You are on page 1of 25

Copper extraction from the 60's.

into the 21'' century


W. G. Davenport
Dept. of Materials Science and Engineering
University of Arizona
Tucson, AZ, USA 85721
ABSTRACT
Changes in copper extraction from 1960 till today are documented. The top ten
changes have been:
replacement of reverberatory smelting by high intensity oxygen rich smelting
growth of the Outokumpu flash smelting to over 50% of the world's smelting
capacity
successful development of single furnace coppermaking but only for low slag fall
concentrates
replacement of the batch Peirce-Smith converter by continuous converting, but
only in a few cases
increased SO2 capture throughout the industry, mainly as sulfuric acid
development of low initiation temperature 'big bight' Cs catalysts for treating the
continuous high-SO2 strength gases from continuous smelting/converting
complete replacement of reverberatory anode scrap and cathode melting h a c e s
by the Asarco shaft furnace
adoption of stainless steel permanent cathodes and automated stripping
technology for electrorefining and electrowinning
complete elimination of wire bar casting by continuous bar castinghod rolling
development and adoption of extractants for turning weak impure leach solutions
into strong pure electrolytes.
It is postulated that the biggest possible change over the next 20 years would be complete
replacement of smelting/converting by hydrometallurgical processing. However, this
seems unlikely due to copper purity, precious metal recovery, and economic concerns.
The increasing value of sulfuric acid to many copper companies gives chalcopyrite
smelting/oxide-supergene leaching a nice synergy especially with the energy credits now
coming from continuous smelters and their acid plants.

Proceedings of
Copper W-Cobre 99 Lnlernational Conference
Volume I-PlenaryLectures/Movement of Copper and Industry Outlook/
Copper Applicalions and Fabrication
Edited by G. A. Eltringham, N.L. Piret and M. Sahoo
The Minerals, Metals & Materials Society. 1 9 9

VOLUME I

INTRODUCTION
Around. 1960, when I first became,interestedin copper smelting, the reverberatory
furnace was king around the world (1). It was very effective .at recovering copper fr.om
concentrate and it was very reliable, having (been used for over a century in various
forms.
As many have found out by thinking that old reverberatory slags might serve as
ore for a money-making enterprise, its slags contained only 0.3 or 0.4% Cu showing just
how efficient it was. Not only that, converter slags could be recycled back through it,
.getting their Cu-in-slags down to that same low level.
However, the reverb didn't make much use of the energy available in its sulfide
concentrate feed to provide the heat for melting slag and matte. More importantly, its
SO2-in-offgasconcentration was so low as to make SO2removal almost impossible. So,
it usually vented all its SO2 to the atmosphere.
There were also a few blast (shaft) h a c e s around in the 1960's. Collection of
their SO2 was also difficult. The Momoda modification allowed quite good SO2
collection but its production rate was so slow as to make it uneconomical. The last
Momoda adoption was in 1967 at the Huelva smelter (2). The Legnica and Glogow I
smelters in Poland still operate six blast (shaft) furnaces for copper concentrate.
It was clear even in 1960 that this reverbblast furnace situation could not go on
forever. Processes that would efficiently capture S& would have to be found and
adopted.
Likewise, the Peirce-Smith converter, though chemically very efficient, was not a
good device for capturing SOz. Its SO2 concentrations of 5 - 8% SO2 were certainly
better than those of the reverberatory smelting furnace. But its discontinuous batch
nature made efficient SO2capture difficult.
So, the writing was on the wall. Processes would have to be developed and
adopted that would pennit (i) efficient capture of smelting and converting offgases and
(ii) efficient fixation of the SO2 from those gases in a useful form, almost always sulfuric
acid.
PROCESSES AVAILABLE IN THE 1960's
Of course, flash smelting already existed in the 1960's. Outokurnpu Oy had
installed its first flash furnace shortly after World War I1 (Table 1) to minimize fossil fuel
consumption. lnco had developed its flash smelting process a little later to make use of
the nearly pure oxygen that could be made cheaply from hydroelectricity. Both produced
continuous streams of strong-SO2offgas from which SO2 could be recovered as

' o p ~ $ u 'oh q p n s u! sdem lapun d a y 8 ~ a q


1111s' p q l a w ayl uo ' s e 8qlaurs
~
qsea o:,q .aJayMasla uayl pm! mder t~! 1Cgs~y'~'09
Alma p m S,OS ale1 ayl u! paldope aq 01O u ~ 8 a sq e Omqams
~
ysen ndnmyolno
Bu!l~aurs qsag

pyl

.sey I! l C e atp
~ 8 q a r n s laddo3 a l e v o p 01auro:, pInoM
a u q spq le lea13 rnoy nj sem I! $nu -20spmbq lo p13e :,"J[~s

VOLUME I

58

As for Outokumpu flash smelting, the role of the Furukawa engineers in making
it a viable economic process cannot be overly praised. Anyone who has scrutinized their
paper 'The Development of Flash Smelting Process at Ashio Copper Smelter, Furukawa
Mining Co., Ltd' (3) cannot fail to realize the huge contribution these engineers made in
advancing flash fUrnace technology for the world.
As far as I can remember, these were the only smelting processes that could be
said to offer the promise of.high efficiency SO2 capture. Electric furnaces held out some
promise but only by shifting SO2 evolution to roasters ahead of the smelting furnace.
So that was it in the 1960's. But there were whispers ofnewcopper smelting
technology coming on the scene, particularly from Noranda and 8Mitsubisbi. And in 1967
patents were issued for Noranda and Worcrasmelting.
THE 1970's

The winds of change began to blow in the late 60's and early 70's. Converter
gases began to be caught and' their SO2 processed. Outokumpu smelting began to be
adopted more fkequently and new processes came on the scene (Table 1)(4). The 70's
may be thought of as the golden age of new copper smelting processes with the adoption
of
Electric furnace smelting
Kivcet smelting
Mitsubishi smelting
Noranda smelting
Outokurnpu flash smelting
TBRC smelting
Teniente smelting
and the .trial ofi
Worcrasmelting.
Worcra smelting never reached the commercial scale and Kivcet smelting for
copper has never developed past one or two small units in the former Soviet Union.
Likewise, TBRC copper smelting was adoptediin several;places but.has not expanded for
smelting copper concentrates.
FATE OF PROCESSES ADOPTED IN THE 1970's

Table 2 follows through the 1980's and 1990's the processes developed and
adopted in the 1970's. It shows that Outokumpu smelting was the most widely adopted
followed by Teniente smelting, Mitsubishi smelting and Noranda smelting. Not shown

PLENARY LECTURESIMOVEMENT OF COPPER AND INDUSTRY OUTLOOK1


COPPER APPLICATIONS AND FABRICATION

in the table is the fact that many electric furnaces went out of service during this period
due to high energy costs.
So the three manifestations of copper smelting:
particulate smelting (Outokumpu flash)
submerged tuyere smelting (Teniente and Noranda)
lance smelting (Mitsubishi)
proceeded together: flash smelting in the lead, submerged tuyere smelting second and
lance smelting third.
Teniente smelting has a somewhat unusual position in this list because it seems to
always need 'seed' matte from another process, e.g. reverberatory smelting (5) or flash
smelting (6): Seed matte production may account, however, for as little as 10% of the
smelter's total smelting capacity. Nevertheless, as now operated, it is mainly a process
for expanding existing smelting.capacity rather than for greenfield smelters.
NEW AND NEWLY ADOPTED PROCESSES IN THE 1980's AND 1990's
The 1980's and 1990's saw continued adoption of Outokumpu flash, Teniente,
Noran& and Mitsubishi smelting. They also brought some new actors on to the scene:
Contop cyclone smelting
Inco flash smelting
Isasmelt lance smelting
Vanyukov submerged tuyere smelting
Contop cyclone smelting was first used to increase the smelting rate of
reverberatory furnaces. The Asarco El Paso smelter then successfully operated a standalone Contop furnace from 1993 to 1999. After early problems with cyclone burner
corrosion, this furnace operated well, albeit with a somewhat high energy consumption
(7). Unfortunately, the El Paso smelter has been temporarily mothballed because of
copper's low price so that the future of Contop smelting is somewhat cloudy.
Inco flash smelting had a brief flurry of activity in the 80's with two retrofit
adoptions as reverberatory replacements, both in the southwestern United States. These
two units have operated reliably since the early 80's. They are now producing at up to
twice their design smelting rates. But no further units have been adopted outside Inco
since this brief flurry. Perhaps Inco's lack of enthusiasm for selling technology has
limited the process' adoption. It should be noted however that Inco recently built two

59

VOLUME I

60

Table 2 Smelting Processes Adopted 1980-1999

198.1
1982
1983
1985*
1986*
.1987
1988.

) Norilsk, Russia

-~

I
1

Camacari, Brazil
Isabel, Leyte, Philippines
Guixi,China
El
Mexico
-.Taio.
...
Srednogorie, Bulgaria
Chuquicamata, Chile
a-7

Gujarat, India
MinasCerais, Brazil
Olympic Dam, Australia
in.progress
Ronnskar, Sweden
in :progress i 110,.Peru (smelting)
in.pr0gres.s : 110, Peru (converting)
znco Flash
1983*
Hayden, Arizona
1984*
Hurley, New Mexico
1995
Sudbury, Ontario (2)
Noranda Process
1990*
I Port Kembla, Ausbalia
199?
1: Daye, China.
2000
I Shenyang, China
Teniente
- - - .- - -Procers
- - --1984*
1 Chuquicamata, Chile
1984*
1 Las Ventanas, Chile
1985.
Polrerillos, chile
1989
I Paiuote. Chile.
Nkana, Zambia
1998
1998
1999*

La Caridad, ,Mexico
1998

Mitsubishi'Smeltinn/Cowerling
.I.Naoshima, Japan.
I, Onsan. Korea
- - - -- '1 Gresik, Indonesia
Isasamelt
1992*
Miami, Arizona
1997
Tuticorin, India
in progress
:Belgium
Contop
1993*
'1. .El,Paso,Texas#
1991*
1998
- -.
.
1998

1
I

PLENARY LECTURESIMOVEMENT OF COPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FABRICATION

furnaces for its Sudbury smelter and plans to use the technology for its Voisey's Bay
operation.
Vanyukov smelting is tuyere smelting. Unlike Teniente and Noranda smelting its
furnace is not rotatable, leading to potential problems with unexpected tuyere-blowing
outages. This process was examined closely by Western Mining for its Western
Australia operations but it was not adopted. The Ballcash Copper Smelter Vanyukovs
were being operated in conjunction with a reverberatory furnace - it was not clear that
they were being operated "stand-alone". Only one (of two) could be operated at a time.
Isasmelt lance smelting was adopted in 1992 by Cyprus's Miami Arizona
operation and it continues to be adopted slowly around the world. It has been adopted for
copper smelting by Sterlite (India) and is being constructed at Union Miniere. Lance
lives were originally rather short but adoption of matte-resistant alloys and better control
procedures have considerably smoothed production. The Isasmelt furnace has small
conductive, convective and radiative heat losses and it is particularly good for low sulfur
concentrates and mixed charges of moist concentrates, residues and scrap.
Of these 198Oys/90's processes, then, only the Isasmelt seems to be moving
ahead.
The 1990's saw improvements to all the previously-adopted processes. However,
they did not see radically new developments. Converting, on the other hand, saw the
radical development of flash converting and continuous submerged tuyere converting,
discussed next.

CONVERTING IN THE 1966's


Just as the reverb was king for copper smelting in the 19607s,the Peirce-Smith
converter was king for converting. As I remember, it was the only available converting
process. It was great except for its :batch nature and its inefficient capture of its SO2
offgas.
This deficiency has been attacked on,nurnerousfronts since then by:
(a)
(b)

(c)
(d)

Hoboken converting - axial flue gas collection and continuous blowing of PeirceSmith type converters
Mitsubishi and TBRC lance converting (8,9)
Kennecott-Outokumpu flash converting (10)
Noranda continuous submerged tuyere converting (1 1).

Hoboken converting began to be adopted around 1970. About 20 Hoboken


converters are currently in place and three more were planned for the Thai Copper
smelter (Rayong, Thailand). The Hoboken's advantages of efficient, continuous gas

61

VOLUME I

collection seem to have been partially offset by problems with splash and dust buildup in
the gas offtake system. This appears to have slowed wider adoption of the process.
One aspect of tuyere smelting/converting that should be mentioned here is
Noranda's late 1960's development of the Gasp6 tuyere-punching system. Operating
almost everywhere now, and,.automaticallyin some smelters, it significantly improved
the rate and uniformity of.air and oxygen flow into the converter.
Mitsubishi converting has followed along with adoption of the Mitsubishi
smelting-converting system, so there are now four MI converters in place (Table 3).
Significantly, the first stand-alone Mitsubishi converter is under construction at the Port
Kembla Copper smelter in New South Wales (12). Its outcome is eagerly awaited.
The 1990's saw two new players come on the scene - Kennecott Outokumpu
flash converting (10) and Noranda continuous converting (1 1). Neither can yet be said to
be 'arms 'length' tested but both are working well for their developers. An 'arms-length'
flash converter is being built at Southern Peru Copper's 110 copper smelter.
Flash converting takes high-grade matte from the smelting flash furnace and:
(a)
(b)
(c)

granulates it with water


grinds and dries it in a vertical roller mill (10)
oxidizes it to high sulfur blister copper in a small, highly-oxygen enriched, single
burner flash converting bee.

There 'have been a few teething problems with the Kennecott flash converting
installation, but the operation.has stabilized enough to convince Southern .Peru Copper to.
include a flash converter in its new 110 smelter.
Noranda continuous converting is submerged tuyere converting (11). It is similar
in many ways to the Noranda smelting process in that it continuously feeds Fe- and Sbearing materials into a rotary fiunace while continuously blowing 02-enriched air into
the bath through tuyeres running about half the length of the rotary furnace. The feeds to
continuous converting will usually be molten Noranda smelting process matte added
through a small mouth, and solid materials added via slinger at one end of the furnace.
Gas is continuously drawn from the converter through a second mouth and hood,
an evaporative atomized water spray cooler, a dust removal system and on into an acid
plant. From Noranda's ,point of view, it extends to converting the expertise gained with
the Noranda smelting system.

PLENARY LECTURESFIOVEMENT OF COPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FABRICATION

Table 3 Continuous Converting Installations


1974*
1979*
199 1*
1998
1998
1999

Mitsubishi Converting
1 Naoshirna,. J a v d
Timmins, Ontario
Naoshima 3a~an
Onsan, Korea
I Gresik, Indonesia
Port Kembla, Australia

Flash Converting
"

1995*

in Droeress

1 Garfield, Utah
I 110. Peru

Noranda Continuous Converting


1 Noranda. Ouebec

1998

*Visited by author. #Closed

Converting Slag
Submerged tuyere converting processes use traditional silica-base slag. It is a
convenient match for the silica base slags used in smelting and it, or a slag concentrate
made from it, can be conveniently recycled back to the smelting furnace for copper
recovery.
Mitsubishi found that this slag was not compatible with lance converting because
solid magnetite formed where the lance oxygen impacted the slag, greatly inhibiting the
converting reactions. They developed a CaO based Ca - Cu - Fe - 0 slag which has a
large liquidus region at converting temperatures. A similar slag is used by the Kennecott
flash converter. This slag appears to be more corrosive than the traditional silica based
slag so it is only used where absolutely necessary. Noranda submerged tuyere
continuous converting does not use it because the turbulence created by submerged
tuyere blowing prevents formation of a solid magnetite layer.
It seems that the traditional silica slag will continue to be used wherever possible.
Through this all we can see the trend that converting is continuing to develop
exactly as smelting, i.e.:
submerged tuyere converting (Hoboken, Nomda Continuous)
lance converting (Mitsubishi)
flash converting (Kennecott/Outokumpu).

63

VOLUME I

SINGLE FURNACE PRODUCTION OF COPPER


The avowed goal of chemical and metallurgical engineers in the 1960's was
single furnace production of copper. This was a natural hope because smelting and
converting both oxidize Fe and S from the smelter's sulfide concentrate feed. Combining
smelting and converting to produce metallic copper by the reaction:
CuFeS2 + oxygen

+ CuO+ iron oxide + SO2

seemed to be the natural way to go.


The ,potential advantages of single furnace metallic copper production were seen
to be:
(a)
(b)
(c)

(d)

maximum use of sulfur and iron combustion energy with consequent


minimization of industrial oxygen and fossil fuel consumptions;
restriction of S& emissions to a single continuous source;
no moving of matte between furnaces with consequently less emission of fugitive
SO2-bearing gas;
elimination of a process step.

As it turned out, single furnace' copper metalmaking ,had two significant


disadvantages:
(e)
(f)

the slags from the process contained up to 20% Cu hence a significant fraction of
the Cu-in-feed;
impurity levels in the copper were high due to their continuous absorption into
molten metallic copper in during smelting.

Through the years a number of single furnace coppermaking configurations were


developed namely:
Noranda continuous submerged tuyere smelting
Worcra continuous lance smelting
flash direct-to-blister smelting.
Worcra lance smelting never left the pilot plant stage due to its,short lance -lives.
The Noranda process, on the other hand, successfulljr produced metallic copper for
several: years before being switched to high-grade matte production. Short tuyere .lives,
slag foaming and high impurity 1evels.h copper seem to have sled to this new direction
However, direct-to-blister flash smelting has operated continuously since 1977,
fust at the Polish Glogow I1 smelter (13) and then also at the Australian Olympic Dam
smelter. The rather small direct-to-blister-flash furnace at Olympic Dam has just been
replaced by a significantly larger unit, showing complete confidence in the process.

PLENARY LECTURESIMOVEMENT OF COPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FABRICATION

In Poland, the flash furnace process produces a Cu-Fe-Pb alloy that requires
converting to metallic copper. It is not clear, then, that the process should really be
called single-furnace copper making. It is operated in competition with the blast furnaces
in the Glowgow I and Legnica smelters, which have turned out to be somewhat better at
removing impurities (specifically As and Pb) than the direct-to-blister furnace (1 3).
The Olympic Dam direct-to-blister flash furnace has, on the other hand, been a
great success, resulting in installation of a new, larger furnace to match an expanded rate
of concentrate production. Initially the Olympic Dam smelter did not have a remelting
unit for melting copper skulls and other revert materials, so it was hampered somewhat in
the handling of its recycle products. Installation of an electric furnace solved this
problem.
Interestingly this smelter operates alongside a uranium leaching operation which
uses all the smelter's sulfuric acid product. In fact, sulfuric acid production from the
flash furnace gases is sometimes -augmentedby burning sulfur. This is a far cry from the
excess acid production situation of many smelters.

FUTURE FOR SMELTING AND CONVERTING


Well, where are smelting and converting going?
competitive trends:

We have seen the ,three

flash smeltinglconverting
submerged tuyere smeltinglconverting
lance srneltinglconverting.
Are there others, I wonder. Or are there combinations of the above that might
make sense? And what about hydrometallurgy? It is already nibbling at the edges as a
good way to treat supergene sulfide minerals (CuS, Cu2S). Might it not also nibble into
chalcopyrite extraction? This will be mentioned again later.

FLASH SMELTING
It has always seemed to me that, in principle, flash smelting is the perfect way to
treat the fine flotation concentrates - because the concentrates are exactly the right size
for flash oxidation. Just blow the concentrate particles uniformly into a space, surround
them with 0 2 and let #themreact. Supply (i) just the right 02/concentrateratio so that the
reaction can proceed exactly to the desired extent and (ii) just the right amount of N2 to
get the heat balance right while producing the desired matte grade and temperature.

65

VOLUME I

Of course the concentrates need to be dried, but thermodynamically, it is better to


dry concentrates in a low temperature dryer (say with steam) than in a #hightemperature
smelting furnace.
The ,technical downsides of the flash furnace have been said to be its. excessive
amount of auxiliary equipment, namely its:
(a)
(b)

(c)

blast heating system


multiple burner feed system
waste heat boiler.

Recent years have seen (i) oxygen replace blast heating and (ii) single concentrate
burners replace multiple concentrate burners. So disadvantages (a) and (b) have
disappeared.
However it is certainly true that waste heat boilers have caused significant
downtime in many Outokumpu flash furnace.smelters. Safe and Jones (14) identify tube
leaks as the most serious problem and indicate the following causes:
(a)
(b)

(c)

excessive rapping of the tubes to remove dust (now of diminished


importance because of single-burner flash smelting dust reduction, 14a)
poor welds
uneven expansion in the boiler during temperature fluctuations.

The temperature fluctuations often arise because of erratic operation of the flash
furnace during startup teething problems.
Overall, however, generation of steam by the waste heat boiler for electricity and
heating has a thermodynamic attractiveness that is absent in processes that cool their
gases by vaporizing water. Its electricity can be extremely valuable where electricity
prices are high. As with metallurgical processing, 1999 boilers are much improved over
1960's era boilers.
INCO FLASH SMELTING

i
h ,the 1976.edition of Extractive Metallurgy of Copper, Anil Biswas and I made
the statement:
"The Outokurnpu design is not well adapted to the use of oxygen because the
combustion tower [reaction shaft] tends to overheat while at the same time remote
comers of the hearth require fuel-fired heating. Thus, for the use of oxygen, flash
furnace designs will likely evolve towards 'the INCO system of horizontal concentrate
burners, which are particularly well suited to heating the entire hearth area."

PLENARY LECTURES/MOVEMENT OF COPPER AND INDUSTRY OUTLOOK1


COPPER APPLICATIONS AND FABRICATION

Well, we could not have been more wrong. It has tuned out that an aerodynamic
central single concentrate burner in an efficiently water-cooled cylindrical reaction shaft
has been a super concentrate combustion combination (15). Oxidation conditions can be
made uniform and they can be adjusted to increase or decrease the thickness of a
magnetite layer on the inside of the shaft. Changes in cooling water temperatures around
the shaft quickly tell the operator when his reactions are not uniform.
Reaction shaft control has, therefore, turned out to be readily manageable
especially with a single central concentrate burner. Furthermore, steady operation of the
Outokumpu furnace at high smelting rates has greatly reduced the need for fossil fuel
combustion in the settler part of the furnace.
With the Inco furnace, smelting rate may ultimately be limited by keeping the
roof above the horizontal burners from overheating. There does not seem to be a way to
run the Inco furnace with a single burner, single feed system.

SUBMERGED TUYERE SMELTING


TenienteMoranda tuyere smelting accounts for about 15% of world smelting
capacity, a very significant amount. The advantages of tuyere smelting area very high
specific smelting rate and violent turbulence in the matte-slag-gas bath. The latter is
especially important for the Noranda process which is designed to smelt large quantities
of copper and precious metals scrap. Teniente smelting seems to be well-adapted to
smelter production rate augmentation as its development throughout Latin America has
shown (Tables 1 and 2).

LANCE SMELTING
Mitsubishi (8) and Isasmelt (16) smelting account for about 10% of the world's
smelting. Both processes continue to be adopted around the world.
The compact Isasmelt furnace with its small heat losses will continue to be
adopted for smelters which treat low sulfur concentrate, revert, scrap charges. The
process' strongestpoint is its ability to treat a wide range of feed materials.
The Mitsubishi process continuously produces copper metal from concentrate. Its
current commercial competitors are (i) direct-to-blister flash smelting, (ii) flash
smeltinglflash converting, and (iii) Noranda submerged tuyere smelting/converting.
Its advantage over direct-to-blister flash smelting is that it can continuously
produce copper from concentrates that have a high slaglmetal product ratio (i,e.
chalcopyrite concentrates) without pushing too much.Cu into slag. A discardable slag is
created between smelting and converting and the small amount of slag produced in the
converting furnace is easily handled by recycle to the smelting furnace.

67

VOLUME IN

68

Its advantage over flash smeltingMash converting is that it does not (i) solidify its
matte between ,the two processes or (ii) recover its Cu-hm-smelting-slag as solid slag
concentrate that requires complete resmelting.
Itsadvantage over Noranda smelting/converting is its well-tested methodology.
So one has to ask why it has not been adopted more. My guess is that in most
locations flash smeltingPeirce-Smith converting have been efficient enough to meet
current environmental requirements. This combination is so well known ,that there has
been little risk in adopting it.
However, the coming of fiash srneltinglflash converting may force new smelters
to avoid the Peirce-Smith converter and go for:
flash smeltinglflash converting (Kennecott/Outokumpu)
lance smeltinflance converting (Mitsubishi)
tuyere smeltinglcontinuoustuyere converting (Noranda)
or some combination thereof.

In this race, the Mitsubishi process might well do well. Whatever, the race will
be interesting.
FUTURE OF CONVERTING
While one can extol the natural virtues of the flash furnace for smelting
particulate concentrates, it is not so easy to do this for converting. After all, the natural
l uJi matte, not solid articulate matte. Thermodynamics would
product of smelting is &
seem to ask that the matte be kept liquid.
So might we see flash smelting followed by 'lance or submerged tuyere
converting? My own view is that, against my thermodynamic wishes, we are likely to
see the above mentioned pairings kept intact just to make the smelting and converting
expertise interchangeable. So far this has been true, except perhaps at Port Kembla
where we are seeing lance converting adopted to follow Noranda submerged tuyere
smelting.

ACIDMAKING
A smelter typically produces 2.5 to 4 tonnes of sulfiuic acid per tonne of product
copper. This underscores acid's importance in the overall smelter flowsheet. Also, a
smelter cannot do business if its gas collectionfacid plant system does not achieve its
government-mandated SO2 removal efficiency.

PLENARY LECTURES/MOVEMENT~OFCOPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FABRICATION

The main success in recent years has been the low initiation temperature, big
bight cesium activated vanadium pentoxide catalyst (17,18). This catalyst permits
greater S02+S03 conversion in the catalytic converter and lowers the emission of SO2
fiom the plant. It has been particularly useful for the continuous high SO2 strength gases
from Mitsubishi and flash smeltinghlash converting.
Catalysts with even bigger temperature bights will be very useful as smelter
offgas SO2 strengths continue to increase.
Production of steam from gas sensible heat and SO,-absorption heat during acid
manufacture is an advantageous aspect of the continuous high SO2 strength gases
produced by continuous smelting/converting. This practice may now considered to be
conventional. It is the methodology of choice for future acid plants.

My first experiences with fire refining were at Canadian Copper Refiners in 1967
and the Gasp6 smelter in 1968. What spectacular changes there have been since then.
CCR was completely dominated by a row of large hearth furnaces for melting
spent anodes and scrap. They resembled open hearth steelmaking furnaces with. their
huge long-armed charging machines. A batch of scrap took forever (about a day) to ,melt
and refine with a spectacular expenditure of labor.
The deoxidation was done with green poles.
Even in the early 1980's the Phelps Dodge El Paso refinery had its row of scrapmelting hearth furnaces, out of service but ready to go. I'm glad to say that they're now
gone.
These changes were all brought about by the Asarco shaft furnace.
The Asarco shaft furnace has had a huge impact on (i) scrap anode melting and
(ii) cathode melting, equally as ,big an impact as the new smelting and converting
processes have had on extraction from concentrate. Its energy expenditure is well under
half that required by the hearth furnaces it replaced and its labor expenditure is very
small indeed.

In the 1960's wood poles were a common way to deoxidize copper during anode
casting. However gas 'poling' had already begun. Wood poling may have been
appropriate in hearth furnaces but submerged.tuyeregas deoxidation in rotary furnaces is
certainly the better process.

69

VOLUME I

70

ANODE CASTING
Continuous casting of anodes by the Hazelett or Contilanode system has been
available for 40 years. Yet only a half dozen smelterslrefmeries have adopted it.
Obviously traditional wheel-and-mould casting, equipped with an expert weighing
system and anode preparation machine is the process to beat.
Perhaps production of a solid strip of solid copper followed: (by cutting oddshaped anodes from it is fundamentally the wrong way to make anodes. It is, after all,
making a two-step process from a one step process. Or it may be simply that the
maintenance:costsof the continuous-castinglcuttingsystem .are rather high.

The CCR rehery in which 1. worked in 6'967 had been built about 1930. Its
defining feature formewas how labor intensive it was. Mind you, I should have known
about extractive:,metallurgy.labor intensity because in the 1950's I worked pulling and
stripping zinc cathodes by hand as a university student.
An interesting sidebar to my CCR employment was a visit to the Phelps-Dodge
refinery in Laurel Hill, New York City. Laurel Hill was still using the series refining
system which had a high productivity (very useful in New York City) but produced a
somewhat low grade product.
The major changes in electrorefining since the 1960's have been (19):
(a)
(b)

(c)
(d)
(e)
(f)

(g)

adoption of stainless steel permanent cathode technology throughout the industry,


along with automated stripping
installation of anode preparation machines to ensure flat, uniform anodes
development of new electrolyte purification techniques, particularly solvent
extractionlion exchange
installation of polymer concrete cells in place of all earlier cell technologies
use of periodic reversal of current to increase plating rate (now rather rare)
on-line determination of addition agent chemistry
electrolyte filtration.

From the point of view of what one sees in the 1999 refinery, it is the degree of
mechanization that is most evident.
Cells are filled automatically, emptied
automatically, drained automatically and so on. So electrorefining has gone from a
labor-intensive industry to an automated industry.
And these trends are continuing.

PLENARY LECTURESFlOVEMENT OF COPPER AND INDUSTRY OUTLOOK1


COPPER APPLICATIONS AND FABRICATION

LEACHINGISOLVENT EXTRACTIONIELECTROWINNING
Copper leachinglcementation has been practiced for centuries, in Spain and Wales
for example. It is still being practiced in a few spots. However, its product is an impure
copper metal that requires complete smeltinglrefining to produce copper for market. It
has been completely overtaken by leachinglsolvent extractiodelectrowinning, which
directly produces market-grade copper metal.
Likewise copper ,leachinglelectrowinningwas practiced extensively in Chile and
Zaire. Unfortunately its .product grade was lower than needed for electrical use.
So until the late 6OYs,copper leaching could not produce the quality of copper
needed for the electrical industry. This greatly inhibited any growth of leaching as a
major influence on copper production. Hydrogen reduction of Cu from solution was
being tried but it did not stick as an industrial-scale process even though reasonable
copper purity could be obtained. In fact, just as Anil Biswas and I were writing the first
edition of Extractive Metallurgy of Copper, hydrogen reduction of copper at the Bagdad
mine in Arizona had just given way to solvent extractiodelectrowinning (20,21).
Unfortunately Anil and I were not exactly brave with our 1976 prediction as to
the outcome of solvent extraction. Our exact words were:
"Industrial copper solvent extraction plants are few in number as of yet, but this process
MAY assume considerable importance in the near future."
We wish now that we had been more adamant.
The coming of solvent extraction, particularly with LIX hydroxy oximes ended
the impurity problems of electrowinning and made leachinglsolvent extraction/
electrowinning the successful process it is today. In 1999, 15% to 20% of the world's
primary copper is produced by this method. Quality-wise, LISXEW copper is fully
competitive with smeltedlrefined copper. Lead is the only significant impurity (from the
Pb-Sn-Ca anodes). This is often complemented by the low-lead cathodes of
electrorefining (if a company produces both refined and electrowon cathodes).
The process has improved markedly since its early days. The extractants and
diluents have improved as have techniques for avoiding their degradation and loss.
Reagent consumptions are down 30 to 40% over the last few years (to about 2 kg
extractant per tonne of copper) and their price has come down a shilar.amount. This has
had one unfortunate side .effect, restriction of extractant production to only .a couple of
manufacturers, a disturbing competitive situation.
Other improvements have been an enhanced ability to prevent C1, Fe and other
impurities from advancing forward to electrowinning. This has improved cathode quality
and made the SXEW process easier to control.

71

VOLUME I

72

Also, leach/SX/EW costs are almost independent of plant size, which means that
a small operation can compete with a large one. As one looks around the Chilean
SXEW plants, it is not unusual to see large plants (100 000 tomes copperfyear)
operating within a few kilometers of quite tiny operations (2000 tonnes copperfyear).
Because most of the electrowinning half of SXEW operations are so new, most
use or are adopting stainless steel permanent cathode technology and polymer concrete
cells. They are also highly mechanized with mechanical stripping machines and cell
loading unloading systems. There are two competitive permanent cathode technologies,
Isa and Kidd. The main difference between them is cathode edge stripping and the
stripping machine technology. They seem to be competitive.
WHERE IS HM)ROMETALLURGY GOING?
In 1999 leaching for copper is treating 'oxide' minerals and supergene sulfide
(Cu2S, CuS). Some mines now have to make a decision as to whether to send the
supergene-rich ore to 1eacWSXIEW or to flotation/smelting!reflning. The tendency today
seems towards leach, the lower cost option. Of course, this depends on ore grade, the
amount of unleachable CuFeS2,acid availability and equipment needs.
Chalcopyrite ores, on the other hand,.:must always be sent tothe concentrator,
because their leach rate is too slow to be economic.

In 1981, when I moved to Arizona, Duval's chloride leach CLEAR Process was
producing about 80 tonnes per day of electrowon copper from chalcopyrite. And several
other processes were operating at about 1/10 that scale.
However, all shut down during-thelow copper price days of the mid-1980's. The
problem was cost and the purity of the copper that was being elkctrowon. frdm the
chloride leach solutions.
It seems to me that a hydrometallurgical process must be able to produce
marketable cathode copper rather than any other form. Based on electrowinning history,
this will require a solvent extraction step between leaching and electrowinning.
A factor not yet clear is as to. whether the byproducts of leaching will be more
acceptab1e:environmentallythan those of smeltinglrefining. It may be that SO2 gas.is as
easy to handle as sulfur-bearing leach residues.

PLENARY LECTURES/MOVEMENT OF COPPER AND INDUSTRY OUTLOOK1


COPPER APPLICATIONS AND FABRICATION

CATHODE MELTINGICASTINGIROD ROLLING


Wirebar casting was king in the 1960's. These 75 to 150 kg bars were easily
transported and fed into rodwire mills for further manufacture. However, development
of the Asarco shaft furnace and Southwire and Hazelett continuous bar casting quickly
put an end to casting individual bars.
The continuous bar is fed hot into a hot rolling machine which produces the rod
feed for wire drawing. The only limitation on the length of rod that is produced is
convenient rod coil shipping length. The coils, once in the wire drawing machine, can be
welded to give as much wire as needed.
Replacement of individual wire bar casting with continuous bar casting is clearly
advantageous. It is, in fact, the opposite of the anode casting situation where individual
anodes are the desired product rather than long (weldable) lengths of rod.
The Southwire system entails casting between a wheel-rim mould and a steel
band. The Hazelett system entails casting between two steel bands. They seem to be
competitive.
Other. operative castingholling systems are Properzi wheel-and-band casting,
dipforming.and Outokumpu upwards vacuum casting. They account for considerably
less tonnage that the Southwire and Hazelett systems..
Concluding about casting, then, the cathode melting and casting area of a refinery
is virtually unrecognizable from the 1960's version. In many refineries it consists of only
two things - an Asarco shaft melting furnace and a continuous bar castinghod rolling
machine.

CONCLUSIONS AND PREDICTIONS


1.

Oxygen-rich smelting will continue to be the process of choice for copper


smelting, whatever the furnace configuration. It increases productivity and
simplifies gas collection.S0~ fixation.

2.

The Outokumpu flash furnace reaction shaft is ideally suited to oxidation of


flotation concentrates. All that is needed is to distribute the particles evenly in the
shaft, surround them with just enough oxygen to produce the desired matte and
add just the right amount of nitrogen to balance the furnace thermally. Feeds such
as reverts and scrap are better handled by other smelting processes namely
submerged tuyere smelting and lance smelting.

73

VOLUME I'

74

Outokumpu flash furnaces will continue to adopt single concentrate burners for
new furnaces and as retrofits of existing furnaces. The single burner
configuration offers:
(a)
a simple feed system
more
uniform feed and blast distribution in the reaction shaft
(b)
better
reaction shaft conGol, particularly of the magnetite coating
(c)
on the inside of the reaction shaft
(d)
significantly reduced dust evolution.
Copper matte converting will move away from discontinuous Peirce-Smith
processing.
converting to continuous
,.. ..

The feed to converting is liquid matte. Thermodynamically this favors.converting


processes that treat liquid matte, i.e. lance converting and submerged. tuyere
converting.
These two processes and flash converting will continue, to compete for 'the
continuous converting market.
Flash converting has the advantage of an easily-controllable oxidation system
with an excellent ability to capture S02. It has the disadvantages of requiring (i)
that its matte feed be solidified and ground and (ii) that Cu be recovered from its
slags by resmelting. It also lacks the ability to melt scrap and revert materials. In
its present manifestation, it seems to require a somewhat corrosive lime-based
slag to prevent solid magnetite formation in the converting furnace.
Lance @litsubishi$ converting has the advantages of being truly continuous with
a molten matte feed. When combined with Mitsubishi smelting it also produces a
discardable smelting slag rather than slag concentrate for resmelting. It does,
however, require the somewhat corrosive lime base slag mentioned for flash
converting.
Submerged tuyere (Noranda) converting has the advantages of using a molten
matte feed and the ability to melt large quantities of scrap. It also has the
advantage of being able to use conventional silica-base slag, because the intense
stirring caused by tuyere-injected aidoxygen keeps a layer of solid magnetite
fiom forming. However, as currently practiced, it requires ladle transfer of matte
rather than continuous flow between smelting and converting. It also requires
resmelting of the copper from its smelting and converting slags.
Single-furnace coppermaking will. be restricted to those concentrates which
produce little slag. Concentrates with large slag falls send too much Cu to slag,
recovery of which,is expensive.

PLENARY LECTURESIMOVEMENTOF COPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FA,BRICATION

If existing smelters are forced to drop Peirce-Smith converting in favor of


continuous converting processes, they will likely stick with their existing
technology, i.e. flash smelters will adopt flash converting, submerged tuyere
smelters will adopt submerged tuyere converting and lance smelters will adopt
lance converting.
Greenfield smelters will now look at smelting and converting as a package, i.e.
flash smelting will be followed by flash converting; lance smelting by lance
converting and submerged tuyere smelting by tuyere converting. These pairings
will greatly simplify a project's .technology.
When choosing a process for recovering the heat fiom a smelter's offgas, a boiler
should always be given first consideration. It has a thermodynamic energy
advantage over water vaporization.
A critical part of any smelter is its sulfuric acid plant.
Continuous
srneltinglconverting technologies produce stronger SO2 .gas, which is putting
increased demands on SO2 conversion and SO3 absorption. Low initiation
temperature, large temperature range ('big bight') cesium base catalysts are
helping with these strong gases but further research into low initiation
temperature, big bight catalysts would be beneficial.
Production of steam from sensible heat and SO3-absorption 'heat during acid
manufacture is an advantageous aspect of the continuous,high SO2 strength gases
produced by continuous smeltinglconverting. This practice may now considered
to be conventional. It is the methodology of choice for future acid plants.
. ..

Anode casting will continue to be predominantly done by mould-and-wheel


casting/followed by an anode preparation machine rather than by continuous
casting of copper sheet followed by cutting of anodes fiom it. Mould-and-wheel
casting is a single unit process while cutting anodes from a continuously cast
copper sheet is a two-step process.
Continuous bar castinglrod rolling, on the other hand, has many advantages over
individual wire bar casting and will always be the process of choice.
Leachinglsolvent extraction/electrowinningwill continue to grow for 'oxide' and
supergene sulfide minerals.
This is the cheapest way to produce
electricaVelectronic grade copper fiom these resources. This process will
continue to grow especially in Chile, Argentina and Peru. The huge leachable
resources in Africa will not be exploited until a solid infrastructure is built up in
that region.
Solvent extraction plants have reduced their extractant losses some 30 to 40
percent in the last few years down to about 2 kg per tonne of copper cathode.
Separation of the pumping and mixing functions in the mixer-settlers has

VOLUME I

76

accomplished much of this along with newly efficient organic-fiorn-aqueous


recovery systems. The price of extractant has likewise dropped 30 or 40% over
the last few years. An unfortunate side effect of this has been reduction of
extractant suppliers down to two, a disturbing trend for the leach/solvent
extraction1electrowinning industry.
20.

Copper electrorefining and electrowinning tankhouses will continue to adopt


stainless steel permanent cathodes. Their main advantage is that they allow
almost complete automation of tankhouse materials handling. They also:
(a)
avoid starting sheet manufacture
give excellent cathode verticality (by avoiding the curling
(b)
of copper starting sheets).
Important benefits of the verticality are:
(c)
purer cathodes
(d)
fewer short circuits
(e)
higher current efficiencies
(f)
tighter anode/cathode separations
(g)
higher productivity.

21.

Polymer concrete will continue to be the material of choice for electrolytic cells.
Polymer concrete cells greatly simplify cell installation, operation and
maintenance.

22.

Leaching of chalcopyrite concentrate was quite successful in the 1980's and work
is continuing today. Impurity problems with the copper from these 1980's
processes lead to the conclusion that electrowinning will always be preceded by a
solvent extraction step.

ACKNOWLEDGEMENTS
The author wishes to acknowledge the following persons who so kindly provided
him with information during the preparation of this paper.
Cameron Hanis (Kvaerner)
Jackson Jenkins (Cyprus)
David Jones (BHP Copper)
Brian Kennedy (Simons)
Matt King (Phelps Dodge)
Gary Kordosky (Henkel)
Phil Mackey (Noranda)
Eric Partelpoeg (Winters Co.)
Tim Robinson (CTI-Ancor)
Akira Yazawa (Tohoku University)

PLENARY LECTURES/MOVEMENT OF COPPER AND INDUSTRY OUTLOOK/


COPPER APPLICATIONS AND FABRICATION

This paper was written while the authorZwasvisiting Professor at the Institute for
Advanced Materials Processing, Tohoku University, Sendai, Japan (Professor Yoshio
Waseda, Director). The author wishes to thank the faculty and staff of 'the Institute for
their wholehearted assistance.

REFERENCES
1. A. K. Biswas and W. G. Davenport, Extractive Metallurgv of Comer, Elsevier
Science Ltd., New York, N.Y., USA, First, Second and Third Editions, 1976, 1980,
1994.
2. P. Barrios, M. Palacios, H. Donicke, and J.-P. Nepper, "Expansion of the Copper
Smelter in Huelva", Sulfide Smelting '98. Current and Future Practices, J. A.
Asteljoki and IRL. Stephens, Eds. The Minerak, Metals and Materials Society,
Warrendale, PA, USA, 1998,559-568.
3. T. Okazoe, T. Kato, and K. Murao , "The Development of Flash Smelting Process at
Ashio Copper Smelter, Furukawa Mining Co., Ltd., Prrometallurgical Processes in
Nonferrous Metallurgy, J. N. Anderson and P. E. Queneau, Eds., A.I.M.E., Gordon
and Breach Science Publishers, New York, 1967,175-195.
4. W. G. Davenport, "Copper Smelting to the Year 2000", CIM Bulletin, January 1980,
152-158.

5. R. Alvarado, G. Achurra and R. Mac-Kay, "Present and Future Situation of the


Teniente Process", Sulfide Smelting '98. Current and Future Practices, J. A. Asteljoki
and R. L. Stephens, Eds. The Minerals, Metals and Materials Society, Warrendale,
PA, USA, 1998,493-501.
6. J. H. Meza Viveros and J. M. Velazquez. V., "Mexicana de Cobre Smelter Expansion
Project", Mining Engineering, Vol. 50 (1 l), 74-76.
7. M. Bmeggerman and E. Caba, "Operation of the Contop Process at the Asarco El
Paso Smelter', Sulfide Smelting '98. Current and Future Practices, J. A. Asteljoki and
R. L. Stephens, Eds. The Minerals, Metals and Materials Society, Warrendale, PA,
USA, 1998,159-166.
8. E. Oshima, T. Igarashi, N. Hasegawa, and K. Kiyotani, 'Waoshima Smelter Operation
- Present and Future", Sulfide Smelting '98, Current and Future Practices, J. A.
Asteljoki and R. L. Stephens, Eds. The Minerals, Metals and Materials Society,
Warrendale, PA, USA,'1998,549-558.
9. "Principles
and
Features
of
the
http://www2.mmc.co.jp/sren/p&f.htm, April 1999.

Mitsubishi

Process",

77

VOLUME I

78

10. C. J. Newman, T. I. Probert, and A. J. Weddick, "Kennecott Utah Copper Smelter


Modernization", Sulfide Smelting '98. Current and Future Practices, J. A. Asteljoki
and R. L. Stephens, Eds. The Minerals, Metals and Materials Society, Warrendale,
PA, USA, 1998,205-215.
11. M. Boisvert, G. Janneteau, J.-P. Landry, C. A. Levac, D. Perron, F. McGlynn, M.
Zamalloa, .and F. Porretta, "Design and Construction of the Noranda Converter at the
Home Smelter", Sulfide Smelting '98. Current and Future Practices, J. A. Asteljoki
and R. L. Stephens, Bds. The Minerals, Metals and Materials Society, Warrendale,
PA, USA, 1,998,569-583.
12. M: Goto, B. Oshima and M. Hayashi, Control Aspects of the Mitsubishi' Continuous
Process, Journal of Metals, Vol. 50(4), 60-65.
13. J. Czemecki, Z. Smieszek, S. Gizicki, J. Dobrzanski and M. Warmuz, "Problems with
Elimination of the Main Impurities in the KGHM Polska Miedz S.A. Copper
Concentrates from the Copper Production Cycle (Shaft furnace Process, Direct
Blister Smelting in a Flash Furnace)", Sulfide Smelting '98. Current and Future
Practices, J. A. Asteljoki and R L. Stephens, Eds. The Minerals, Metals and
Materials Society, Warrendale, PA, USA, 1998,3 15-343.
14. P. Safe and D. M. Jones, "Process Off-Gas Cooling Design Considerations for-NonFerrous Metallurgical Processes", Sulfide Smelting '98, Current and Future Practices,
J. A. Asteljoki and R. L. Stephens, Eds. The Minerals, Metals and Materials Society,
Warrendale, PA, USA, 1998,401-415.
14.a D. M. Jones and W. G. Davenport, "Minimization of Dust Generation. 'in
Outokumpu Flash Smelting", EPD Congress 1996, Garry Warren, Ed. TMS Annual
Meeting, Feb. 4-9, 1996,81-94.
15. Y. Suzuki, C. Suenaga, M. Ogasawara and Y. Yasuda, "Productivity Increase in
Flash Smelting Furnace Operation", Sulfide Smelting '98. Current and Future
Practices, J. A. Asteljoki and R. L. Stephens, Eds. The Minerals, Metals and
Materials Society, Warrendale, PA, USA, 1998,587-595.
16. R. R. Bhappu, K. H. Larson and R. D. Tunis, "Cyprus Miami Mining Corporation
Smelter Modernization Project: Summary and Status", paper prepared for 1997 TMS
Annual Meeting, San Francisco, February 27-March 4, 1994.

PLENARY LECTURES/MOVEMENTOF COPPER AND INDUSTRY OUTLOOK1


COPPER APPLICATIONS AND FABRICATION

17. S. M. Puricelli, R. W. Grendel and R. M. Fries, "Pollution to Power: A Case Study of


the Kennecott Sulfuric Acid Plant", Sulfide Smelting '98. Current and Future
Practices, J. A. Asteljoki and R. L. Stephens, Eds. The Minerals, Metals and
Materials Society, Warrendale, PA, USA, 1998,451-462.
18. P. M. Ritschel, R. C. Fell, R. M. Fries, and N. Bhambri, "Metallurgical Sulfuric Acid
Plants for the New Millennium" paper presented at Sulphur 98, Tucson, Arizona,
November 1-4, 1998..
19. J. M. Schloen and W. G. Davenport, "Electrolytic Copper Refining - World
Tankhouse Operating Data, Hvdrometallursw. Comer 95, Proceedings of the Copper
95 Conference, Santiago, Chile, November 1995.
20. Enmneerina and Mining Journal, "Fii-st Commercial-Scale Hz Reduction Plant for Cu
On Stream", Vol. 168(1), 97-100.
21. World Mining, "How Bagdad Uses LIX to Recover Copper fiom Dump Leach
Solutions", Vol. 24(4), 46-48.

79

You might also like