The early development of the reverberatory furnace for smelting copper ores was the work of the Welsh smelters, particularly those of Swansea. The first record of a reverberatory furnace is made by Jars, who states that copper-smelting was effected in reverberatory furnaces at Middleton-Tyas, in Yorkshire, England. The first patent, of any importance, for improvements in reverberatory furnaces was granted to Thomas Williams for the granulation of the regulus. The next was that of William Evetts, in Sheffield, for the cooling of the fire-bridge by the admission of air.

Charge-hoppers above the furnaces were in common use. The size of the hearth of the furnaces about the beginning of the 19th century was commonly 11 by 8 feet.

No material development was made until the Welsh process was brought to Colorado, where Richard Pearce, as manager of the Argo works, near Denver, developed the furnace to meet the requirements of a custom plant, under keen competition with lead-silver smelting-plants, using blast-furnaces.

His improvements are outlined in Fig. 1, elaborated by permission from E. D. Peters’s excellent work, to which the reader is referred for further details. This figure shows that the furnaces were, 9 ft. 8 in. by 15 ft. in hearth; 16 by 35 ft., while the capacity of the furnace had been increased from 12 tons per 24 hr. to 50 tons in the same period.

The next step in development was made in Butte, Mont., by the Colorado Smelting Co.—this plant being, at that time, affiliated with the Argo works, so that Mr. Pearce’s influence was apparent. The step referred to was the lengthening of the hearth to 50 ft., with consequent increase in capacity to 105 tons in 24 hours.

The first furnace of this size—built from the Colorado Smelting Co.’s plans—was constructed at the Butte & Boston plant in Butte, Mont., in the year 1900. Details of Montana practice, will be found in Prof. H. O. Hofman’s paper.

The 50-ft. furnace became very popular in Butte and Anaconda, and when the Washoe smelter was built, in 1900-1902, the 50-ft. reverberatory was adopted and 14 of these furnaces were installed.

The next improvement was the addition of lime-rock to the charge of the calciners, so that it was thoroughly mixed into the charge before dropping the same into the reverberatories. This apparently insignificant change resulted in increasing the capacity of the furnace greatly.

The next step was the building of a furnace with a 60-ft. hearth, the results from which were so encouraging that a furnace with a hearth 85 ft. long was tried, with corresponding increase in tonnage. Then a radical step was taken at the Washoe plant, of connecting two 50-ft. furnaces, making a single furnace with a hearth 102 ft. long. The saving in fuel, and the increased tonnage therefrom, caused the management to build a furnace with a hearth 116 ft. long. The results of these experiments are given in the tabulated data below, compiled by William Wraith, Superintendent of the Washoe smelter.

Effect of Lengthening Reverberatory Furnaces

The reverberatory furnace as originally built at the Washoe smelter had a hearth-area of 19 by 50 ft. After some months of operation it was decided that a longer furnace could be operated to a better advantage. To determine the length best suited to the conditions, one furnace was lengthened to 60 ft.; another to 85 ft.; another to 102 ft.; another to 112 ft.; and, finally, to 116 ft. in length.

The draft at the bridge-wall was from 0.75 to 1 in. of water; the fire-box area, 7 by 16 feet.

The coal used was from the Anaconda Copper Mining Co.’s mine at Diamondville, Wyo., having the following average proximate analysis and thermal values :

There was some variation in the quality of the coal—ash ranging from 6 to 16 per cent., and the thermal value from 10,750 to 12,000 B.t.u. per lb. of wet coal; or, from 11,000 to 18,200 B.t.u. per lb. of dry coal.

There is also some variation in material smelted and in manipulation by the different furnace-crews.

The copper-assays of the slags from the different reverberatory furnaces averaged :

Reverberatory furnaces in different localities present their own conditions, and the length of furnaces will be found to be established by the type of fuel, draft, tonnage to be smelted, and investment.

About this time, the Garfield plant of the American Smelters Securities Co. was being built near Salt Lake City, Utah, and for that plant furnaces with 100-ft. hearths were adopted. Then, at Humboldt, Ariz., the Arizona Commercial Copper Co. built two furnaces of 100-ft. hearth, using oil for fuel. The Steptoe plant, at McGill, Hev., followed with five furnaces with hearths 110 ft. long, coal-fired, changed to oil-firing in 1911. The Cananea Copper Co. put in one furnace with hearth 100 ft. long, for flue-dust, in 1906; followed by a second in 1911. Experiments were tried with coal-dust firing, but cheap oil being available, the coal-dust experiments were discontinued and oil substituted.

The reader is referred to Dr. Ricketts’s article on experiments in reverberatory practice at Cananea.

Table I. gives important data on the best modern installations of reverberatory furnaces, and Table II. presents the details of the chemical composition of the charge, the slags, the mattes, and the fuels. Fig. 3 is a diagram showing the tonnage- and assay-curves of the Anaconda furnace.

The features that distinguish the modern reverberatory are : its length; solid bottom (monolith of quartz fused); structural steel conker-plates for bridge-wall; frequent charging; infrequent skimming of slag; the large body of molten matte retained in the furnace to assist in melting and distributing the charge; no leveling of charge by means of rabble; thick roof (from 15 to 20 in. of silica brick); fettling only once, a month in many plants, practically continuous operation; recovery of waste heat in the form of steam; and, where coal is used for fuel, the recovery of the unburnt coal from the ashes.

During the development of the reverberatory furnace to its present state of efficiency, a great many schemes were tried out and abandoned; for example, the preheating of the air, by passing it under the furnace-bottom or around the walls of the flue or fire-box, and forced blast under the grate.

A great variety of fuels has been used, and is being used, in reverberatory practice; for example, at Kyshtim, Perm Government, Russia, producer-gas from wood is in use, with very gratifying results. At this plant the hearth of the furnace is about 35 ft. between ports and 15 ft. wide; the gas is made from pine wood, and the moisture and tars are scrubbed from the gas before going to the furnace.

The following are data regarding the Kyshtim furnaces : Surface of air checker, 869 sq. m.; useful area, 688 sq. m.; sectional area of air checker, 49 sq. m.; surface of gas checker, 698 sq. m.; useful area, 559 sq. m.; sectional area of gas checker, 46 sq. meters.

Average gas-analysis: CO2, 7.47; CO, 26.22; H, 8.3; CH4, 5.86 ; C2H4, 0.79 per cent.
Analysis of products of combustion in stack: CO2, 12.2; O, 6.3; 1ST, 71.5 per cent.; S02, ?
Pressure at reversing-valve, 3 mm. of water. Suction at stack, 25 millimeters.
Temperature of gas entering regenerator, 68° C. Temperature entering stack, 385°C.
Average temperature in furnace, 1,600° C. Maximum temperature at ports, 2,000° C. Temperature entering checker, 1,300° C.

The charge was mainly flue-dust, but there was some green ore, some siliceous oxidized copper-ore, some converter-cleanings, etc., mixed with it. I cannot give the composition of the charge, but 90 per cent, of it was flue-dust and green fines.

From January to September, the furnace was in operation 166 days, with results as follows:

Amount smelted, tons……………………………………………………………..10,690
Wood consumed in producers, cords………………………………………….3,417
Matte produced, tons……………………………………………………………….3,326
Contents of matte, tons of copper………………………………………………..283
Assay, Cu, percent………………………………………………………………………8.5
Contents of charge, copper, tons………………………………………………..330.8
Average assay of charge, Cu, per cent……………………………………………3.1

The fuel is chiefly interesting for the extremely low fuel-cost, which has been verified by a first-class engineer on the ground ; wood at Kyshtim costs $1.05 per cord delivered, so that the average fuel-cost over the nine months’ period was 33.6 cents per ton of 2,000 lb. of charge. It comes about from the fact that the local Russians in charge of the furnace have had a lot of experience with gas-producers in iron-works, and made a gas of very high caloric value.

An analysis of a sample of Kyshtim flue-dust taken from one month’s production gives: SiO2, 8.8; Fe, 50.9; Cu, 313; S, 9.6 per cent.

A sample from a 6,000-ton pile of flue-dust gives : SiO2, 8.9; Fe, 48.9; Cu, 3.3; S, 10 per cent.

An average sample of raw ore used with flue-dust gave : SiO2, 2.1; Fe, 38.5; CaO, 0.5; S, 47.0; Cu, 3.2 per cent.

The data in regard to the Kyshtim plant have been recently communicated by John H. Allen.

At Cananca, Texas oil has been used; and at Humboldt, fuel-oil from California. At the plant of the Steptoe Co., at McGill, Nev., fuel-oil from California was introduced during 1911, until gratifying results.

A great many different varieties of burners were tried, but a simple home-made burner (arranged as shown in Fig. 2), using air for atomizing the oil, has given the best results.

A record performance at the Steptoe plant, at McGill, is communicated by Superintendent Sorensen: on Dec. 11, 1911, No. 1 furnace smelted 666 tons of total charge on an oil-consumption of five-eighths of a barrel of oil per ton of charge.

At the Washoe plant, in Anaconda, coal is used for fuel, and a record run of one furnace was the smelting of 402.5 tons of charge in 24 hr., with a fuel-ratio of 1 ton of fuel for 6.45 tons of charge.

At Great Falls, Mont., the fuel is producer-gas, but the producers are located too far away from the furnaces to get the best results; new furnaces are contemplated for this plant, in

which the producers will be placed in close proximity to the furnaces and a much greater efficiency will be obtained.

The Canadian Copper Co. is experimenting with pulverized coal as fuel, but the furnaces have not been in operation long enough to give data of any great value. Supt. David H. Browne writes that he is greatly pleased with the performance of the furnaces up to date.

The objections to the use of pulverized coal in reverberatory practice have been the settling of ash and unburnt fuel on the top of the charge and the clogging of the flues with ashes. The Canadian Copper Co. has a rather peculiar condition, in that the charge is basic, and that the addition of silica, in the form of coal-ash, is rather welcome than otherwise, and to avoid the other objection the coal is pulverized extremely fine, and the surplus ash goes out with the gaseous products of combustion.

At nearly all modern plants waste-heat boilers have been installed, and the diagrams, Figs. 4 to 18, show the important variations of these installations.

Silica brick is used almost exclusively in the construction of the lining and roof of the modern reverberatory furnaces for copper-smelting, and it is now possible to obtain silica brick, of excellent quality, all over the United States at reasonable cost.

I desire here to thank the following gentlemen for valuable data supplied : Dr. E. D. Peters, Dr. H. O. Hofman,Dr. L. D. Ricketts, C. B. Lakenan, S. Severin Sorensen, George B. Lee, William H. Howard, A. E. Wheeler, William Wraith, H. N. Thomson, John H. Allen, David H. Browne.