Preliminary Refining and Casting into Anodes —
Electrolytic Refining — Bringing to Pitch,
and Casting of Merchant Copper.
The further treatment of the converter metal depends to a large extent
upon its composition, and the purpose for which it is intended. The
matte-smelting operations on copper ores bring about the elimination
of the greater part of the constituents accompanying the copper. The
converter-grade matte may, however, in addition to the copper, iron
and sulphur, also contain considerable proportions of easily reducible
impurities of the ore, possessing a greater tendency to enter the matte
than to be oxidised and eliminated in the slag. Such constituents may
include gold and silver (practically all concentrated and retained in
the cupriferous product), arsenic, antimony, bismuth, selenium and
tellurium (retained to very considerable extent), as well as lead,
zinc, nickel and cobalt (in much smaller proportions). The amount of
these latter impurities ultimately retained in the converter matte
depends very largely upon the proportions originally present in the
ore, and upon the smelting conditions.
Under the strongly oxidising conditions of the Bessemer process the
copper retains but small quantities of impurity, and those which do
remain in ordinary converter metal may be broadly divided into two
classes[17]—(_a_) those which are oxidisable with comparative ease, and
(_b_) those which persist in the metal even under oxidising influences,
unless treated by special means. The former include iron, sulphur, and
zinc; the latter, arsenic, antimony, bismuth, selenium, tellurium,
gold, and silver. Keller gives the following figures for the average
elimination of the impurities in the converter:—
Iron, 99 per cent.
Sulphur, 99 "
Zinc, 99 "
Cobalt, 99 "
Bismuth, 97 "
Lead, 96 "
Arsenic, 81 "
Antimony, 71 "
Selenium, 47 "
Tellurium, 40 "
Nickel, 37 "
Of the persistent elements, the retaining of the gold and silver in
the converter-copper is a factor of much economic advantage, but the
other impurities are curiously just those which are characterised by
possessing most injurious effects on copper intended for electrical
work—for which purpose most of the material is employed.
The demand for particularly pure metal in electrical and conductivity
work therefore usually necessitates a further purification of the
converter-copper (unless it be an exceptionally pure brand) and the
production of metal specially free from the injurious constituents
which persist to a small but sometimes very appreciable extent in the
metal under the ordinary oxidising conditions. The presence of silver
and gold in the copper may afford in many instances sufficiently good
reason for a separating process independently of the market for the
pure copper itself.
In modern practice, electrolytic methods are almost universally
employed for the purification of the crude copper. By this means the
large demands of the present day can be conveniently met, and the
copper be obtained in a condition of remarkable purity. The frequent
presence of gold and silver in the metal, and the convenience and
completeness with which they are separated on electrolytic treatment
of the copper are particularly advantageous features which recommend
the adoption of electro-refining, and may in some cases be the reason
for this procedure even though the metal might otherwise be already
quite up to specification for electrical service. In the large majority
of cases, these bullion-values constitute a welcome and independent
bye-product, the returns from which may be set against the expenses of
the refining operations on the copper, which might, in any case, be
necessary.
The process may, therefore, be operated with one of the following
objects—
(_a_) Of purifying converter-copper.
(_b_) Of recovering from copper, the bullion-values which
have been collected in the metal.
(_c_) Of manufacturing pure copper, and recovering the gold
and silver as profitable bye-products.
Under the present industrial conditions, the electrolytic refineries
are located at centres often at very considerable distance from
the smelters. Situations for the refineries are chosen where the
local conditions as regards power supply, technical resources, and
particularly proximity to markets and distributing centres, allow
of the operations being conducted under the most advantageous
circumstances, and it is customary for smelters situated in the remoter
mining districts to ship the crude copper to these custom refineries,
instead of conducting the process themselves. At Anaconda, the
well-equipped electrolytic refineries have been closed down, and the
anode metal shipped to the Eastern refineries for treatment.
=Preliminary Refining of Converter Copper and Casting into Anodes.=—For
modern electro-refining practice, the crude metal must be prepared into
anodes, which are usually in the form of plates about 2 feet 6 inches ×
3 feet by 2 inches thick. It is found that the metal as produced in the
converters, on being cast into such plates, does not as a rule yield
anodes which work satisfactorily in the tanks. This is largely owing
to the impure and crude condition of the metal, which results in the
production of plates which are spongy, coarse, and exceedingly rough
and uneven on the surface. In consequence the direct employment of such
metal would occasion irregularity and difficulties in the operation of
the tanks, giving rise to short circuits, uneven wear, breaking off in
large pieces, and similar troubles. Furthermore, the tank liquors and
slimes become badly contaminated if large quantities of impurity be
present in the anodes, and the deposition of good clean metal is thus
greatly interfered with. All these reasons render it advisable that
the converter-copper should, as a rule, undergo a preliminary furnace
treatment before being cast into anodes.
The Anaconda practice is representative of the manner in which these
preliminary refining and casting operations are conducted, except that
the enormous scale and organisation of the operations are practically
unique. The principles involved and the general method of operation are
in all essentials those of the old Welsh furnace-refining process.
_The Furnace._—The finished metal from the converters is teemed into
ladles, and from these is poured directly into one of three casting
furnaces. Two of the furnaces are in constant use, one of them engaged
in refining, one being filled, and one in reserve or repair. Two of
the furnaces are 14 feet × 22 feet 8 inches hearth dimensions, with a
capacity of 95 tons; the third has a 14 feet × 28 feet hearth, and a
capacity of 110 tons—the fire-boxes being 5 feet 6 inches × 7 feet.
The furnace bottom is constructed of the local silica brick (which is
claimed to be the finest in the world) laid down in four beds, the
three lower being each 12 inches thick, whilst the working bed is
constructed of 20-inch bricks; brick being found to be better than sand
in this class of work. The bottom is curved to a depth of 2 feet. These
furnaces are, in consequence of their different function, constructed
on somewhat different principles to the reverberatory smelting
furnaces.
[Illustration: Fig. 69.—Sectional Plan, Elevation, and Transverse
Sections of Refining and Anode-Casting Furnace, Anaconda (_Peters_).]
Owing to the high conductivity of copper, and to the fact that the
functions of the furnace are either largely as a medium for simple
fusion or as a receptacle for molten metal, and further, that but
little slag is produced, that no settling and separation of the fluid
materials are required, and that there is no danger of dusting-losses,
the furnace may conveniently be built with a deep hearth which
need not be of very considerable length. The main requirements are
refractoriness of the building materials, particularly careful
construction so as to avoid breakouts, and very strong bracing indeed
on account of the deep and heavy bath of material which is carried on
the furnace hearth.
OPERATIONS—(_a_) _Oxidation Stage._—The furnace is loosely filled with
scrap copper which has accumulated round the works (8 to 12 tons), and
converter metal (of composition say about 98·3 per cent. copper) is
then poured in at the side door from ladles bringing it in quantities
of about 5 tons at a time, as teemed from the converters. When the
furnace is about half-filled, a blast of air at 90 lbs. pressure is
injected through the metal by means of iron pipes, which at this
stage just dip below the surface. These pipes are gradually eaten
away by oxidation and slagging action, but as the end wears down,
the pipe is pushed further in. The function of this air blast is to
supply oxygen for the purpose of acting upon the small quantities of
oxidisable impurity which remain in the metal after bessemerising, and
which consist chiefly of iron and sulphur, in addition to the small
quantities of metalloids. The oxygen partly acts directly on these
constituents, but as already indicated, the scouring action is to a
great extent performed by copper oxide which is produced and which is
itself a powerful oxidising agent. The iron appears to be one of the
first elements to be removed, and then a little sulphur, but this is
chiefly eliminated after the iron has been oxidised. The interaction
between the copper oxide and the sulphides liberates metallic copper
and yields SO_{2}, which bubbles up through the metal and gives to it
an appearance of “boiling,” by which name this stage is known. Too
rapid an oxidation during the early stages is dangerous if much sulphur
be present, owing to the evolution of sulphur-dioxide assuming a degree
of explosive vigour. Up to this point, the oxygen has been utilised
in removing iron, sulphur, etc., which are eliminated as oxides, so
that but little of the oxygen is retained in the metal, but after
the boiling stage is passed, oxygen is actually absorbed, the copper
now becoming oxidised, and the oxygen contents of the metal rapidly
increase. As in the analogous instance of steel bessemerising, it
appears essential to introduce some excess of oxygen into the metal in
order to ensure the complete removal of the oxidisable impurities, so
in copper-refining, an excess which amounts to about 0·7 per cent. of
oxygen (equivalent to about 6 per cent. of Cu_{2}O) must be introduced.
In the refining practice as conducted by the Welsh process, much of
this aëration took place during the slow melting down of the crude
blister copper, and subsequently during the flapping operations with
the rabble; but the use of the air-blast hastens this oxidation
considerably, especially as the metal is now often directly poured
into the furnace in a molten condition, so that oxidation during
melting is not possible. It is essential to defer this final oxidation
and elimination until it can be conducted at the refining furnaces
rather than to attempt it in the converter, since the refining furnace
allows of the operation being performed much more gradually and under
better control, whereas if conducted in the converter, the necessarily
vigorous action would occasion unduly heavy losses of copper in the
slag and probably excessive oxidation of the metal.
During the aëration, the furnace contents are continually added to, by
additional ladles-full of metal, and usually by the time the furnace
is filled the air-blast has oxidised most of the impurities from the
metal. These have entered the slag, and the copper has become “dry,”
owing to the necessary super-aëration. If this stage has not been
reached, the sample often shows “sprouting” (also known as “spewing”
or “throwing a worm”), which is caused by the escape of SO_{2}, and
indicates that all the sulphur has not been eliminated. In that case
the blowing is continued until small samples ladled out from the bath
exhibit the characteristics of dry copper, viz.: the depression down
the middle line of the ingot, brittleness of the metal, and a purplish
brick-like fracture.
These preliminary operations may occupy some three or four hours or
more.
(_b_) _Poling and Bringing to Pitch._—The oxidation having proceeded to
the required stage, the highly cupriferous slag is skimmed off, after
being first thickened with ashes from the fire-grate, and the poling of
the metal is then commenced. This operation is conducted by immersing
poles of timber, three to six at a time, in the metal, holding them
well under the surface and pushing them further in as the ends burn
away. It is essential that the timber should be green and not dry, and
preferably it should be hard wood, such as birch, beech, or oak. The
poles are usually as long as possible, and are from 6 to 8 inches
thick at the butt end. The function of the wood, particularly during
the early stages is, to a great extent, mechanical, and any chemical
changes effected are by indirect action.
The poling operation really consists of two stages, the first of which
is the final elimination of SO_{2} retained by the metal, and the last,
the actual reduction of the excess oxide and the “bringing of the metal
to pitch.”
The green timber, when inserted into the copper, liberates large
amounts of moisture and reducing gases which agitate the bath
considerably and “shake” the gas out of the metal more or less
mechanically, replacing, at the same time, some SO_{2} by CO and
hydrocarbons which copper possesses the power of absorbing. When the
SO_{2} has been satisfactorily eliminated, the reduction stage is
arrived at, and this is conducted in a manner similar to the familiar
poling operation of the Welsh process. The surface of the bath is
completely covered over with a layer of coke, anthracite, or charcoal,
and more poles are inserted. The exact mechanism of the operation has
not yet been definitely traced, but the action of the wood at this
stage is partly of a mechanical and partly of a chemical nature. The
reducing gases liberated by the charring and destructive distillation
of the wood have themselves a reducing action on the oxides which are
dissolved in the dry copper, but an important feature of the action of
these gases is the agitation and splashing which they occasion, thus
bringing the molten metal into close and vigorous contact with the
layer of reducing carbonaceous material maintained upon the surface of
the bath.
Poles are inserted usually two or three at a time, and samples are
constantly taken and examined for surface indications and for fracture.
This preliminary refining operation usually has for its chief object
the preparation of a fairly pure metal which will yield a sound, clean,
and even anode casting, and which is not required at this stage to pass
the rigid mechanical tests essential for the market product. In this
case it is therefore usual to carry the poling operation only to such a
degree that the samples ladled out and cast into small ingots solidify
with the even, smooth surface desired and which is characteristic of
“tough-pitch copper”—irrespective of any special mechanical properties
of the metal. If the test is satisfactory, the metal is ready for
casting.
The poling occupies some hours, and usually from 40 to 50 poles of wood
are used up before the metal is in a suitable condition for casting.
During these operations the coal fire in the grate is manipulated in a
manner best suited to the various stages of the process; there may thus
be an oxidising flame during the early part of the refining, but the
flame must be of a reducing character whilst poling is in progress.
_Casting._—Until comparatively recent years, the size of the refining
furnace has been necessarily limited to small dimensions, owing to the
difficulty in emptying the furnace of large charges. The practice,
as conducted hitherto, has been based on the familiar method of the
old Welsh process, viz., that of ladling out the metal by small hand
ladles. This involves so much hand labour, and requires such a long
period of time for its operation as to make practically impossible
any attempt to deal with large quantities of metal, or to lead to any
considerable increase in furnace capacity. The chief difficulties to be
overcome when operating on large charges of copper by this hand-lading
method are those of maintaining the metal at the correct pitch during
the lengthy period of ladling; whilst the large amount of time during
which the finished copper has to remain within prevents the furnace
being used for its chief purpose, that of refining more metal.
The method of hand ladling was employed for so many years on account
of the difficulties of controlling the stream of metal and of tapping
the furnace in the usual way—_i.e._, through a tap-hole at the lowest
point of the bath. These difficulties were due to the very high working
temperatures, to the great weight of metal behind the stream, which
forced it out under great pressure, and to the high melting point,
conductivity, and tendency to chill of the copper, which was apt to
cause setting of metal in the tap-hole, and led to the latter becoming
rapidly closed up and useless. Regulation of the stream of metal to a
gentle flow was impossible under such conditions.
With the introduction of casting machines by Walker, and the
improvements in the methods of tapping by the adoption of the vertical
tapping-slot, these difficulties have been removed, and the casting
of 100 tons of metal from one of the modern large casting furnaces
presents, to experienced workers, little practical difficulty.
The modern casting machine brings a series of moulds continually under
the supply of metal which issues from a large ladle fed continuously
from the tapping-slot of the furnace.
The method of tapping now used is to allow the copper to gently run
out of the furnace, by gradually lowering the level of a temporary
retaining wall which is constructed in a narrow vertical slot in the
tapping side of the furnace.
[Illustration: Fig. 70.—Indicating Tilting and Pouring Mechanism for
Ladle of Casting and Refining Furnace.]
This slot, which is /‾\-shaped in plan, extends from the lowest
point of the hearth to well above the highest possible level of the
liquid metal. It is about 3 feet high and 4½ inches wide, and whilst
the furnace is working it is kept rammed with a mixture of loam
and anthracite, this filling being supported by a series of short
transverse bars, 16 inches long and 1 inch square in section, which
are set 3 inches apart and rest upon lugs fixed to the iron plates
which strengthen the furnace-wall. During the operation of casting,
this hard filling can be readily cut away as required and the level
of the dam thus gradually lowered at will, permitting the gentle and
continuous overflow of the molten metal. The stream is also regulated
by inserting a pole of wood in the opening, should the flow become too
rapid, and by this means it is kept under absolute control. The molten
metal flows along a spout which feeds a small suspended ladle of about
800 lbs. capacity, the supply being so regulated that this ladle is
filled sufficiently slowly as not to get ahead of the moulds. The ladle
is supported hydraulically, and is pivoted so that it can be brought
forward and tilted for pouring, and then lowered and moved a slight
distance backwards, to allow the next mould to come into position.
On tilting the ladle, the metal flows gently and without splashing
through a three-hole grid in the front—which keeps back slag or
cinders—and runs into the mould, which is rapidly filled. In order
to prevent the metal overflowing in the mould, and also to rapidly
cool that portion which forms the lugs of the anode-plate, a hollow
water-cooled block 2 feet 6 inches long and of 6 inches square section,
situated opposite the ladle, is brought forward hydraulically into such
a position that it rests on the mould just against the edge of the lugs.
At many smelters the circular form of anode casting machine introduced
by A. L. Walker is employed. This apparatus consists of a horizontal
wheel which can be rotated slowly, carrying a series of arms at the end
of which the moulds are supported, so that they form a broken ring. By
the rotation of the machine, one mould after another can be brought
under the ladle and filled. The moulds are pivoted, so as to allow of
tilting, and when the metal has set, the ingot is thus dropped into
a cold water bosh, whence it is carried to the yards by a conveyor.
At Anaconda, the casting machine consists of a series of moulds
carried on a platform conveyor which is operated hydraulically—the
moulds are attached by bolting them on to the belt through lugs fixed
underneath. The moulds are constructed of 1 inch cast-iron, and allow
of the production of anode ingots 2 feet 6 inches × 3 feet by 2 inches
thick, provided with lugs at the corners of one end for the purpose of
supporting the plates in the tanks.
Each mould holds about 560 lbs. of metal, and when the anode has been
cast, the ladle is dropped back into position and the mould is moved
forward by means of the conveyor belt. After traversing a distance
equal to three times its own length, the ingot becomes fairly solid,
and at a point corresponding to this position the conveyor base
inclines slightly upwards. The cake is sprayed gently during its
passage over a distance of about 8 feet, the conveyor belt then passes
over a pulley-wheel, and when in a vertical position, the anode is
forced out of the mould by a crowbar and falls into a water bosh, from
which it is carried by another conveyor on to a platform. Here it is
wheeled to stacks, examined for flaws, and weighed. Sample anodes are
placed on one side, and the others are packed for shipment to the
Eastern refineries.
[Illustration: Fig. 71.—Walker’s Anode Casting Machine.]
[Illustration: Fig. 72.—General View of Tank-room of Electrolytic
Refinery, Perth Amboy, N.J.]
The furnace deals with one charge (usually of 100 tons, but
occasionally much more) per eight-hour shift, and the casting machine
yields 25 tons of anodes per hour.
Samples weighing from 4 to 6 ozs. are taken three times per shift from
the stream of copper running into the moulds, by batting the metal into
water with a wooden paddle. This method checks very well with drillings
taken from the anode plates, the chief discrepancy feared having been
with respect to silver contents, owing to the tendency of this metal
to segregate. The assay of the anode metal at Anaconda averages copper
99·3 per cent., silver 80 ozs. per ton, and gold 0·5 oz. per ton.
=Electro-Refining.=—Electrolytic refining was introduced on a
commercial scale by Elkington at Pembray in 1865, and with the general
adoption of the dynamo for the production of power, dating from about
1870, the process was greatly developed. Most of the copper now placed
on the market has passed through the electrolytic refinery.
_System of Working._—The method of arranging the electrodes in the
depositing tanks which is usually adopted at the great refineries at
the present day, is that known as the _parallel_ or _multiple_ system.
In this method of working, the anodes are all connected to one pole of
the circuit, and the cathodes, situated between them, are all connected
to the other. In this way, each tank comprises in reality one large
anode and one large cathode, and the voltage as measured between any
two neighbouring electrodes will be the same. The system thus allows of
currents at low voltage being employed, since the voltage is a factor
of the number of electrodes in series, and in consequence danger of
short circuiting is lessened. This allows of plates being placed closer
together in the tank, with less danger from this source of trouble.
A large number of tanks are employed at the refineries, and they are
usually arranged in series, the anode plates of one vat being connected
to the cathode plates of the neighbouring one, the current thus passing
from one vat to the other through the entire system.
Various other methods of arranging the electrodes have been favoured
from time to time, and of these the _series-system_ is the most
important, this being still in use at several large refineries, though
it has been generally superseded by the multiple method.
The plan underlying the series method was that of avoiding the
trouble and expense of preparing and working with the special cathode
sheets of pure copper as are necessitated by the multiple system.
In the series-method, each anode was made to serve as a depositing
surface for the pure cathode copper produced by the operation, so
that as impure anode copper was dissolved away on one side of the
“anode”-plate, pure copper was gradually deposited upon the other side.
This system appeared, therefore, to have several marked advantages to
recommend it, but in practical operation many difficulties in working
and several serious disadvantages were encountered. The chief points in
favour of the series-system are—
(_a_) Smaller first cost of the installation, particularly in the
matter of electrical connections, since the multiple system requires
heavy leads running along each side of the tank, as well as close
attention to the providing of good contacts, in order to connect all
the anodes and all the cathodes together with a minimum of current
leakage. In the series-method, the plates are readily connected one to
the other.
(_b_) The great saving of the cost of preparation and arranging for
specially pure cathode plates, this constituting a very important
factor in the costs of the multiple process.
(_c_) The output of metal per vat is greater.
On the other hand, the disadvantages of the system, except under
special conditions, are very serious.
(_a_) More scrap is produced and requires re-treating, owing to the
difficulty of separating the new deposit from the remaining portions of
old anode, which often adhere very firmly.
(_b_) Higher voltage through the tanks is required, owing to the
large number of electrodes in series in the bath. Hence the danger of
current-leakage and short circuiting is greatly increased, especially
when impure anodes are used, since they tend to produce conducting
layers of mud on the bottom of the vat.
(_c_) The anode plates have to be made particularly smooth and even on
the surface, since in order to lessen the voltage required, the plates
are brought as close together as possible, in consequence of which,
any excrescences upon the surface greatly increase the danger of short
circuiting. Preliminary furnace-refining and special straightening of
the anodes are therefore essential in connection with the series method.
(_d_) Special tanks are required, as the protecting lead liner
cannot be employed, since the danger of current-leakage through it
is increased, owing to the higher voltages required. Hence special
acid-resisting material, such as slate, is necessary, the expense of
which is considerable.
(_e_) The cost of stripping the cathodes is high, and the operation is
often difficult.
(_f_) The cost of maintaining the plant is greater.