Part 4 — SMELTING ZINC RETORT RESIDUES[13] By E. M. Johnson
Reading edition
SMELTING ZINC RETORT RESIDUES[13] By E. M. Johnson
(March 22, 1906)
The following notes were taken from work done at the Cherokee Lanyon Smelter Company, Gas, Kansas, in 1903. It was practically an experiment. The furnace was only 36 × 90 in. at the crucible, with a 10 in. side bosh and a 6 in. end bosh. There were five tuyeres on each side with a 3 in. opening. The side jackets measured 4.5 ft. × 18 in. The distance from top of crucible to center of tuyeres was 11.5 in.
The blast was furnished by one No. 4½ Connellsville blower. The furnace originally was only 11 ft. from the center of tuyeres to the feed-floor, and had only been saving about 60 per cent. of the lead. This loss of lead, however, was not entirely due to the low furnace. As no provision had been made to separate the slag and matte, upon assuming charge I raised the feed-floor 3 ft., thereby changing the distance from the tuyere to top of furnace from 11 ft. to 14 ft. Matte settlers were also installed. These two changes raised the percentage of lead saved to 92, as shown by monthly statements. The furnace being small, and a high percentage of zinc oxide on the charge, the campaigns were naturally short. The longest run was about six weeks. This was made on some residue that had been screened from the coarse coal, and coke, and had weathered for several months. This particular residue also carried about 10 per cent. lead. The more recent residue that had not been screened and weathered, and was low in lead, did not work so well. Although these residues consisted of a large proportion of coal and coke, it seemed impossible to reduce the percentage of good lump coke on the charge lower than 12.5 or 13 per cent. At the same time the reducing power of the residue was strong, and with the normal amount of coke caused some trouble in the crucible.
When residue containing semi-anthracite coal was smelted, the saving in lead dropped, and the fire went to the top of the furnace, burning with a blue flame, thereby necessitating the reduction of this class of material. This residue had been screened through a five-mesh screen, and wet down in layers, becoming so hard that it had to be blasted. The low saving of lead with this class of material was a surprise, as it has been claimed that the substitution of part of the fuel by anthracite coal did not affect the metallurgical operations of the furnace.
The slag was quite liquid and flowed very well at all times. However, there was a marked variation in the amount at different tappings. This, I am satisfied, was not due to irregular work on the furnace, but may be accounted for in the following manner. The residue (not screened or weathered to any extent), consisting approximately of one-half coal and coke, was very bulky, and while there was about 35 per cent. of it on the charge by weight, there was over 50 per cent. of it by bulk, not including slag and coke. In feeding, therefore, it was a difficult matter to mix the whole of it with the charge. Several different ways of feeding the furnace were tried. The one giving the most satisfactory results was to feed nearly all of the residue along the center of the furnace, in connection with the lime-rock, coarse ore and coarse iron ore, and the fine and easy smelting ores along the sides. The slag was spread uniformly over the whole furnace, while the sides were favored with the coke. The charge would drop several inches at a time, going down a little faster in the center than on the sides.
It is possible that a small proportion of the residue in connection with the easy smelting, leady, neutral ore, iron ore and lime-rock formed the type of slag marked No. 1.
SiO2 FeO MnO CaO ZnO Pb Ag
1 33.7 34.1 1.0 16.5 7.5 0.9 0.7
2 31.0 36.1 1.2 16.0 9.6 1.3 —
This being tapped with a good flow of slag, the charge would drop, bringing a proportionately large amount of residue in the fusion zone which formed the type of slag marked No. 2. There was also a marked variation in the slag-shells from different pots. The above cited irregularities of course exist to a certain extent in any blast furnace.
AVERAGE ANALYSIS OF MATERIALS SMELTED
Name SiO2 FeO CaO MgO ZnO Al2O3 Fe2O3 S Pb Cu Ag Au
Mo. iron ore 10.0 65.0 ; ; ; ; ; ; ; ; ; ;
Lime rock 1.5 ; 52.0 ; ; ; ; ; ; ; ; ;
Mo. galena 1.5 2.4 ; ; 9.5 ; ; 11.0 74.0 ; ; ;
Av. of beds 50.8 16.2 ; ; 4.6 ; ; 3.3 9.1 ; ; ;
Residue[14] 10.5 38.5 ; ; 18.0 ; ; 4.8 2.2 1.0 10.0 0.03
Roasted matte[15] 9.0 48.0 3.0 ; 10.0 ; ; 4.0 9.9 3.0 21.0 0.06
Barrings 18.8 24.4 5.0 ; 14.5 ; ; 6.0 25.4 ; 13.0 0.07
Coke ash 27.0 ; 14.9 4.5 ; 19.7 31.6 ; ; ; ; ;
H2O V.M. F.C. Ash S ; ; ; ; ; ; ;
Coke[16] 1.2 2.3 85.7 11.1 0.9 ; ; ; ; ; ; ;
ANALYSIS OF BULLION, SLAG AND MATTE PRODUCED
Bullion Slag Matte
Ag Au SiO2 FeO MnO CaO ZnO Pb Ag Ag Au Pb Cu
Feb. 90.0 1.15 31.2 35.9 1.0 14.5 10.3 0.88 0.98 19.0 0.04 8.7 1.5
March 93.1 1.63 31.3 37.2 1.0 13.9 11.1 0.71 1.30 21.0 0.06 8.0 2.5
April 104.3 1.59 29.8 37.7 2.7 13.9 11.4 0.52 1.40 23.0 0.07 7.0 3.5
May 90.0 1.24 30.0 37.3 2.2 14.1 9.3 0.86 1.10 25.4 0.07 5.1 4.0
July 78.7 1.00 32.2 37.4 1.0 13.9 9.8 0.50 1.15 21.3 0.03 8.9 4.0
Aug. 90.8 1.21 31.2 37.1 1.7 13.7 9.6 1.10 1.60 23.1 0.08 9.8 3.0
Sept. 65.3 2.58 32.0 39.7 0.8 14.1 8.1 0.80 1.30 18.6 0.06 7.6 2.3
Average 87.5 1.49 31.1 37.5 1.5 14.1 10.0 0.77 1.26 21.6 0.06 7.8 3.0
MONTHLY RECORD OF FURNACE OPERATIONS
Blast Ounces Tons per F. D. Per Cent. Pb. on Charge Per Cent. Coke on Charge Per Cent. Slag on Charge Per Cent. S on Charge Matte Produced Saving
Ag Au Pb
Feb. 21 42.5 9.0 12.0 30.0 3.7 8.0} 84.4 83.0 90.3
March 21 44.8 9.7 13.5 37.0 4.0 9.0}
April 21 43.7 9.0 13.5 35.0 4.3 10.0 97.9 70.5 96.6
May 21 49.4 10.0 13.5 30.0 3.5 6.5 95.6 109.5 88.8
July 17 41.0 9.8 12.5 34.0 3.8 6.0 97.9 90.0 92.9
August 18 47.0 9.3 13.0 32.0 3.7 6.3 86.2 107.5 87.6
Sept.[17] 15 51.0 7.3 13.0 30.0 2.8 4.6 92.9 94.0 95.6
Average 45.6 9.1 13.0 32.6 3.7 7.2 90.8 92.4 92.0
I believe that, in smelting residues high in zinc oxide, better metallurgical results would be obtained by using a dry silicious ore in connection with a high-grade galena ore, provided the residue be low in sulphur. This was confirmed to a certain degree in actual practice, as the furnace worked very well upon increasing the percentage of Cripple Creek ore on the charge. This would also seem to indicate that alumina had no bad effect on a zinky slag.
ZINC OXIDE IN SLAGS By W. Maynard Hutchings
(December 24, 1903)
From time to time, in various articles and letters on metallurgical subjects in the Engineering and Mining Journal, the question of the removal of zinc oxide in slags is referred to, and the question is raised as to the form in which it is contained in the slags.
I gather that opinion is divided as to whether zinc oxide enters into the slags as a combined silicate, or whether it is simply carried into them in a state of mechanical mixture.
For many years I have taken great interest in the composition of slags, and have studied them microscopically and chemically. The conclusion to which I have been led as regards zinc oxide is, that in a not too basic slag it is originally mainly, if not wholly, taken up as silicate along with the other bases. On one occasion, one of my furnaces for several days produced a slag in which beautiful crystals of willemite were very abundant, both free in cavities and also imbedded throughout the mass of solid slag, as shown in thin sections under the microscope. In the same slag was a large amount of magnetite, all of which contained a considerable proportion of zinc oxide combined with it. Magnetite crystals, separated out from the slag and treated with strong acid, yielded shells of material retaining the form of the original mineral, rich in zinc oxide; an inter-crystallized zinc-iron spinel, in fact. I have seen and separated zinc-iron spinels very rich in zinc oxide from other slags. They have been seen in the slags at Freiberg; and of course everybody knows the very interesting paper by Stelzner and Schulze, in which they described the beautiful formations of spinels and willemite in the walls of the retorts of zinc works.
I think there is thus good ground for concluding that zinc oxide is slagged off as combined silicate, and that free oxide does not exist in slags; though zinc oxide does occur in them after solidification, combined with other oxides, in forms ranging from a zinkiferous magnetite to a more or less impure zinc-iron, or zinc-iron-alumina spinel, these minerals having crystallized out in the earlier stages of cooling.
The microscope showed that the crystals of willemite, mentioned above, were the first things to crystallize out from the molten slag. The main constituent was well-crystallized iron-olivine-fayalite.