
    tube and the system was purged with high-purity helium.  When  all
    oxygen was removed from the system, the  concentrate was heated  to
    the desired temperature.  When the reaction temperature was reached,
    the water flow was started while maintaining the helium flow.   The

-------
                          -28-
off-gas, after cooling to room temperature to condense most of the
water, was analyzed for h^S and S02 by gas chromatography (GC).
Then the gas was scrubbed with a copper sulfate-sodium chloride
solution to remove the h^S and S02, and the  residual  gas flow
(principally helium and hydrogen) was measured using  a wet-test
meter.  From the gas flow, GC analysis and water flow, the h^S
yield as a function of water fed to the reactor could be calculated.
     In most of the experiments very little  S02 could be detected
in the off-gas by gas chromatography.  This  indicates that if any
S02 was present in the off-gas leaving the reactor, it reacted with
H2S as the gas cooled, to form elemental sulfur according to the
reaction
              2 H2S + S02 2 2 H20 + 3 S                       (1)
The h^S content of the gas stream, as determined by gas chronato-
graphy, represents, therefore, the net h^S yield after any S0
initially present has reacted.
     Figure 6 shows the results of a typical steam oxidation run
using desulfurized Morenci concentrate.  The h^S yield is essentially
constant for the first portion of the run, and then decreases with
time as the sulfur content of the concentrate is depleted.  This
indicates that the composition of the off-gas leaving the reactor
is constant during the first portion of the  run.  By  carrying out
similar runs at various water flow rates it  was established that,
during the initial phase of each run, the off-gas leaving the reactor
was in thermodynamic equilibrium with the fresh desulfurized concen-
trate.  The length of time in which the H2S  yield was constant was
a function of the water flow and volume of concentrate used.
     By carrying out a series of runs at various temperatures with
a desulfurized concentrate, it is possible to obtain  a measure of
the equilibrium H2S yield as a function of temperature.  This was
done for each of the six concentrates.  Figure 7 shows the results
obtained with three of the concentrates:  Morenci, Anaconda, and
Lavender Pit.  The results obtained with the other three concentrates

-------
                            -29-
o
 CO
    0.9
    0.8
    0.7
    0.6
    0.5
o
o
z:
ce.
O
"Is,  -4
LU
_J
O
    0.3
    0.2
    0.1
        - o
TEMPERATURE
CONCENTRATE
WATER FLOW
HELIUM FLOW
800C
DESULFURIZED MORENCI
0.585 g/MINUTE
118 cc/MINUTE
                                I
                           I
                     I
               20     40       60     80       100
                        REACTION TIME, MINUTES
     FIGURE 6.   H2S Yield as a Function of Reaction  Time
                                          120
                                    140

-------
                             -30-
o
o
   1 .0
   0.9
   0.8
   0.7
   0.6
   0.5
   0.4
   0.3
   0.2
  0.1  -
       400
O  MORENCI
O  ANACONDA
A  LAVENDER PIT
 500
600
                                     700
800
                       REACTION  TEMPERATURE, C
                                       900
     FIGURE 7.  Effect of Reaction Temperature  on  the
                Equilibrium H^S Yield  (Experimental)

-------
                         -31-

are similar to those obtained with the Morenci concentrates as can
be seen by referring to Table 4.  It must be remembered that the
equilibrium HS yields presented in Figure 7 and Table 4 are the
net yields after any hl,S and SO,, in the cooled gas have reacted to
form sulfur.
     If copper concentrate, which has not been desulfurized, is
used in the steam oxidation reaction, the initial H?S yield is in-
creased.  However, the HLS yield decreases rapidly with time and
after about 50 minutes reaction time the H2S yield approximates
that obtained with a desulfurized concentrate.  The off-gas from
the reaction contains both H^S and SO- and as the gas cools ele-
mental  sulfur is formed.   The reaction is incomplete however, and
the GC analysis shows that cooled off-gas contains both SOp and
H^S.   The HpS-SO- ratio in the gas is approximately 2:1.  It is
hypothesized that the SO- and the bulk of the H?S is formed by a
reaction between the labile sulfur vapor and water vapor:
           2 H20 + 3/2 S2 t 2 H2S + S02
During the initial stage of the run, the release of labile sulfur
is high and the SO- and H-S yields are high.  As the release of
labile sulfur decreases to zero, the H?S yield decreases until  it
equals the H~S yield from the steam oxidation reaction.
     To determine if the sulfur-water reaction could account for the
high H~S yield, experiments were performed in which sulfur vapor and
water vapor were heated together at 300-500C, cooled to room tempera-
ture, and the gas stream analyzed by GC.  The results show the gas
stream contained high concentrations of both S02 and H-S,  indicating
that the water-sulfur reaction is responsible for the high H-S yields
observed when untreated copper concentrate is steam oxidized.
     Reports in the literature    indicate that the addition of lime
or other alkaline agents to the copper concentrate can increase the
yield of H?S obtained during steam oxidation.  However, in a number
of tests in which calcium hydroxide (10 wt%) was added to  the con-
centrate, no increase in hLS yield was observed.  This was true
using both untreated and desulfurized concentrate.  A substantial

-------
                      TABLE 4,  Equilibrium H^S  Yield  As  A  Function  Of  Temperature
Temperature
               Morenci
                                        H?S Yield,  Moles  H?S  /  100 Moles  of Water  Fed
                             lyrone
Pi ma
Battle Mountain
Lavender Pit
Anaconda
450
500
550
600
650
700
750
800
850
900
-
0.07
0.14
0.23
0.35
0.49
0.62
0.82
0.89
1.05
-
0.07
-
0.22
0.35
0.51
0.60
0.77
0.81
1.09
0.02
0.07
0.12
0.20
0.33
0.51
0.65
0.82*


_
0.08
0.14
0.23
0.32
0.61
0.85
0.90
0.96

0.07
0.16
0.28
0.42
0.57
0.68
0.90
0.86
0.94

0.08
0.14
0.23
0.33
0.47
0.54
0.64
0.71
0.72
0.74



CO
ro
i





  Concentrate bed partially sintered  at  800C.

-------
                         -33-

amount of S0?, however, was detected in the off-gas when the lime
was added to the desulfurized concentrate.
     The steam oxidation of desulfurized copper concentrate is pre-
sumed to occur by the reaction
           3 FeS + 4 H90 t Fe70, + 3 H~S + H,                    (2.)
                      L.       yield, as
a function of temperature, is given in Figure 8.    The H2S yields
obtained with Morenci and Lavender Pit concentrates are also shown.

-------
                                      -34-
             TABLE 5.   Calculated Equilibrium  Gas  Compositions*
     Basis:    3FeS(s)  + 4H20(g)
                                  &F
Temp

 C
 400


 600


 800


1000


1200


1400
        AFR'
        kcal
31.46  6.11x10


35.96  1.00x10
                      -11
                      -9
                       -9
        40.53  5.578x10


        45.12  1.779xlO"8


        49.61  4.370x10-8


        54.19  8.355xlO-8
>(r\
^9 1 o
Kp
1 C0U ' ~>"o-J/n
34(s) 2(g
Gas Composition, atm
) 2(g)
H2S H2 H20
0.00366
0:00734
0.01121
0.01495
0.01856
0.02173
0.00122
0.00245
0.00374
0.00498
0.00619
0.00724
0.99512
0.98021
0.98505
0.98007
0.97575
0.97103
* Assuming ideal  gas behavior,  unit activity of solid  constituents, and

  no side reactions which affect the gas  composition.

-------
                             -35-
   1 .8
   1.6
   1 .4
   1 .2
O  CALCULATED H2$ YIELD

A  DESULFURIZED MORENCI CONCENTRATE

D  DESULFURIZED LAVENDER PIT CONCENTRATE
o
o
     400
   500
600
700
800
900
1000
1100
                        REACTION TEMPERATURE, UC
     FIGURE 8.  Comparison of Experimental  and Calculated
                (Thermodynamic) H^S Yields  at Various
                Reaction Temperatures

-------
                         -36-

     The experimentally measured hLS yields are much less than the
calculated yields.  There are several possible explanations for this:
     1.  Inaccuracies in the free energy data which affect the cal-
         culated yields.
     2.  The calculated equilibrium yield assumes ideal behavior of
         reactants and products, which may not be the case even at
         the low pressure (^ 1 atm) of the reaction system.
     3.  Thermodynamic equilibrium is not attained in the reaction
         system due to kinetic limitations.
     4.  The steam oxidation reaction does not proceed as written.
     5.  Other reactions occur which affect the H~S yield.
                                                      (38)
     The compilation of free energy data by Rosenqvist^    appears
to be the best source of data available.  Even so, the suggested
accuracy of the data is such that the AFR for Reaction (2) could
                          +
be in error by as much as - 14.5 kcal at the various temperatures.
An error of this magnitude would have a drastic effect on the cal-
culated H2S yields.  At 800C, the calculated H2$ yield could vary
from 0.207 - 5.745 moles H2$/100 moles H20 fed to the reactor,
depending on the error in the calculated free energy of reaction.
     At the low reaction pressures involved (about one atmosphere
absolute), the deviation from ideal gas behavior would be slight.
Corrections for non-ideal gas behavior would have only a slight ef-
fect on the calculated H2S yield.  Deviations from unit activity
for the solid constituents could have a substantial effect on the
calculated H2S yield, but not to the same degree that inaccuracies
in the free energy data would have.
     The experimental data indicate that the off-gas composition
does approach the  thermodynamic equilibrium during the initial stages
of each run.  Since decreasing the water flow to very low levels
does not increase  the initial H,,S yield, the off-gas from the
reactor must be at equilibrium.
     All experimental evidence  indicates  the  steam  oxidation  re-
action  proceeds  as written  as  shown  above  (2)  and  as supported  by

-------
                         -37-


the work of Thompson and Tilling.   '  However, to prove or dis-

prove that the reaction proceeds as assumed would require a study

beyond the scope of this program.

   There are a number of other reactions which could occur in

the reaction system and affect the H2S yield.  These include the

following:

                                            AFj
                                           - R       |/o
                                           (kcal)      KP


2 H2S    2 H2 + S2                      17.7   2.534 x 10"4  (3)



                 K4
2 H20 + 3/2 S2  ; - ^  2 H2$ + S02            4.34  0.1307        (4)



                 K5                                          5
2 H20 + 1/2 S2  ~^-"  S02 + 2 H2            22.0   3.31 x 10"D   (5)
3 Fe,0.. + FeS   ^  10 FeO + S0?           36.6   3.51 x 10"   (6)
    0 H        \                 ^


              K7
Fe.O. + FeS  - *  4 FeO + 1/2 S?           19.2   1.23 x 10~4   (7)
  *5 H        ~\ ~                  ^-


                K8                                           5
3 FeS + 2 S02  " - "  Fe304 + 5/2 S2         22.8   2.27 x 10"s   (8)



            K9                                               ,

FeS + H00  - "  FeO + H9S                  12.3   3.13 x 10"J   (9)
       L.   ^             t-


            K10
Fe304 + H2  " - '  3 FeO + H20               -2.3   2.94          (10)




3 FeS + 10 H20  "  ^  Fe304 + 3 S02 + 10 H2    132 1.33 x 10"27  (11)
Calculated at 800C
Similar reactions can be written involving copper.  If Si02, MgO and

CuO are present in the concentrate, the formation of complex oxides

must also be considered.

-------
                         -38-

     Because of many possible reactions which could be involved in
the steam oxidation of desulfurized copper concentrate, a rigorous
thermodynamic analysis of the reaction system is impossible.  How-
ever, it is known that under certain conditions sulfur and SOp are
reaction products in addition to magnetite, hLS and hL; also that
some HpS decomposition could occur at the reaction temperatures in-
volved.  Therefore, the equilibrium off-gas compositions were cal-
culated assuming that reactions 2, 3 and 4 occur in the reaction
system.  The equilibrium constants for the three reactions were
determined and the results are given in Table 6.  The equilibrium
gas composition was calculated as a function of temperature.  The
calculated gas compositions and H2$ yields are shown in Figure 9.
     The calculations show the hL and S02 concentrations, at equi-
librium, increase rapidly with temperature.  The sulfur concentration
is low at all temperatures, while the HpS concentration remains
relatively constant.  Assuming all of the SCL formed reacts with
KUS as the gas cools, the net H2S yield decreases with increas-
ing temperature, and would become zero at about 875C.  This is
in complete disagreement with the experimental data which shows the
net HLS yield increases with increasing reaction temperature.
     At reaction temperatures above 800C, the amount of SCL formed
increases rapidly.  At 800C all of the S02 formed reacts with HL.S
to form sulfur as the gas cools, and analysis of the cooled off-gas
shows no residual S02.  Above 800C the cooled off-gas contains
H?S plus some unreacted SOp.  If the residual SO,, was reacted with
H2S, the net H2S yield would be lower than shown in Figure 9, and
could possibly approach the predicted value.  Even so, the dis-
crepancy between the experimental and calculated gas compositions
show either the free energy data used for the calculations are in
error or the assumed reactions are incorrect.  A costly and time-
consuming experimental program would be required to determine why
the calculated and experimental gas compositions do not agree.

-------
                                   -39-
Table 6.   Equilibrium Constants* As A Function Of Reaction Temperatures
 Reactions Considered
          3FeS
 K,
 ^t
V '

 K.
    3H2S
          2H20 + 3/2S2 ---^  2H2S + SC>2
          K3 = (PH/(PY/(V2
     Temp, C

       400
       600
       800
      1000
      1200
      1400
6.11 x 10
         -11
1.003 x 10
5.578 x 10
1.179 x 10
4.370 x 10
8.355 x 10
                   -9
                   -9
                   -8
                   -8
                   -8
1.393 x 10
2.206 x 10
2.534 x 10
6.453 x 10
0.06964
0.4206
-9
-6
-4
-3
3.901 x 10
0.3585
0.1307
0.05283
0.02569
0.01520
                                                     -3
* Assuming ideal  gases and unit activity of solid  constituents.

-------
                               -40-
   0.020
   0.015
UJ
Of
Of.
a.
<
O_
o:
CD
   0.01
   0.005
            PARTIAL PRESSURES  OF  S2+H20 NOT SHOWN)

            PS  IS LOW: <0.001  ATM.  AT ALL TEMPS)

            P '   =1-P   -P   -P    -P
                                                          SO,
                                         \NET H2S


                                            \    /  NET  SO-

                                      I        \Xl	_\
            600
       FIGURE 9.
      700
800
900
1000
        REACTION  TEMPERATURE, UC

Calculated Equilibrium Gas Composition as
a Function of  Reaction Temperature

-------
                             -41-

d.  Reaction Kinetics
         The kinetics of the steam oxidation reaction were evaluated
    using the fixed bed reactor shown in 2.  The nature of the reaction
    system and the uncertainty of the reaction mechanism makes a rig-
    orous analysis of the reaction impractical.  Instead it was neces-
    sary to make empirical  estimates of the steam oxidation reaction
    kinetics based on the fixed bed data.
         If the HpS is produced by Reaction (2) and the reaction pro-
    ceeded to completion, one mole of water would yield 0.75 moles of
    H~S.  However, the thermodynamic equilibrium apparently limits the
    H2S yield that can be obtained.  At 800C, about 0.0082 moles of
    hLS per mole of water fed is the thermodynamic limit.  As the steam
    oxidation reaction proceeds in the fixed bed reactor, the pyrrhotite
    concentration in the concentrate bed decreases continuously with
    time.  The reactor system is, therefore, in an unsteady state through-
    out the course of the reaction.  In addition, at any given time the
    pyrrhotite concentration will vary from the top to the bottom of
    the concentrate bed.
         It can be visualized that for a given concentrate particle, a
    layer of magnetite builds up on the particle as the pyrrhotite reacts.
    The HpS yield from the single particle will decrease with time as
    magnetite layer builds up and the pyrrhotite concentration (or re-
    active surface area) decreases.  In a fixed bed reactor the off-gas
    leaving the reactor will approach the equilibrium composition if
    the gas-solid contact time (residence time) is sufficient.  The con-
    tact time required will, in turn, depend on the pyrrhotite concen-
    tration with the time increasing as the concentration decreases.
    The contact time is determined by the gas (water) flow rate and the
    concentrate bed depth.   For a given experimental run, in which both
    the water flow and bed volume are fixed (thus fixing the contact
    time), the H-S yield will be constant as long as the gas is in
    equilibrium with the concentrate.  However, a point is reached in
    time at which the contact time required to maintain the equilibrium
    exceeds (due to the decrease in the average pyrrhotite concentra-
    tion) the actual contact time and the H2$ yield drops below the

-------
                         -42-

equilibrium level.  The HpS yield will continue to decrease until
all of the pyrrhotite is consumed.  For a given experiment, the
water flow and concentrate volume will determine the length of
time the H^S yield is constant.  Figure 10 shows typical steam
oxidation runs at 800C using desulfurized Morenci concentrate.
With a small volume of concentrate and a high water flow, the H?S
yield is below the equilibrium level even at the start of the run.
For the three runs shown in Figure 10 the gas flows corresponded
to superficial gas velocities of 0.09-0.44 ft/sec at 800C.  With
the Morenci concentrate fluidization of the bed began at a superfi-
cial gas velocity of about 0.82 ft/sec (at 800C) which corresponded
to a water flow of 6.14 g/min.
     From the fixed bed experiments it is possible to calculate the
HpS yield, under various conditions, as a function of the pyrrhotite
concentration if one assumes an average pyrrhotite concentration
across the bed.  This assumption is not completely valid for a fixed
bed system, but the error introduced by the assumption should not be
excessive.  Figure 11 shows the results of such calculations for de-
sulfurized Morenci concentrate which was steam oxidized at 800C.
     By making a number of runs with a given concentrate at various
contact times, it is possible to estimate the contact time required
to reach the equilibrium hLS yield as a function of the pyrrhotite
concentration (process economics dictate that the ^S content of
the reactor off-gas should be a maximum, and it is necessary to know
the contact time required to reach equilibrium in order to estimate
production costs).  This was done with the desulfurized Morenci con-
centrate and the results are shown in Figure 12.  The estimate assumes
the pyrrhotite concentration is uniform throughout the bed.  This
would be approximately true for a fluidized bed reactor, but not for
a fixed or moving bed system.  Similar estimates can be made for
other reaction temperatures and concentrates, but this was not done
in the current study.
     An attempt was made to determine the effect of temperature on
the kinetics of the steam oxidation reaction.  The problem is com-
plicated by the fact that at reaction temperatures above 800C a

-------
                     -43-
                   EQUILIBRIUM H-S YIELD
                  (EXPERIMENTALLY DETERMINED)
                   20 g DESULFURIZED MORENCI
                   0.585 g/MINUTE H20
                   CONTACT TIME  -0.37
                    10 g DESULFURIZED MORENCI
                    0.825 g/MINUTE H20
                    CONTACT TIME  -0.13 SECONDS
                    10 g DESULFURIZED MORENCI
                    3.12 g/MINUTE H20
                    CONTACT TIME  -0.034
                    REACTION TEMPERATURE=800
              100             200

               REACTION TIME,  MINUTES
300
FIGURE 10.  H9S  Yield  as a Function of Reaction
            TTme for a Batch Reactor

-------
                          -44-
0.8
 THERMODYNAMIC  EQUILIBRIUM
(EXPERIMENTALLY  DETERMINED
                     CONTACT  TIME=0.33 SECONDS
       CONTACT TIME
       0.13 SECONDS
        CONTACT TIME=0.017  SECONDS
                      REACTION TEMPERATURE  -  800"C

                    I	I	I	1
                    20       30       40      50

                   PYRRHOTITE REACTED,  PERCENT
  FIGURE 11.  H?S Yield as a  Function of the Pyrrhotite
              Reacted for Various  Water Flow Rates

-------
                              -45-
          2.0
                                REACTION TEMPERATURE=800UC
                                        J_
                                                  I
                      20        40       60        80       100

                       AMOUNT OF PYRRHOTITE REACTED, PERCENT
FIGURE 12.  Contact Time Required to Reach  the Equilibrium H
            Concentration as a Function  of  Pyrrhotite Reacte
            at  800C

-------
                         -46-

substantial fraction of the pyrrhotite may be consumed in the
formation of SCL which reduces the net hLS production.  Therefore,
while the experimental data show that pyrrhotite reacts more
rapidly at 900C than at lower temperatures, the increased rate is
of no benefit because the net yield of hLS is greatly decreased at
the higher reaction temperatures.  The optimum reaction temperature
for the steam oxidation reaction appears to be in the range of 700-
800C.
Fate of Concentrate Impurities
     The various concentrates contain certain impurities whose
presence must be considered in the steam oxidation process.  The
impurities of interest are gold and silver (for their economic
value) and arsenic and antimony (for their potential health hazards
and detrimental effects on the process and resulting by-products
such as sulfur and iron oxide).
     Arsenic and antimony are present in the concentrates as sul-
fides.  These sulfides are quite volatile, however, and should be
eliminated from the concentrates during the desulfurization reaction.
Analysis of the various desulfurized concentrates showed only trace
levels of arsenic and antimony, indicating the sulfides had vola-
tilized as expected.  During the desulfurization of arsenic-antimony
containing concentrates, yellow orange solids condensed as thin films
in the cold sections of the system.  These solids were not analyzed,
but were found to be soluble in aqua regia as are the arsenic and
antimony sulfides.
     The various concentrates studied contained up to about 11 troy
ounces of silver per ton and 0.3 oz of gold/ton.  Analysis of the
desulfurized concentrates and steam oxidation residues indicated
that the gold and silver remained with the solid residue through-
out the process.  This is to be expected since the gold and silver
are not volatile at the temperatures under consideration.  While
exact material balances for the gold and silver were not obtained
due to analytical inaccuracies, there was no experimental evidence
to show that the gold and silver did not stay with the concentrate
residue.  These analytical results were as follows:

-------
                                 -47-
        Suppl^er of Concentrate
        Battle Mountain (Duval)
           In Troy Ounces - Ag
           Per Ton        - Au
        Anaconda
           Weight % - As
                      Sb
Concentrate  Desulfurized   Residue After
 Dry Basis   Concentrate   Steam Oxidation
  10.79
   0.31
   1.86
   0.037
12.3
 0.39
trace
trace
12.8
 0.56
trace
trace
5.   Interpretation of Experimental  Results
        Process Limiting Factors
             The objective of the steam oxidation concept is to produce H2S
        from iron sulfides (pyrrhotite) without need for a specific chemical
        reductant so that the H2$ can be used to react with S02 to produce
        elemental sulfur.  Any source of iron sulfide could be used, but
        copper concentrates containing pyrite or chalcopyrite were used as
        the iron source in this study.  The commercial value of the steam
        oxidation process would be determined by the cost of producing the
        H2S.
             The experimental study has shown that FLS can be produced from
        pyritic copper concentrates by the steam oxidation reaction.  Es-
        sentially all of the iron sulfide in desulfurized concentrate can
        be converted to H^S and magnetite by the reaction, while the copper
        sulfide is unchanged.  There are, however, several limitations to
        the process which reduce its potential value.
             The most serious problem concerning the steam oxidation process
        is the limit placed on the hLS concentration of the off-gas leaving
        the reactor by thermodynamic considerations.  All of the tests show
        that the H2$ concentration in the off-gas will never exceed about 1.1%.
        This corresponds to a consumption of about 1.5% of the water fed to
        the reactor.  The low equilibrium hLS concentration of the off-gas
        places severe limitations on the heat economy of the process and on
        the reactor design.  It means that at least 90 moles of water must
        be fed to the reactor to produce one mole of H2$.  Unless very high
        waste heat recovery can be accomplished, the heat load required to
        produce the H?S will be extremely large.

-------
                         -48-

     The tests show the initial hLS yields  increase with increasing
 temperature.  However, at reaction temperatures above 800C, the
 rate of SCL formation increases and much of the pyrrhotite is con-
 sumed  in S(L production.  As a result the optimum temperature for
 the steam oxidation reaction appears to be between 700 and 800C.
 At 800C the equilibrium H~S concentration of the off-gas is only
 about  0.82%, which places even more restrictions on the process.
     If the steam oxidation reaction involves only the water and
 pyrrhotite, the equilibrium hLS yields should be about the same
 for all the desulfurized concentrates.  Variations in factors such
 as surface area, crystal modifications, pyrrhotite concentrations,
 etc.,  should not affect the equilibrium hLS yield to any degree.
 However, the results show that there is a significant difference
 in the equilibrium H^S yields for the various concentrates (see
 Table  4).  This would seem to indicate that the reaction(s) which
 produce the H2S is more complex than anticipated.  This view is
 further substantiated by the fact that the equilibrium H~S yields
 are much different from those calculated on a basis of available
 thermodynamic data.  The one major conclusion that can be drawn from
 the H2S yield data is that the mechanism(s) by which H^S is pro-
 duced  in the steam oxidation process is not known.  The apparent
 complexity of the process indicates that a time-consuming experimental
 program would be required to define accurately the reaction mechanism;
 however, such work should yield results useful in understanding the
 process.
     The rate of the steam oxidation reaction(s) appears to be quite
 rapid  at temperatures of 700C and above.  However, the low equilib-
 rium H2S yield makes it quite difficult to take advantage of the
 rapid  reaction rates.  At a temperature of 800C the contact time
 required to achieve the equilibrium HpS concentration is less than
 3 seconds even when 90% of the pyrrhotite in the concentrate has been
 reacted.  As a result, any reactor design considered which is to
 produce substantial quantities of H2$ will require a very large cross-
 sectional area and only a small depth of concentrate.  The cost of
building and operating reactors of the  required  cross-sectional  area

-------
                                 -49-

        will make the cost of producing H2S by the steam oxidation reaction
        much higher than desirable.  Operation at higher temperatures does
        not improve the situation due to the formation of S0~.   At reaction
        temperatures above 800C a substantial fraction of the  pyrrhotite
        is consumed in the formation of SCL which greatly reduces the HLS
        yield obtainable.
             In summary, the major limitation of the steam oxidation process
        is the low concentration of H,,S in the reactor off-gas.   Since this
        low concentration is due to thermodynamic limitations,  it does not
        appear possible to increase the HUS concentration without changing
        the reaction(s) involved.

                              SECTION III
Economic Assessment of the Process Concept
     In view of the rather unfavorable prospects for achieving the hydrogen
sulfide yields and concentrations desired for a likely process involving the
steam oxidation concept, it was considered unnecessary to make a detailed
economic analysis.  Instead a preliminary evaluation based on a comparison
with another likely method of achieving elemental sulfur production for which
data were at hand was regarded as suitable.  The following evaluation com-
pares the steam oxidation concept with the ASARCO process for very large-
                                                             (42)
scale sulfur dioxide reduction as reviewed in the A. G. McKeex    report.
1.  Basis
        As the basis for this evaluation, an arbitrary figure of 676 tons
    of sulfur  (about 400 tons of copper per day) in a concentrate of a com-
    position representative of Morenci material was selected (Table 3) which
    on a weight basis was:
               Cu                          20.82%
               S                           36.2 %
               Fe                          27.3 %
               Other, including moisture    6.56%
               Inerts                       9.12%
2.  Material Balances
    Assumptions
    a.  All  sulfur was  assumed to exist only as  chalcocite  (CU2S) and
        pyrite (FeS2).

-------
                                   -50-
    b.  The following principal reactions were assumed for the overall
        process:
             FeS2 - FeSj^ + 0.9 S (labile sulfur evolution durinq
                                    neutral roasting and pyrrhotite
                                    formation)
             3 FeSul + 4 H20 -> Fe304 + 3.3 HZ$ + 0.7 \\2 (the steam oxi-
                                                          dation reaction)
             2 H?S + S09 + H90 + 3 S (the Glaus reaction for elemental
                d      *    *         sulfur)
    c.  The equilibrium composition of gas leaving the steam oxidation step
        of the process was assumed to be one mole of hUS per 99 moles of H?0.
         On the basis of these assumptions, the distribution of the sulfur
    from 100 tons of concentrate was calculated to be as follows:
                    Labile sulfur      13.95 tons
                    Chalcocite sulfur   5.20 -tons
                    Pyrrhotite sulfur  17.05 tons
         In order to balance the H2S and S02 input to the Claus reaction
    for elemental sulfur production, two-thirds of the sulfur remaining in
    the residue after neutral roasting of the concentrate must be  evolved
    as H2S in the steam oxidation step.   It is assumed that the sulfur re-
    maining in the residue after steam oxidation would be evolved  in the
    converting step as S02.   Therefore,  the residue from the steam oxi-
    dizing step would contain about 7.42% sulfur.
         The results of the  material  balance calculations are shown in
    Figure 13.

3.   Major Equipment Sizing
    Fluidized Beds
         The total cross-sectional  area  of the fluidized beds needed for the
    steam oxidation process  was estimated to be 28,246 square feet, as
    follows:
         The volumetric flow rate of steam at 800C and atmospheric pres-
    sure would be 28,246 cubic feet per  second.
         The gas velocity needed for fluidization was taken as 1 foot per
    second based on laboratory experience.   Therefore, the  total

-------
                                   BASIS: PROCESSING OF 1867 SHORT TONS PER DAY OF PYRITIC COPPER CONCENTRATE
                                          (QUANTITIES GIVEN ARE TONS/DAY EXCEPT AS NOTED)
   CD

   73
O> rt-
r+ O>
CD cu
-S 3
CD CL
dl QJ
 ' rt-
o> -i.
3 O
O 3
-n -s
 ' O
o o
S. fD
 V)
3" in
ft)
n>
CONCENTRATE  (1867)
      S  (676)

w/

1
1
1
S (260.5)
1
1
1
k NEUTRAL
~* ROASTING



H2S (2
, 	 H20 (1
1 H2 <
1
STEAM
^ OXIDI Z ING
806C
1
^ STEAM
~~w GENERATOR

1
1
D4'4) CLAUS
5430) 	 -^CONVERTER
i T\ 270C
?> *
S(676)
H20 (15586)
INERT GAS
(3.4xl05SCFM)
S02 (277.2)
INERT GAS (2. lx!05SCFM)
1
SMELTING
PROCESS



-------
                                  -52-

                                           po 246
    cross-sectional area required would be -^  - = 28,246 square feet.

         The bed depth is estimated to be 2.75 feet based on the follow-
    ing calculations:
         The estimated time for conversion of pyrrhotite to magnetite is
    900 minutes.
         The amount of sulfur needed to be retained as unconverted pyrrho
    tite per hundred parts of steam oxidation residue would be:
              7.42-5.20 = 2.22 parts
         The degree of pyrrhotite conversion in the steam oxidation step
    would then be:
              17.05-2.22 x 10Q = BJ%
                 17.05
         Assuming ideal or perfect backmixing in the fluid bed, the re-
    quired holdup time for conversion of the pyrrhotite in the neutral -
    roasted residue to 87% is about twice the conversion time or about
    1.25 days.  This holdup time corresponds to about 2334 tons in the
    total fluidized beds.  Assuming a bulk density of 60 pounds per cubic
    foot, the total bed volume would be:

              2334 x 200 = 77,700 cubic feet
                   60
    thus the average bed depth would be:
    (The reference for these calculations is the book by Levenspiel (43).)

4.  Capital Cost Comparison
         The capital cost of the ASARCO process for 676 tons of sulfur per
    day is estimated to be $49.0 million based on the following calculations:
         The capital cost of ASARCO process processing feed gas containing
    4 vol% S02 at a rate of 116.6 tons of sulfur/day is $12 million
    (1969).  Costs increase on the basis of 0.8 power of capacity.

-------
                                  -53-
    Therefore, the cost of a plant processing 676 tons of sulfur per day
    would be $49.0 million (1969).
         The minimum capital  cost of a steam oxidation plant is $60.3
    million based on the following calculations:
         Forty 30-ft-diameter fluid-bed units are needed.  The 1969 cost
    of one unit = |Z2. (230,000) = $285,000.  The installed cost of forty
                  t.ob
    units would then be $11.4 million (1969).  The heat requirement for
    the 28,246 cubic feet per second of 800C steam is 2.27 x 109 Btu/hr.
         Steam-generating furnaces are assumed to produce 10^ Btu/hr each.
                                          9OC
    The estimated cost of each furnace is     (360,000) or 380,000.  Twenty-
    three furnaces are required for steam production.  The cost of twenty-
    three furnaces would then be 23(380,000) or $8.7 million (1969).  These
    cost data are adjusted to 1969 using Marshall  and Stevens cost indices.
    The cost data source was a book by Dryden and Furlow.   '
         Thus, the minimum installed equipment cost is $20.1 million and
    the total minimum plant cost is  (3.0) (20.1)  or $60.3 million (1969)
    for the steam oxidation process.  The factor (3.0) in the above calcu-
    lation represents the ratio of total plant cost to equipment cost and
    is used to arrive at total minimum plant cost.

5.  Comparison of Fuel Plus Depreciation Cost of Process
         On the basis of plant costs as estimated above and on an estimate
    of fuel costs, the annual fuel and depreciation costs of the two pro-
    cesses are as follows:

                                          Millions  of Dollars Per Year
                                           ASARCO     Steam Oxidation
         Depreciation (Linear 15 yr)        3.27          4.02
         Fuel Cost/Yr ($0.40/mil lion        4.12          7.91
                             Btu)
             Fuel + Depreciation            7.39         11.93

6.  Estimate of Lower Limit of H2S Concentration in Steam Leaving the
    Oxidizer
         The following calculation was made to show the minimum h^S con-
    centration which would have to be attained to make the steam oxidation

-------
                                 -54-

    process competitive (or equivalent) in cost with the ASARCO process as
    estimated above.  Since many equipment units in parallel are assumed to
    be required, the process capital and fuel costs should be proportional
    to the steam throughput.  To reduce the cost of the steam oxidation pro-
    cess to that of the ASARCO process, the steam throughput would have to
    be reduced by at least a factor of  7.39 or 0.62.   To maintain the pro-
                                       TT753"
    duction rate of H?S, the mole % H?S in the steam leaving the steam
    oxidation units would then have to be raised to  1   or 1.61 mole %.
    In other words, if the oxidizer can be operated to produce more than
    1.61% H?S in steam, the process may be as economical as the ASARCO process.
    However, the ASARCO process has not been operated beyond the large
    pilot-plant scale and is expected to be superceded by improved processes
   'in the future.  Thus the comparison is realistically even more unfavorable
    than indicated here.

                              SECTION IV
Recommendations
     In view of (1) the lack of expectations for the steam oxidation process
to provide a competitive means for sulfur dioxide air pollution abatement,
(2) the desirability of a means of producing hydrogen sulfide on a large
scale and at characteristically low cost for use in sulfur dioxide air pol-
lution abatement, (3) the encouraging possibilities of an alternative process
for such production and (4) the full utilization of existing background and
experience developed to date in this important facet of air pollution control,
the following recommendations are made:
     1.  Further consideration of the steam oxidation process as a means of
         air pollution abatement for copper smelters under the present pro-
         ject should be abandoned.
     2.  The scope and objective of the present study should be broadened
         to cover the production of hydrogen sulfide from pyritic copper
         concentrates without the use of a reductant.
     3.  The area of study recommended for investigation would be on a concept
         involving the following combination of processes:
         (a)  The neutral roasting of pyritic copper concentrates.

-------
                                 -55-
         (b)  The reaction of hydrochloric acid with the neutral roasted
              concentrates to produce hydrogen sulfide and an enriched con-
              centrate via iron dissolution and removal.
         (c)  The conversion of the enriched concentrates to blister copper
              by conventional means thus recovering the precious metals also.
         (d)  The production of pure elemental sulfur by the reaction of the
              hydrogen sulfide from step (b) with the sulfur dioxide from
              step (c).
         (e)  The regeneration of the hydrochloric acid from the ferrous
              chloride solution by step (b) and the concurrent recovery of
              pure iron oxide for sale.

                              SECTION V
Identification of Additional Required Work
     It is reasonably clear from the results of this present study that the
steam oxidation process concept does not provide an economical means for
sulfur dioxide air pollution abatement at copper smelters and furthermore
holds little promise for improvements in the process chemistry to assure a
more favorable view.  However, this conclusion does not in any way detract
from the overall desirable objective of producing hydrogen sulfide from
pyritic copper concentrates as an attractive possibility for sulfur dioxide
air pollution abatement at copper smelters.  This observation gains consid-
erable support when the rather major work being done by others to achieve
a similar result is noted.  For example, the U.S. Bureau of Mines Citrate
process^  ' for elemental sulfur production from smelter flue gas involves
the reduction of a major fraction of the recovered sulfur by use of natural
gas to produce hydrogen sulfide which is essential to operation of the pro-
cess.  This process is scheduled to be demonstrated on a large scale.
Similarly, the co-sponsor of this present study, The Phelps Dodge Corporation,
is also concurrently engaged in a substantial cooperative effort with others
to demonstrate an improved process for sulfur dioxide reduction with natural
gas also to yield elemental sulfur.   '  Thus these two efforts which are
of considerable magnitude have two common, general featureseach is directed
at the conversion of smelter waste sulfur dioxide to elemental sulfur.  Al-
though each accomplishes this in different ways, the end result is achieved
by use of natural gas as the reductant.  With increasing concerns about the

-------
                                 -56-


efficient use of fuels in general and natural gas in particular, availability
and economics may dictate the use of reductants other than natural gas with

accompanying complications in the overall reduction process and processing
economics.

1.  Suggested Additional  Work

         Continuation of the present study along the lines of an alternative
    approach which will  produce the required amount of hydrogen sulfide but

    still without need for a chemical reductant is suggested.   The work would

    involve the laboratory study of the chemistry of the less  familiar of

    the following steps  which would comprise the overall  process:

         (a)  Neutral roasting of pyritic copper concentrates  to remove
              volatiles  including a large portion of the sulfur as ele-
              mental  sulfur and to convert the  iron sulfides  to acid-solu-
              ble forms.

         (b)  Acid leaching of the residue from step (1)  to yield  concen-
              trated  hydrogen sulfide gas and to separate the  iron in  a
              soluble form.

         (c)  Regeneration and recovery of the  acid for  recycle and the con-
              current production of a fairly pure iron oxide  residue.

         (d)  Converting  of the enriched insoluble copper sulfide  residue
              to blister  copper containing the  usual  level  of  precious
              metals.

         (e)  Reacting of the sulfur dioxide from the converting step with
              the hydrogen sulfide from the leaching step to yield elemental
              sulfur.

         Only step (b) among the above is regarded as an  area  which requires

    substantial  laboratory study and development.   All of the  other steps
    are regarded as near  present state-of-the-art processing requiring  only

    a  small  amount of laboratory development.

         The studies  of the neutral  roasting step as  applied to a  variety of

    representative pyritic copper concentrates  as detailed  in  this present

    report are directly applicable to the follow-on  work  proposed  here.

         Means of copper  recovery involving hydrogen  sulfide production  to-

    gether with  nearly conventional  means for blister copper production  with

    or without subsequent electrorefining appears to  have been suggested in

-------
                             -57-
                (41)
the literature,    ' but a detailed presentation of such a process has
not been noted.  It is this area in which further efforts should be
directed.  Specifically, the work should comprise a laboratory and pre-
liminary economic  study of a process involving the steps as listed above.
At this time the preferred acid for hydrogen sulfide production by pref-
erential dissolution of the iron appears to be hydrochloric.  Regeneration
of this acid from  the corresponding ferrous chloride is already practiced
on a substantial scale in the steel pickling industry.  The principal
problems to be  studied would be concerned with (a) assuring the effective-
ness of both hydrogen sulfide production and preferential dissolution of
the iron, (b) development of suitable means of satisfactorily removing
or assuring the minimum carryover of copper into the iron chloride solu-
tion, (c) establish the fate of the precious metals, (d) investigation
of the response to the process of representative concentrates on the
basis of their  copper mineralization:  chalcocite, covellite, chalcopy-
rite, enargite, tetrahedrite,  bornite, etc., (e) determine the
preliminary economics of the preferred  process.
     A few scouting tests have been conducted in the laboratory on the
hydrochloric acid dissolution process on desulfurized (neutral roasted)
Morenci concentrate.  These tests showed that the evolution of hydrogen
sulfide is rapid and apparently fairly complete.   These tests, in which
no effort was made to assure the exclusion of air, also showed that less
than 10% of the copper was dissolved with over 90% of the iron being
solubilized.  A copper sulfide concentrate containing over 50% by weight
copper was obtained after removal of this iron by this simple treatment
with the hydrochloric acid.  The dissolution of some of the copper is
attributed to the presence of ferric iron presumably formed by air
oxidation.  It  is believed that essentially all  of this dissolved copper
can be reprecipitated by contacting the solution with hydrogen sulfide or
by cementation  with scrap iron.
Impact on Conventional Smelting Practices
     Assuming a successful outcome from the proposed research, the follow-
ing impact or significant changes would be required for adoption of the
process in place of conventional smelting processes:

-------
                                 -58-


         (a)  Neutral roasting with sulfur recovery instead of oxidative
              roasting would be required.

         (b)  Hydrogen sulfide production via the aqueous acid dissolution
              of the soluble iron sulfides as produced in (a) above,
              together with

                   (1)  conversion of hLS to elemental sulfur,

                   (2)  regeneration of the acid and recovery of fairly
                        pure iron oxide which should be marketable and

                   (3)  recovery of potential dissolved copper.

         (c)  Reverberatory furnaces would be eliminated.

         (d)  Converters and the converting process would be changed to ac-
              commodate the very rich and very low iron content  concentrates.

         (e)  S0?-containing gases from the converters would be  combined
              with excess hLS in step (b-1) above to produce marketable
              elemental sulfur.

3.  Advantages

         Although firm advantages cannot be claimed without confirmation

    resulting from further laboratory work, some apparent significant ad-
    vantages can be visualized on the assumption that hoped-for  results

    of the laboratory work will be realized.

         (a)  For those concentrates characterized by high pyrite content,
              a substantial fraction of the total sulfur will be obtained
              as elemental sulfur from the neutral roasting step.

         (b)  The hydrogen sulfide obtained in the acid dissolution of the
              pyrrhotite produced in (a) above would be more than enough to
              react with the sulfur dioxide obtained from the converting
              step.  If the quantity is marginal, it is probable that more
              pyrite could be retained with the copper at the concentrator.
              It is understood that most sulfide concentrators are operated
              to reject pyrite which exists in the ore.

         (c)  The essentially iron-free copper concentrate obtained after
              acid treatment and anticipated to contain over 50% copper,
              would be unique as a feed to most conventional converters.
              Although the converting procedure would probably be changed,
              the changes could result in very high capacity from a converter,
              with a greatly reduced requirement for slag-forming agents
              such as silica and limestone.  Electro-converting  may also be
              attractive with concurrent production of a rich and low-volume
              sulfur dioxide off-gas.

-------
                        -59-


(d)  No reverberatory furnaces with their high fuel  and slag-forming
     material requirements would be needed.   At present the rever-
     beratories comprise the principal  source of low-strength, large-
     volume sulfur dioxide-containing waste  gases.

 (e)  Waste heat from neutral  roasters  and converters may contribute
     substantially to the evaporation economy of the acid regenera-
     tion step.  The net heat  (fuel) requirements for this alterna-
     tive smelting process may even be  less  than that for a conven-
     tional smelter.

(f)  The lower total off-gas load and minimum concentration of waste
     sulfur dioxide in these off-gases  may make unnecessary the
     requirement for the costly very tall  stacks at  smelters,  with
     their attendant and also  costly electrostatic  precipitators.

(g)  The process should yield  high-quality sulfur and iron oxide
     by-products from which significant sales value  should be  realized.

-------
                                    -60-


                                 REFERENCES


  1.   Regnault,  V.,  "Research  Relative  to  the  Action  of  Water  Vapor  on Metals
      and Metal  Sulfides,"  Annales  de Chim.  et de  Ph.ys..  378-88  (.1836).

  2.   Gautier, M.  A.,  "Effect  of  Water  Vapor at Red Heat  on  Sulfides, Formation
      of Native  Metal  -  Application to  Volcanic Phenomenon," Comptes rendu,
      94, 1465-70  (1906).	

  3.   McLarty, J.  A.,  U.S.  Patent 987,156,  "Treating  Nickel  Sulfide  Ores With
      Steam  Under  Pressure  Without  Roasting."

  4.   Hall,  W. A., "A  New Desulfurization Method,  Eng. Mag., 4_5 876-8 (1913).

  5.   Wilson, A. W.  G.,  "Hall  Process for Desulfurizing Ores," Summary Report.
      Mines  Branch,  Canada  Dept. Mines,  27-30  (1913).

  6.   Hall,  W. A., U.S.  1,083,247,  "Process  for the Production of Sulfureted
      Hydrogen"  (1913).

                  U.S.  1,083,248,  "Process  for the Extraction of Sulfur from
      Metallic Sulfides" (1913).

                  U.S.  1,083,249,  "Process  for the Production of Sulfur"  (1913).

                  U.S.  1,083,250,  "Process  for Recovering Sulfur" (1913).

                  U.S.  1,083,251,  "Process  of  Obtaining Sulfur from Sulfides" (1913).

                  U.S.  1,083,252,  "Process  of  Desulfurizing and Briquetting Ores" (1913)

                  U.S.  1,083,253,  "Process  for Extracting Sulfur" (1913).

  7.   Bacon, R. F., U.S. 1,235,953,  "Production of Hydrogen Sulfide"  (1917).

                   U.S. 1,731,516,  "Recovery of Sulfur from Iron Pyrites"  (1929).

                   U.S. 1,782,226,  "Recovery of Sulfur from Iron Pyrites"  (1930).

                   U.S. 1,939,033,  "Recovery of Sulfur"  (1933).

 8.  Leese, L. F.  W., U.S.  1,947,529,  "Sulfur Recovery from Sulfides"  (1932).

 9.  deJahn, F.  W.,  U.S. 2,851,349, "Obtaining Iron and  Sulfur from  Iron Sulfide
     Ore Without Fusion" (1958).

10.  Woltereck,  H. C., Ger. 275,503, "A Process of Separating  Zinc and Other Metals
     Oxidizable  by Steam."

11.  Arnold, F.  P. and Wideman, G.  F.,  U.S.  1,104,239 and 1,104,287, "Removing
     Sulfur and  Arsenic with Alkalies  and  Superheated Steam."

12.  Hooten, W.  M.,  Brit.  18,007,  "Treatment of Metallic Sulfides with Steam
     Mixed with  Other Gases."

-------
                                    -61-


 13.   Perkins, W. G. , U.S.  1,478,295,  "Selective Treatment of Ores Containing
      Fe  Pyrite with Steam."

 14.   Petersen, E.  F. and Field, S., Brit. 206,207, "Sulfide Ores are Treated
      with  Steam to Recover h^S with a Limited Quantity of Air."

 15.   Lindemann, L., Norw. 53,577, "Freeing Ores and Oxides from Sulfur with
      Hydrogen and  Superheated Steam."

 16.   Goetz; C., Ger. 636,773 (Also Ger. 589,448) "Treatment of Sulfidic or
      Arsenical Ores with Steam and a Reducing Gas."

 17.   Horakova, J. and Rajecky, F., Czech., 90,809, "Separating Elements from
      Ores  by Roasting in Steam and Inert Gases."

 18.   Birkeland, K. , U.S. 1,121,606, "Process for Treating Sulfid Ores and
      Other Metallic Sulfids" (1914).

 19.   Millar, W. S., U.S. 1,567,378, "Desulfurizing Sulfide Iron Ores with
      Sul fur Dioxide. "

 20.   Thompson, F. C. and Tilling, N., "The Desulfurization of Ores," J. Soc.
      Chem. Ind., , 27-46T  (1924).

 21.   Ruhrmann, J., "Elimination of Sulfur from Spathic Iron Ore by Roasting,"
      Stahl u. Eisen, 4, 1118-9 (1926).

 22.   Grunnert, E. , "Desulfurization of Coal," J. Pract.  Chem., 122,  1-120 (1929).
23.  Kirillov, I. P. , and Makarov, M. M., "Preparation of HS from Pyrite by
     the Action of Water Vapor," J. Applied Chem. (USSR) ^9, 71-8 (1946).

24.  Rama Murthi , R. K. , "Decomposition Studies of the Iron Pyrites of Bihar
     with a View to Their Utilization," J. Indian Inst.  Set., 3, 32-5 (1954).

25.  Inaba, T. and Kono, K., "Research on the Reaction of Iron Sulfides,"
     J. Chem. Soc. Japan, 57_, 1-21 (1954).

26.  Inaba, T., and Kono, K. , "Mechanism of Wet Roasting of Pyrrhotite,
     J. Chem. Soc. Japan, 58^, 952-9 (1955).

27.  Budon, V. D. , "Interaction Between Water Vapor and  Metal Sulfides,"  Bulletin
     from the Institute of Metallurgy and Enrichment,  Academy of Sciences of the
     Kazhak Socialist Republic, 2_, 36-46 (1960).

28.  Uno, T., "Equilibrium Between FeS and Mixed  Gas of  H2  and I-^O,"  Mem.  Fac.
     Eng. Hokaido Univ.,  9_, 84-90 (1952).

29.  Diev, N. P.,  Paduchev, V.  V., and Toporova,  V.  V.,  "Use of Steam in  the
     Bessemerization of Copper  Mattes in Oxygen,  Acad  of Sci . of the  Soviet
     Union, Reports from Eastern Branch, 6,  79-84 (1957).

-------
                                    -62-

30.   Frey, W. H., "Production of Hydrogen Sulfide (Thermodynamics)," Chimia,
      2., 265-88  (1948).

31.   Park, C. F. Jr., "Hydrothermal Experiments with Copper Compounds, Econ.
      Geol., 26, 857-83  (1931).

32.   Kunda, W., Mackiw, V. N. and Rudyk, B., "Iron and Sulfur from Sulfidic
      Iron Ores," The Canadian Mining and Metallurgical (CIM) Bulletin for
      July, 1968, pp. 819-835.

33.   I'iehta, B.  l\. , and  O'Kane, P. T./'Economics of Iron and Sulfur Recovery
      from Pyrites," ibid pp. 836-845.

34.   Graves, J. T., and Heath, T. D.,  "Sulfur Dioxide and Sulfur from Fluosolids
      Systems,"  presented at a meeting of the American Institute of Mining,
      Hetal1urgical  and  Petroleum Engineers in Los Angeles, Calif., Feb.  2, 1967.
      Technical  Preprint No. 7008-P by Dorr-Oliver Incorporated.

35.   Kassila, K., "Flash Smelting Process," a paper presented at the "Seminar
      on Copper  Production and Group Study Tour of Copper Plants in the USSR,"
      Tashkent,  USSR 1-15 October, 1970, United Nations Industrial  Development
      Organization.

36.   Guccione,  E., "From Pyrite:  Iron Ore and Sulfur Via Flash Smelting,"
      Chemical Engineering, Feb. 14, (1966),  pp.  122-24.

37.   Rosenqvist, T., J. Iron Steel Inst.. 175, pp. 37-56 (1954).

38.   Rosenqvist, T., "Thermochemical Data for Metallurgists," Tapir Forlag (1970).

39.   Rosenbaum, J.  B. et.  al., U.S. Bureau of Mines Report of Investigations
      (RI-7774), "Sulfur Dioxide Emission Control  by Hydrogen Sulfide Reaction
      in Aqueous Solutionthe Citrate  System," 1973.

40.  Henderson,  J.  M.,  "Reduction of SO, to Sulfur,"  Mining Congress Journal,
     March 1973, pages 59-62.          L

41.  Levy, S.  I.,  U.S. Patents 1746313 (1930), "Treatment of Copper-Rich
     Material,"  and 1980809 (1934), "Production  of Ferric Oxide and Other
     Metal Values  from Pyrites."

42.  PB-184885,  "Systems Study for Control  of EmissionsPrimary Non-Ferrous
     Smelting IndustryVolume II," Arther 6. McKee  & Co.,  June 1969.

43.  Levenspiel, 0., "Chemical Reaction Engineering," John  Wiley and Sons,
      Inc., (1962),  page 367.

44.  Dryden & Furlow, "Chemical Engineering Costs,"  (1966).

-------
                      -63-
                    APPENDIX
WORLD PRIMARY COPPER SMELTERS AND SMELTING CAPACITY

-------
                                               WORLD PRIMARY COPPER SMELTING CAPACITY
COUNTRY

EUROPE

Austria

Belgium


Bulgaria


Finland

France

Germany F.R.
Italy

Netherlands
Norway  .


Rumania

Spain
Sweden

Yugoslavia
COMPANY
Montanwerke BMxlegg G.m.b.h.
Metal lurgle Hoboken
Metallo-Chimique S.A.
George Damijanov Copper
Medet Copper Works
Outokumpu Oy

Societe Francais d/Affinage du Culvre
Norddeutsche Affinerie
Berliner Kupfer-Raffin G.m.b.H.
Duisburger Kupferhutte
Metallhutte Kail G.m.b.H.
Huttenwerke Kayser A.G.
AMMI

Alcu Metaal N.V.
Orkla Metal Aktieselskap
A/S Sulitjelma Gruber
Combinatul Chimico
Metallurgic "Georghin Dej"
Rio Tinto Patino S.A.
Rio Tinto Patino S.A.
Electrolysis del Cobre S.A.
Industrie Reunidas Minera-
 Metalurgicas S.A.
Boliden Aktiebolag

Rudarsko Tapionicarsk Bazen Bor
                                      SMELTER
Brixlegg
Hoboken
Beerse
Harjavalta
Polssy
Hamburg
Berlin-WUlmersdorf
Duisburg
Kail (Eifel)
Lunen

Aussa-Corno
Thamshamm
Sulitjelma
Bor
Bara Mare
Huelva
Rio Tinto
Barcelona
Asua

Ronnskar
Bor
                       1971 CAPACITY
                       (METRIC TONS)
12,000

45,000


25,000
20,000

50,000

11,000

75,000
15,000
30,000
20,000
42,000


10,000

 4,000
 6,000
 40,000
 18,000
 15,000
 18,000


 65,000

105,000
                    POTENTIAL
                    CAPACITY
                 (METRIC TONS)*      DATE
                      18,000
100,000





 60,000





200,000


 75,000
                1975
1972





1975





n.a.


1972-73
                                                                                                                                                        i
                                                                                                                                                       CT>
                                                                                                                                                       -P.
                                                                                                               140,000
                                                                                                     1974
                                                                TOTAL
                                                                643,000

-------
USSR
                           Government
                        1965 CAPACITY
                        (METRIC TONS)
 Pechenga                   12,000
 (Kola Peninsula)
 Monchegorsk                11,000
 (Kola Peninsula)
 Munusinsk                   50,000
 (Central Siberia)
 Pitkaryanta                10,000
 (Karelo-Finnish Republic)
 Mednogorsk  (Urals)          55,000
 Balkhash (Kajakhatan)      125,000
 Dzezhkazgan-Karsakpai      200,000
 Combine  (Kazakhsdan)
 Almalyk  (Uzbekistan)       100,000
 Norilsk                     5,000
 (North Central Siberia)
 Srednyevralsk  (Urals)      50,000
 Krasnovralsk (Urals)       45,000
 Kirovgrad (Urals)          25,000
 Yuspenskiy  (Kazakhstan)     5,000
 Karabash (Urals)           30,000
 Blyava (Urals)             50,000
 Sibay (Urals)              25,000
 Kadzharan (Azerbaldzhan)   25,000
 Zangezur (Azerbaidzhan)    10,000
Allaverdi  (Armenia)         15,000
                                                                                                               100,000
                                                                                                               300,000
                                                                                                                                                i
                                                                                                                                               en
                                                                                                                                               en
                                                                                                                                                i
                                                               TOTAL
                                                                                         848,000
                                                                                       1,085,000** (1971)
                                              1,389,000**
1977

-------
AFRICA

Rhodesia

S. African Rep.



S. W. Africa

Uganda

Zaire


Zambia
Messina-Rhodesia Smelting &
 Refining Co., Ltd.
Messina Transvaal Development Co.
Palabora Mining Co., Ltd.
O'OKiep Copper Co., Ltd.

Tsumeb Corporation, Ltd.
Kilembe Mines Ltd.

Gecamines
Gecamines
Gecamines
Roan Consolidated Mines
Roan Consolidated Mines
NCCM (Nchanga Consolidated Copper
NCCM                       Mines)
Alaska
23,000
Messina
Palabora
Nababeep
Tsumeb
Jinja
Likasi-Shituru
Lubumbashi
Liulu
Mufulira
Luanshya
Chingola\
N'Kana [
20,000
89 ,500
40,000
35,000
18,000
185,000
125,000
125,000
230,000
122,000
420,000



170,000
155,000
247,001
500,000



1972
1973
1973
1974-75
                                                                 TOTAL
                                                              1,432,500
                                                                                   01
                                                                                   i
ASIA
India



Iran
Japan
Indian Copper Corp.
Hindustan Copper Corp.
Hindustan Copper Corp.

Government
Mitsubishi Metal Mining  (2 p
Onahama Smelting &  Refining
Furukawa Mining Co.
Dowa Mining Co.
Dowa Mining Co.
Hibikyodp Smelting  Co.
Mitsui Mining Co.
Nippon Mining Co.
Nippon Mining Co.
Sumitomo Metal Mining  Co.
Sumitomo Metal Mining  Co.
Sumitomo Metal Mining  Co.
Rasa  Industry.
Toho  Zinc
Moubhander
Khetri
Ingladhal
Sar Chesmeh
Naoshima
Onahama
Ashio
Kosaka 1
OkayamaJ
Tawanao
Hibi
Hitachi
Saganosekl
Kuni tomi
Besshi
Toyo
Miyaki
Onahama
10,000
	
	
	
146,000
90,000
38,000
54,000
47,000
60,000
120,000
16,000
108,000
96,000
24,000
18,000
                                                 31,000
                                                  n.a.
                                                  n.a.
                                                                                                                288,000
                                                                                                                 84,000

                                                                                                                 120,000
                                                                                                                 240,000
                                     1974
                                     n.a.
                                     n.a.
                                                                n.a.
                                                                1972

                                                                1972
                                                                n.a.

-------
Korea S.
Philippine Republic
Turkey
Government
Government
Etibank
Etibank
Black Sea Copper Mines
Musan


Murgul
Erganl
Samsun


TOTAL
   11,000
   19,000
                                                                                           857,000
    15,000
60-100,000



     n.a.
                                       1974
                                                                                                                                n.a.
NORTH/CENTRAL AMERICA
Canada
Mexico
Puerto R1co
U.S.A.
Inco
Inco
Falconbridge Nickel Mines, Ltd.
Gaspe Copper Mines, Ltd.
Hudson Bay Mining & Smelting Co.,Ltd.
Noranda Mines, Ltd.

Asarco Mexicana S.A.
Asarco Mexicana S.A.
Campania Minera de Cananea S.A.
Cia Minera Macocozac S.A.
Cia Minera de Santa Rosalia

Ponce Mining (Kennecott-Arnax)
American Metal Climax, Inc.
Asarco
Asarco
Asarco
Anaconda
Inspiration Consolidated
Kennecott Copper Corp.
Kennecott Copper Corp.
Kennecott Copper Corp.
Kennecott Copper Corp.
Magma
Phelps Dodge
Phelps Dodge
Phelps Dodge
Phelps Dodge
Copper Range
Bagdad (electrowinning)
Copper CUff j
Coniston    J
Falconbridge (Ont.)
Murdochville (Que.)
Flin Flon
Noranda (Que.)

San Luis Potosi
La Cavidad
Cananea
Conception del Oro
Santa Rosalia

Ponce
Cartaret
Tacoma
El Paso
Hayden
Anaconda
Miami
Garfield (Utah)
Hurley (Chino)
McGill (Nev.)
Hayden (Ray)
San Manuel
Douglas
Morenci
Ajo
Southern Hidalgo
White Pine
Arizona

TOTAL
  170,000

   30,000
   63,000
   45,000
  204,000

   30,000

   37,000
   25,000
   10,000
   40,000
  105,000
   77,000
   80 ,000
  170,000
   70,000
  300,000
  100,000
  100,000
  100,000
   85,000
  140,000
   80,000
   70,000

   81,000
    6,500

2,218,500
                                                                                                                200,000
   254,000

    42,000
     n'.a.
                                                                                                                  n.a.
                                                                                                                135,000
                                                                                                                 91,000?
                   1974-75
1973-74

n.a.
n.a.
                                                                                                                                n.a.
                    1974
                    1974

-------
SOUTH AMERICA
Brazil

Chile
Peru
AUSTRALIA
Plgnatari Group
Pignatari Group
Andes Copper Mining Co.
Chile Exploration
Sociedad Minera El  Teniente S.A.
Cia Minera Disputada de las Condes
 S.A.
ENAMI
ENAMI
Mantos Blancos
Southern Peru Copper Corp.
Cerro de Pasco Corp.
                           Mt.  Isa Mines,  Ltd.
                           Electrolytic Refining  & Smelting  Co.
                           Peko
Caraiba
Caraiba
Potrerillos
Chuquicamata
Caletones
Chagres
Las Ventanas
Paipote
Mantos Blancos
Ilo
La Oroya
---
100,000
319,000
Z55.000
30 ,000
43,000
33,000
30,000
160,000
50,000
                                                                 TOTAL
                                                                 TOTAL

                                                                 GRAND  TOTAL  1971

                                                                 GRAND  TOTAL  1977
                                                              1,020,000
35,000
70,000
                                                                                                               100,000
                                                                                                                 n.a.
1973?
1976?
               n.a.
               n.a.
Mt. Isa
Port Kembb
Tennant Creek
100,000
40,000
 v_ 
150,000

25,400
1974

1972
                                                                                                                                                   OO
                                                                                                                                                   I
                                                                140,000

                                                              7,369,100

                                                              8,899,800   (1,530,700  increase)  =  >21% increase
                                                                          in  next  6  yr =. 3.5%/yr growth
 * For major increases
** Assumes 4%/yr increase in capacity from 1965 to 1977
Data are obtained from the following sources:
     (1)  Survey of Free World Increases in Copper Mine,  Smelter and  Refinery  Capacities  1971-1977,  International Wrought Copper
          Council, April 1972.
     (2)  Nonferrous Metal Works of the World,  First Edition  1967,  H.  G.  Cordero,  Editor.
     (3)  A. D. McMahon, CopperA Materials Survey, U.  S.  Dept.  of the,Interior,  Bureau  of Mines  IC-8225, 1965.

     (4)  Metal Statistics 1971, 64th Annual Edition, The American  Metal  Market Co.
     (5)  Yearbook of the American Bureau of Metal Statistics, 50th Annual  Issue for the  year 1970,  Issued June  1971.

-------
                                 TECHNICAL REPORT DATA
                          (Please read Instructions on the reverse before completing)
 i. REPORT NO.
 .
 EPA-650/2-74-085-a
                                                       3. RECIPIENT'S ACCESSION NO.
 4. TITLE AND SUBTITLE Control of ^^^ Dioxide Emissions
 from Copper Smelters; Volume I--Steam Oxidation of
 Pyritic Copper Concentrates
            5. REPORT DATE
             September 1974
            6. PERFORMING ORGANIZATION CODE
 7. AUTHOR(S)

 C. A.Rohrmann, H.T.Fullam, and F. P.Roberts
            8. PERFORMING ORGANIZATION REPORT NO
 9. PERFORMING ORONIZATION 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 AND PERIOD COVERED
             Final;  6/71-7/72	
            14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
 is. ABSTRACT The report presents results of a laboratory study on production of hydrogen
 sulfide by reaction between water vapor at 700-800C and iron sulfide contained in
 neutral-roasted pyritic copper ore concentrate.  Hydrogen sulfide thus obtained was
 to be reacted with sulfur dioxide emitted from copper smelter converters. In this
 manner sulfur emissions from a smelter could be controlled and recovered in the
 form of elemental sulfur. It was determined that the above treatment of copper ore
 concentrate could yield necessary quantities of hydrogen sulfide. The concentration
 of yielded hydrogen sulfide, however, was limited in the stream to less  than 1.0 per-
 cent by reaction equilibrium conditions, thus implying excessively high  energy
 requirements and causing this control approach to be significantly more expensive
 than known conventional processes. Alternative  means of hydrogen sulfide production
 are dealt with in the second volume of the report.
 7.
                              KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Group
 Air Pollution
 Sulfur Dioxide
 Copper Ores
 Smelters
 Smelting
 Hydrogen Sulfide
Air Pollution Control
Stationary Sources
Steam Oxidation
Pyritic Copper
13B
07B

11F
13H
 8. DISTRIBUTION STATEMENT

 Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
     74
                                           20. SECURITY CLASS (This page)
                                           Unclassified
                                                                   22. PRICE
EPA Form 2220-1 (9-73)

-------