calories computed by Yngstrom as being necessary to produce 1 ton of pig iron with the number that can be calculated from JungeHermsdorf's "Heat Balance of a Blast Furnace of 250-ton Capacity." A furnace producing 250 tons in 24 hours will produce, in round numbers, 10 tons an hour, and as 10 tons of coke is needed per hour, this would practically correspond to 1 ton of coke for each ton of iron produced.

One ton of coke is equivalent to 1,000 kg., and as 1.18 kg. of coke is equivalent to 1 kg. of pure carbon, 1 ton of coke is equivalent to practically 850 kg. of pure carbon. If 4,153 kilogram-calories, as estimated by Yngstrom, be taken as the number of calories that are given off when 1 kg. of carbon combines with oxygen, forming a mixture that contains 30 per cent (by volume) of CO2 (0.3×8,080+0.7×2,470= 4,153), the heat necessary for producing 1 ton of pig iron in the blast furnace is 850X4,153 = 3,530,050 calories. As 2,121,284 calories represents the amount of heat that is necessary to produce 1 ton of pig in the electric furnace, there is a difference of 1,408,766 calories between the heat necessary for producing 1 ton of pig iron in the blast furnace and that required for producing 1 ton of pig iron in the electric furnace.

PROBLEMS IN THE ELECTRIC SMELTING OF IRON ORES.

Having thus traced the history and evolution of the electric pigiron furnace up to the present time, and having stated the fundamental chemical principles upon which the reduction of iron is based, the authors will now briefly consider some of the difficulties that had to be overcome in order to bring the Swedish type of furnace up to its present stage of development and some of the problems that yet remain to be solved.

EARLY DIFFICULTIES.

An inspection of Harmet's drawings, and of the drawings of practically all others who gave their attention to the development of an electric furnace for the reduction of iron ores, shows that the first idea was to construct a shaft similar to a blast-furnace shaft, and then to substitute electrodes for tuyères. This seemed feasible, but in practice the furnace wall in the neighborhood of the electrodes proved to be short lived. Water cooling did not obviate this difficulty, but rather tended to increase it, due to jackets burning out and to other reasons. As was early observed by Héroult, the proper way to maintain the walls of an electric furnace crucible is to remove the electrodes as far as possible from the side walls. In the development of the furnace in California and in Sweden it was found necessary not only to do this, but also to keep the charge as far as possible

from the roof of the crucible, as otherwise the roof was short lived owing to the intense local heat generated at the point where the electrodes enter the charge. In overcoming this most serious difficulty the present shape of the roof of the crucible and the manner in which the electrodes are introduced into the crucible have been evolved. As a further protection to the roof of the crucible, a part of the gases that were escaping from the top of the shaft, as previously stated, was returned to the crucible in order to cool the walls, as well as to assist in the reduction of the charge.

ELECTRODE PROBLEM.

Although the Swedish experimenters did not have so much trouble with electrodes, in California the difficulty of procuring suitable electrodes greatly interfered with the progress of the work. When experiments were begun in California the manufacturers of electrodes in this country had not previously been required to furnish electrodes of such large size, namely, about 20 inches square and about 72 inches long. Of those first furnished, that part of the electrode projecting into the crucible would either break off completely after it became heated, or else would spall off in large chunks and give trouble in operating the furnace. That others also encountered this difficulty is evident from the following quotation from a paper presented by W. R. Walker at the April, 1912, meeting of the American Iron and Steel Institute:

a

Our problems-mechanical, metallurgical, and otherwise-proved many, and our experience soon demonstrated that the conditions surrounding the successful operation of a large electric furnace were in many respects entirely different from those involved in the use of smaller units. In illustration the demands of a 15-ton electric furnace proved to be far in advance of the art of manufacturing electrodes. Our necessities represented a requirement that the electrode manufacturers of America and Europe had not been called upon to meet, and it took much time and money before there was finally accomplished the 20-inch round amorphous-carbon electrode that is now being used at South Chicago.

At the plant of the Noble Electric Steel Co. it was finally decided to try graphite electrodes. These worked satisfactorily in all respects except that on account of the angle at which it was necessary to insert them in the crucible they were subjected to a severe strain which caused them to break at the threaded joints. The electrode problem is, however, no longer a serious matter. As stated by Walker, the 20-inch round electrodes now in use at South Chicago give satisfaction, and the engineers at Trollhättan also report that the large carbon electrodes of about the same size as those used at South Chicago meet their requirements.

a Walker, W. R., Electric furnace as a possible means of producing an improved quality of steel; Metall, Chem. Eng., vol. 10, 1912, p. 371.

UNSOLVED PROBLEMS.

In connection with the unsolved problems in the electric smelting of iron ores it may be well to take up first a discussion of the results obtained at Trollhättan and in California.

A most careful record was kept of the work done at Trollhättan during the period November 15, 1910, to April, 1911, and the data thus obtained were incorporated in a report submitted to JernKontoret by two of its engineers-J. A. Leffler and E. Odelberg. The essential parts of the report have been translated by Dellwik." This report has been ably discussed by Robertson in a paper that he presented before the Toronto meeting of the American Electrochemical Society, September, 1911. The work at Trollhättan is also the subject of a paper by Frick. In this paper Frick also discusses the results obtained in California. Some of the points brought out in his paper are discussed below.

VOLUME OF SHAFT AS COMPARED WITH CAPACITY OF FURNACE.

Theoretically, the volume of the shaft of an electric furnace should be great enough to permit the practically complete reduction of the ore before it enters the crucible. In other words, the ideal condition in electric reduction furnace work would be to have the charge in such a condition as to require only melting by the time it comes into proximity with the electrodes. So far the degree of reduction that has taken place in the shaft of the electric furnace has varied all the way from nothing up to a considerable proportion less than complete reduction. Granted that the size of the shaft at Trollhättan has been properly calculated, and that the ratio of the volume of charge in 24 hours to the volume of the furnace should be 1:55, then in order to insure the best economical working conditions of the furnace, the reduction of the charge in the shaft must take place in the proper

manner.

REDUCTION OF ORE IN SHAFT.

In an ordinary blast furnace the weight of the gases produced exceeds the weight of the charge by 30 to 50 per cent, whereas in electric-furnace reduction the gases evolved amount to only about 40 per cent, by weight, of the charge. In other words, in the blast furnace three to four times as much gas is given off in the production of 1 ton of iron as is given off in the production of 1 ton of iron in the electric furnace. Moreover, the temperature of the gas as it leaves the vicinity of the tuyères may be as high as 1,600° C., whereas the

a Dellwik, C., Electric iron smelting at Trollhättan, Sweden: Iron and Coal Trades Rev., vol. 82, June 9, 1911, pp. 957-962.

Robertson, T. D., Recent progress in electrical iron smelting in Sweden: Trans. Am. Electrochem. Soc., vol. 20, 1911, p. 375.

c Frick, Otto, Electric reduction of iron ores with special reference to results obtained in Electro-Metals furnace at Trollhättan, Sweden, and Noble furnace at Heroult, Cal.: Metall. Chem. Eng., vol. 9, Dec., 1911, p. 631.

16282°-Bull. 67-14- -3

highest temperature stated to have been attained at the lower end of the stack in the work at Trollhättan was 965°. Granting that the temperature of the gas as it enters the stack may be even 1,000°, inasmuch as the weight of the gases produced is only about 40 per cent of the weight of the charge, the charge would not be heated to more than about 350° C., as pointed out by Frick, owing to the fact that the specific heat of the gas and that of the charge are about the same. For this reason complete reduction in the shaft can not be accomplished solely by the heat from the gases generated in the regular manner. In other words, the shaft acts merely as a preheater, most of the reduction taking place in the crucible by means of solid carbon. Thus it is seen that an electric furnace may be operated in one of two ways: The ore may be reduced in the crucible by means of solid carbon, with no attempt at reduction in the stack, as is done in Cal

FIGURE 14.-Curves showing variation of fuel required and of heat evolved.

ifornia at the present time by the Noble Electric Steel Co., or there may be part or complete reduction in the stack, this being the principle on which the Swedish furnaces are operated.

As is well known, a method that may seem best theoretically is not always best in practice. In this instance it is not possible to judge by results, as sufficient data are not available, and so the matter can be viewed only theoretically. In the reduction of any given iron ore, a definite amount of oxygen must be removed for every ton of iron obtained. Therefore it would theoretically be best to remove all the oxygen in the form of CO. However, as the oxidizing action of CO2 gas varies with different temperatures, such removal is not possible, as experiments show that the ratio of CO2 to CO should not be greater than 1:1, whereas 1:2 is the determined ratio in which they are usually present in blast-furnace gases. The curves in figure 14,

the work of Prof. Richards, show that in electric-furnace work the greater the proportion of CO2 in the gas the less is the amount of carbon required to remove the oxygen from the ore. Figure 15 (also prepared by Richards) shows that the consumption of electrical energy is proportional to the production of CO. Therefore, judging from deductions made from the curves, the efficiency of the electric furnace will be increased as reduction is effected by the gases in the shaft, and the most desirable ratio of CO2 to CO in the escaping gases is 1:1. If this be granted, it would seem that reduction in the shaft is necessary, and the problem becomes how to effect the most satisfactory reduction in the shaft.

2

FIGURE 15.-Curves showing variation of power required according to the composition of the waste gases.

The essential factors in accomplishing reduction in the shaft are temperature, time, and the proportion of CO in the gas.

As has just been pointed out, the volume of the gas generated by the reduction of the oxides is not sufficient to maintain a reduction temperature in the stack. As a remedy for this difficulty the present system of gas circulation was adopted. The system is briefly discussed below.

GAS CIRCULATION.

As has been pointed out by Frick," the idea of using circulating gas in the electric furnace was conceived by Harmet and incorporated by him in his papers on the reduction of iron ores in the electric

furnace.

a Frick, Otto, Electric reduction of iron ores, with special reference to results obtained in Electro-Metals furnace at Trollhättan, Sweden, and Noble furnace at Heroult, Cal. : Metall. Chem. Eng., vol. 9, 1911, p. 631.

Report of the commission appointed to investigate the different electrothermic processes for the smelting of iron ores and the making of steel in Europe: Mines Branch, Department of the Interior, Canada, 1904, pp. 124, 139.