THE TRANSFORMERS.

In the rear of the furnace, and as close thereto as is possible, are the three service transformers which supply three-phase current at 40 to 80 volts to the electrodes. These transformers are oil immersed and water cooled and have a capacity of 750 kilowatts each. The low-tension connections to the electrodes consists of eight pieces of flat copper bar, three-eighths of an inch thick and 5 inches wide, bolted together. On the 2,400-volt primary side there are brought out eight current taps for voltage regulation. These lead to an oil-immersed switch group, each unit of which is operated by a solenoid. This . arrangement, with autotransformer compensator, gives 15 steps for voltage variation.

ELECTRIC CONTROL.

Electric control is through a switchboard, there being a panel for each furnace. As the current and power factor in each phase must be under observation at all times during operation, separate meters are installed in each phase. The requirements for one panel are 3 ammeters, 3 voltmeters, 3 wattmeters, 3 power-factor meters, and 3 recording wattmeters. The meters are mounted across the panel in rows of three each. Under the first four sets named are three handwheels to control the voltage variation, and under these three switches that control the entire load, and under the latter are the recording wattmeters.

For operating the voltage control and the main circuit breakers there is a 7-kilowatt motor-generator set, comprising a 125-volt directcurrent generator, directly connected to a 10-horsepower induction motor. This set has a small panel board mounting a circuit breaker, ammeter, voltmeter, and two single-pole knife switches. In the event that line voltage should fall, or if for any other cause the directcurrent supply should become deranged, there is a storage-battery set. having a capacity of 73 kilowatts, which may be instantly switched in, and thus prevent the furnaces from cutting out in the case of low voltage.

OPERATION.

The operation of the furnace is continuous. After a period of eight hours the hearth contains a full bath of molten metal, and therefore the metal is tapped three times each day. Charging is done at regular intervals, and the current is never shut off during operation. During the period of smelting the change in electrical conditions is interesting. At the beginning of the charge the power factor is almost unity. This gradually lowers as melting continues, until with a full bath of metal a power factor of 65 per cent is reached. If coke is used in place of charcoal, or if a mixture is employed, a dif

ferent set of power-factor conditions exists. By studying these conditions it is possible to know the exact condition of the charge by looking at the meters. The load is, of course, a function of the voltage, and with half voltage the load will drop one-quarter.

In charging, the ore cars are run on the mixing floor, which is level with the charging floor. Charcoal and lime and quartz for flux are brought in on the lower floor in cars, hoisted on an electrically operated elevator to the mixing floor, and are dumped into their respective bins. The mixing is done in a car that is run on platform scales. The charge is placed in layers, the proportions depending on the tests made in the laboratory by the chemist. Following is a representative charge: Five hundred pounds of iron ore (magnetite); 135 to 150 pounds of charcoal; 33 pounds of lime (well burned); 12 pounds of quartz.

USE OF CHARCOAL AS A REDUCING AGENT.

Although, as explained later, the use of the electric furnace in the production of pig iron permits the manufacture of 3 tons of pig iron from 1 ton of coke or charcoal, the cost of a suitable reducing agent is still an important factor, especially in California. Charcoal is used at the plant of the Noble Electric Steel Co., and its economical manufacture in retorts so constructed as to save the distillation products has been almost as much of a problem as the production of the iron itself. Pit charcoal was used at first. A battery of beehive charcoal ovens, each holding 60 cords of wood, was then constructed. Later it was decided to use by-product retorts, and a vertical retort system was accordingly built. The retorts were arranged in two batteries of four each, each retort being a vertical cylinder 6 feet in diameter and 16 feet high. The volatile products were led from the bottom of the retorts to condensers. Altogether, wood distillation and the attendant recovery and working up of the by-products has not seemed to be a profitable enterprise on the Pacific coast. Soft woods are used for the most part, and these are generally lean in acetic acid and wood alcohol, although fairly rich in tarry products. A splendid grade of charcoal was made at the plant of the Noble Electric Steel Co. in the retorts above mentioned, but the system proved to be rather cumbersome, and the time necessary for retorting was excessive. Moreover, owing to the manner in which the retorts were constructed, all the products resulting from distillation of the wood had to pass down through the hottest part of the retort. As a result the volatile products were broken up by heat, undesirable products were thus formed, and an inferior grade of by-products was produced. Consequently the company has decided to replace this system with what is known as the Yost system. The retort of the Yost system consists of a horizontal steel cylinder, about 5 feet

in diameter and about 20 feet long, mounted in brickwork. One end of the retort is closed and the other is provided with doors. The wood is loaded on a car, and the car is run into the retort, which is then closed and sealed. The volatile products resulting from the heating of the wood are conducted to the condensers by means of a copper pipe; the liquid tar formed in like manner is drained from the bottom of the retort into a tank. Owing to its simplicity this type of retort is claimed to be much less expensive than the type formerly used, and because of the fact that the wood will not have to be handled as much as formerly the cost of producing charcoal is claimed to be much less.

USE OF CRUDE PETROLEUM AS A REDUCING AGENT.

As is well known, California is favored with a seemingly abundant supply of petroleum. Due to the scarcity of suitable reducing agents such as are at present used in metallurgical work attempts to use crude oil for this purpose have from time to time been made. It is quite possible that a successful process of this kind will be devised. The Bureau of Mines has been conducting investigations along this line and expects to publish the results in the near future. As soon as a commercially successful process of this nature has been devised, it is quite likely that the electric smelting of iron ores on the Pacific coast will receive an impetus, for the use of crude petroleum will not only permit the introduction of electric smelting in those districts where its use, except at prohibitive prices, is not now possible owing to the lack of suitable reducing agents, but should also cheapen the cost per ton of metal produced as compared with the cost when charcoal is used.

COMPARISON OF THE CALIFORNIA AND SWEDISH FURNACES AND PROCESSES.

As to the construction of the two types of furnaces, the difference is so apparent that no comment is necessary. As to the process, the California practice differs from the Swedish practice very decidedly in the following respects, namely: No attempt is made to reduce the ore in the stacks of the furnace; no artificial circulation of gases is employed; limestone calcined outside of the furnace is used. In other words, the California practice is just the reverse of the Swedish practice.

REDUCTION IN SHAFT.

As has previously been noted, the ideal condition in electric reduction furnace work would seem to be to have the charge in such a condition as to require only melting by the time it comes in proximity to the electrodes. The electric current would then be used to furnish only the additional heat necessary to bring about the melting

of the hot previously reduced oxide. Of course the heat for the reduction must come from some source, and its origin would be in the crucible and as a result of the heat developed during the melting of the charge, but the excess heat so developed can be transferred by the circulation of the gas to the charge in the shaft instead of being carried away by radiation and by circulating water.

CIRCULATION OF GASES.

As has also been mentioned on page 16, in the Swedish practice a part of the gases rising from the top of the shaft is returned to the crucible and made to pass through it and up through the charge in the shaft. This is done to cool the superheated brickwork of the roof of the crucible and to carry through the stack a volume of gas large enough and hot enough to bring about reduction in the stack.

As stated elsewhere, one of the most serious difficulties that had to be overcome in the development of the Swedish type of furnace was the destruction of the crucible resultant on local heating in the vicinity of the electrodes. Therefore, by returning to the crucible a part of the gases resulting from reduction, it is possible to prevent excessive local heating and at the same time to make use of the heat instead of wasting it. Although gas circulation has its objections, its advantages in the way of economy of operation would seem to outweigh its disadvantages. On the other hand, if reduction previous to smelting be considered the proper practice, this reduction will have to be made in the stack, and the present California practice will not find application except when warranted by the local conditions. As has been previously stated, the present demand in California is for a soft gray foundry iron. Although the furnace now used gives that product, such a construction and such a process are, of course, not absolutely necessary for the production of such an iron.

That the California type of furnace is better suited to the production of a soft gray foundry iron than is the Swedish type is probably due to the fact that reduction is performed solely by solid carbon, and the reduction of silica to silicon takes place at the same time as the reduction of iron oxide to metallic iron. The silicon is dissolved in the iron and by its presence causes the carbon to be precipitated as graphitic carbon, and there is thus obtained the soft gray iron desired. It is also quite probable that in a furnace having a rectangular crucible, such as has the California furnace, the temperature of the charge in the crucible as a whole is much higher than in a furnace of the Swedish type, and this higher temperature is more favorable to the reduction of silica to silicon. As before stated, the latter reaction is largely the controlling element in the production of soft gray foundry iron.

16282°-Bull. 67-14- 4

USE OF CALCINED LIMESTONE.

The use of calcined limestone has been repeatedly suggested, and theoretically seems desirable, but in operating a furnace of the Swedish type there are two objections to its use, namely, it increases the proportion of fine material in the charge and it makes the charge hang. As to the former of these objections, the idea seems to be prevalent that there is no limit to the percentage of fine material that may be used in the electric reduction furnace. However, judging from observations of the smelting of black sands when the furnace consisted simply of a crucible and no shaft, some difficulty was experienced on account of the charge being made up entirely of fine material. In the Swedish practice the proportion of fine material that can be used in a charge was definitely determined in the work at Trollhättan, and it was found that a large proportion of fine material was detrimental to smooth running and good results; but Leffler, one of the engineers in charge of the work at Trollhättan, is of the opinion that calcined limestone may be advantageously added to the charge if no reduction is attempted in the shaft.

In view of the above facts, it would seem that the California practice as compared to the Swedish practice presents the following advantages: It permits the use of limestone calcined outside the furnace and it does not require the circulation of gases.

As to the circulation of gases and also as to reduction in the shaft, the California practice might perhaps prove more efficient if both were done. Although a very complete record of the working of the furnace at Trollhättan has been published, no such record of the operation of the furnace at Heroult is available, and so it is not at present possible to make a definite comparison of the two practices.

ELECTRIC FURNACE AS COMPARED TO THE BLAST

FURNACE.

As has been mentioned in the section pertaining to the history and evolution of the electric pig-iron furnace, the electric furnace was not developed as a competitor of the blast furnace, but for the purpose of finding a furnace and a process that would be able to produce iron in those localities where blast-furnace practice was not feasible, or, as in Sweden, where the increasing cost of suitable fuel was becoming prohibitive to the existing practice of smelting in blast furnaces.

Broadly speaking, it may be said that the feasibility of smelting iron ores in an electric furnace depends upon the relative cost of either charcoal or coke and of electric power. As regards the latter, it must be cheap. As is well known, electric power at the present time can be developed more cheaply than it could when Capt. Stassano made