The furnace described in this paper was installed and has been used by the writer for the melting of rich gold-bearing material, and has given complete satisfaction. It may, therefore, prove of interest to other metallurgists who are faced with a similar problem. A furnace of this type should be most suitable for the reduction of “ cyanide precipitate,” “ charcoal ash,” “ anode sludge”. Formerly the melting was done in fire-clay crucibles, using coke or crude oil as fuel, but this method is costly, slow, and unsatisfactory. A small 18-in. water-jacketed blast-furnace was also tried on this material after briquetting it, but was discarded in favour of the electric furnace.
The furnace may be described as a simple form of the Heroult resistance furnace, with an adjustable, upper electrode and a fixed bottom electrode. This is in accordance with Stansfield’s classification of electric furnaces. The heating is mostly due to the resistance offered the current by the molten slag.
The melting chamber consists of a steel drum 14 in. in diameter and 21 in. high, and is lined with silica bricks specially made to fit the shell. The internal diameter is 6 in. The bricks contain 92 % silica, are made from calcined quartzite bonded with fire-clay, and are similar in composition to those used for lining steel converters. They are set in a mixture of finely-ground quartzite and clay, and well dried out.
The shell is bolted to a steel plate, which forms the decking of a truck running on rails, so that the furnace can be moved under a hood for carrying off the fumes. A cast-iron slag spout is bolted to the shell below the tapping hole, which is 6 in. from the bottom of the shell. Opposite this another hole ¾ in. in diameter is drilled through the shell and lining for the passage of the lower cable. This is made up of four strands of No. 4 copper wire, about 24 in. long, and sweated into a “ lug ” at one end. The other ends are spread out near the bottom of the crucible, and are tamped round with a hot mixture of crushed graphite and distilled tar and covered for about 3 in. This acts as the lower electrode.
The electrical installation consists of a 20 k.v.a. transformer, immersed in oil, and supplied by single-phase current at 400 volts. The board is fitted with fuses, knife-switch, ammeter, and voltmeter.
From the secondary side a cable connects direct to the furnace bottom. From the board another cable connects to the fixed plate of a water rheostat. This is simply a wooden trough 24 in. square and 24 in. deep, with two steel plates 21 in. by 18 in. One is movable backwards or forwards, and the other is fixed. The movable plate is connected by a flexible cable to the upper electrode of the furnace. The rheostat is only used in starting operations by cutting down the current, and can be thrown out of circuit.
The upper electrode is of 2-in. round Acheson graphite, suspended over the furnace by a 3/16-in. flexible steel cable running over pulleys, and is raised or lowered by means of a small worm and wheel.
The electrode holder is a strip of copper 2 in. wide, 1/8 in. thick, and 18 in. long. It is bent at right-angles, and clamped for 6 or 8 in. down the side of the electrode. A 3/8-in. stud is also screwed through the holder and into the graphite to prevent slipping. The other end of the holder has a hole drilled through it so that the lug on the cable can be bolted to it. This seems a very crude type of holder, but answers better than several others tried. A fire-brick serves to keep the electrode central whilst the furnace is operating.
The material which is melted is finely divided, and varies greatly in composition. It contains from 20 % to 60 % of silica, together with iron oxides, alumina, &c. Air-slaked lime and a very small proportion of borax are the only fluxes used. A little sulphur or
galena is sometimes added to form a matte with any iron that may be reduced. No particular type of slag is or can be adhered to. If it is too basic the linings cut out rapidly. The higher the percentage of silica that can be used the better, as this means longer life to the linings, a greater resistance to the current, and a higher temperature. As the temperature attained is very great, the question of fusible slags is not a difficult one. Litharge and old cupel bottoms, together with sufficient coke dust to reduce the lead, are added to the charge to supply the carrier for the valuable metals. The amount of lead used depends on the richness of the charge ; usually 12 % to 15 % is sufficient.
On starting the furnace the rheostat plates are separated, the power switched on, the upper electrode is lowered until a steady arc is obtained, and a pound or two of slag is thrown into the crucible. When this melts, the crucible is filled with dampened charge. As the charge melts and the resistance becomes steady, the load in amperes is increased by adjusting the electrode until 225 amp. show on the meter, the voltage being about 80.
The furnace top is open, but the electrode is kept fairly cool by the unmelted charge surrounding it. This also protects it from excessive oxidation. Gases and steam are evolved, and cause a small amount of dusting. The furnace is surrounded by a frame lined with iron and provided with a flue for carrying off the fumes. All dust is swept up and returned to the furnace.
As the charge melts and subsides further material is fed in until the crucible is nearly full of molten slag. In tapping, a steel bar is driven through the tapping hole, and the molten slag and metal are run into a conical mould, provided with a lip. This acts as a settler or forehearth, and the excess slag overflows and is caught in other moulds, similarly shaped. The crucible is not completely emptied except at the end of the day. There is usually no difficulty in tapping, and the slag runs out white hot and very fluid. The separation of the values is good, considering the richness of the material melted. As an example, 231 lb. of material, containing 69 oz. 4 dwt. 12 gr. of gold, equal to 671.3 oz. per ton, gave slags worth 7 oz. per ton.
The amount of energy used per lb. of charge averages 1.1 kw. and the quantity melted per 8 hours varies between 60 and 100 lb. Slags are melted at a much faster rate. The silica-brick linings last from 36 to 72 hours.
Electrodes are joined by drilling and tapping 3/4-in. holes and screwing in a threaded iron dowel, 4 in. long. Before tightening up, a paste of graphite and tar is spread on the ends. When screwed together, good electrical contact is ensured. By this means very little of the electrode has to be discarded.
The metallic bottoms separated from the slag are placed in fire-clay crucibles and heated to liquate the lead bullion which is poured off. The residue is refined by stirring in nitre. This metal and the leady bullion is subsequently cupelled in a small gas-fired cupel furnace. The resulting rich bullion is treated chemically to separate the gold, silver, and, if present, platinum.
The advantages of electric melting are many. There is practically no danger from electric shock, no skilled labour is required for handling the melting is rapid, and the temperature is ample, and under perfect control.
Based on results obtained from this small furnace, a much larger one is to be installed.
MR. W. Poole said Mr. Mitchell had done well to bring that subject before the Institute, because it was a very important class of work that had come to stay in Australia. In addition to the works he mentioned there was also a company in Sydney— namely, the Australian Electric Steel Ltd., which, had been making steel castings for over two years, and had been doing very satisfactory work. An engineering firm in Newcastle was also about to install an electric furnace for making steel castings. Experimental work bad been made on zinc-sulphide ore from Queensland. The production of metallic zinc in an electric furnace from fully roasted sulphide ore was now a successful process, and a plant had been erected in Tasmania for that purpose. The experimental work carried out on the Queensland ore was an attempt to commercially produce zinc direct from the unroasted sulphide ore. It was reported that a British company was about to install works in North Queensland for the electrical-furnace reduction of wolfram and molybdenite ore.
MR. Mitchell, in reply, said one reason why these industries were not more heard of was the cost of current. Taking the minimum rate of the Melbourne Electric Supply Co., which supplied current at .66 of a penny, that worked out at £24 per h.p. per year. He understood that the lowest rate of the hydro-electric plants in Tasmania was .067 of a penny per unit, which worked out at a little over £2 per h.p. per year. That showed the disadvantage under which they laboured in starting such works on the mainland.