surfaces of the fuel lumps. Also the combustion space must take into account the inevitable excess of air that is required to secure any sufficient volume of air in actual contact with the coal. Stack draft is determined, other considerations aside, from the necessity of drawing this large volume of combustion and excess air through the fuel bed. With pulverized coal, this requirement for stack draft is eliminated and if a burner is such as to require more commingling space than with the

old methods of burning lump coal, it is evidently a most inefficient piece of apparatus. After all is said and done, why should we spend money to pulverize the coal to the ultimate practical commercial degree of fineness, if we propose to use a feeding apparatus that will feed masses of the fine particles into the furnace as a body? If several thousand particles are to be thrown into the furnace as a lump, we may as well save the cost of grinding that lump.

A Confession of Inefficiency

Again the statement that excess air must be used in burning powdered coal, is on its face a confession of inefficiency in the feeding and mixing apparatus. The opinion seems to be

that some excess air is required particularly in boiler applications and more or less in general heating furnaces, in order to avoid a destructive temperature on the brickwork or tubes. In malleable practice, we are concerned only with the brickwork and, referring again to Fig. 7, it will be observed that the center of heat must be at or near the tip of the ignition body, or the center of the flame production. All combustion is in suspension in the body of gas and the high combustion temperatures of burning particles of carbon, hydrogen, etc., are formed in the center of the gas volume. Thus the brickwork in the surrounding walls of the furnace comes in contact with the lowest temperature gases-the outer expanded film of gas in the furnace. This balanced condition of internal high temperature in the body of gas causing uniform expansion in all directions, is readily maintained by reason of the low discharge velocity employed. Otherwise the hot burning material would be thrown against the bridge wall at the very moment of combustion and thus cut it away as has been experienced with high-pressure jet feeders. It is a remarkable fact that the interior walls take on a glaze and show practically no signs of erosion.

It seems hardly necessary to point out that such a method of controlling temperature and flame is far superior to any method of introducing excess air to chill the gases-usually burning in contact with the brick work-in order to save the furnace lining. The older method of injecting a high-speed jet of material into the furnace and against the bridge wall, means that the wall is cut away and the flame curls upward and to the sides, thus cutting out both roof and side walls. The action is mechanical and also, due to the high temperatures of combustion-with carbon perhaps 1000 and with hydrogen 2000 degrees above furnace average temperature-the limit of brick endurance is passed and of course the brick gives out. Evidently this high temperature must be removed from the brick, but is the injection of cold air on the surface of the brick the right method?

A natural assumption would be that the length of combustion chamber and the volume of discharge should be arranged so as to have this combustion occur in contact with the charge

of iron in a melting furnace, thereby securing a rapid cutting down. There is, however, a great difference between burning the material in contact with a hot refractory substance and burning it in contact with the cold iron which it is desired to cut down. The chilling effect of the iron on the flame is such that heating is much retarded and the time of melting would be increased. Furthermore the greatest heat would be at the back end of the charge.

Tests

Having developed this type of apparatus, arrangements were made to conduct a series of tests on an air melting furnace at a plant at Meadville, Pa. Owing to some misunderstanding as to the method to be used in controlling the top blast, the original burner installed was designed to admit half the air at the top blast and take the remainder through the burner. This sacrificed at least half the mixing efficiency of the apparatus. A 14-inch duplex burner was used, capable of taking air enough for mixing and burning 1200 to 1500 pounds of coal per hour and a proportionately greater amount according to the amount of air admitted at the top blast. Pulverized coal was purchased from an outside plant, being that used ordinarily for annealing. It analyzed as follows:

Owing to the fact that the coal was stored in paper bags in an open foundry and samples were taken from the bags, it is believed a considerable amount of the coal contained more moisture than shown, as many of the bags were broken. The coal had been in this storage for some six months before the

test.

Air was supplied by a No. 7 Sturtevant fan at a pressure on the fan outlet of five ounces, the line running some 40 feet to a Y-division supplying the two parts of the burner. The existing branch of the line supplying air to the ash pit was not disturbed. The burner was first lit with the furnace empty, the total air pressure at the burner being 8 inches and at the over

head blast pipe 4.25 ounces (difference due to pipe layout) The furnace was operated on this basis for two hours 12 minutes, the average coal fed being about one ton per hour.

At 1 hour and 20 minutes from starting time, a pig of iron was placed in the rear side door of the furnace and another in the front door. These were dripping freely in eight minutes. When feeding the full amount of coal, combustion was not complete until the flame came in contact with the top blast just over the bridge wall. Thus the principal heat zone was too far in the rear of the furnace. Coal was filled into the hopper by hand from bags of weights averaged, the amount of coal left at end of run being deducted. It was found that the feed screw had been delivering 0.3 pound

FIG. 8-DIAGRAM SHOWING COMPARATIVE ECONOMY OF PULVERIZED COAL BURNERS AND HAND FIRING ON ANNEALING OVEN

per turn. tons.

The furnace was built for a charge of 10 to 12

The following day, a 2-ton charge made up of pig iron, hard scrap and railway malleable, showing an average of 0.914 per cent silicon and 0.627 per cent manganese was put in the furnace, which was then fired, without skimming bath, for 3 hours and 29 minutes, when the iron was poured, the fire being continued for 15 minutes while tapping out. A total of 5914 pounds of coal were used, or say three tons of coal for two tons of iron. The general character of the metal was satisfactory but it was necessary to leave considerable of slag in the furnace, owing to the condition of the bottom and the tap holes.

The brick covering that had been placed over the grates was removed at the bridge wall across the fire box for a

space 1 foot wide lengthwise of the furnace. This made it possible to admit air through the ash pit under the fire in regulated amount. A 4-ton charge showing 0.902 per cent silicon and 0.62 per cent manganese was charged on the following day and on starting the fire it was at once seen that the large volume of air rising at the bridge wall deflected the flame to the roof and formed a cold blanket on the bath, thus making a slow heating furnace. The rear bridge wall had been built up two courses of brick and after running one and a half hours it was decided the fire was choked too much. Twelve minutes were lost in removing some of the brick, after which the fire was continued for a total of 5 hours and 52 minutes, including a short shut down to replace bung. The roof showed distress while the bath showed dull, attributed to the air condition above mentioned. Owing to the short charge it was not possible to skim the heat effectively, although it was partially skimmed after four hours. The total coal consumption was 10,300 pounds or 5.1 tons for four tons of iron melted in a 12-ton furnace. A test bar poured at 4 hours and 40 minutes showed silicon, 0.54 per cent, sulphur 0.102 per cent, phosphorous, 0.131 per cent, and combined carbon 3.50 per cent.

Changes Were Made

The opening over the grates was filled up and the eight 2-inch tuyeres on top blast were changed to 3-inch, the floor of combustion chamber was filled up to within 10 inches of top of front bridge wall. The outer end of the burner was raised and the discharge pointed at a downward angle. against the front end of the charge. The burner was then lighted on the empty furnace for 1 hour and 15 minutes, it being desired to start a charge with a hot furnace. A 5-ton charge of the same composition as the previous heat was then fired for 4 hours and 55 minutes, with shut down of 5 minutes on account of the loss of a bung. The total coal fed was 9088 pounds or an average of 1725 pounds per hour, or a total of 4.5 tons for 5 tons of iron in a 12-ton furnace, with a considerable amount of slag left in the furnace from previous heats that could not be skimmed properly. The