Coke in a Residence Heater Designed For Coal By DONALD B. PRENTICE, M.E. The high price of anthracite coal has led many householders in the Eastern cities to seriously consider the feasibility of substituting some other less expensive fuel for warming their homes. Gas companies, naturally, have taken advantage of the situation to urge the merits and use of coke, their most important byproduct. At the present time and also under normal conditions coke can be sold at a price 20% to 25% lower than stove or chestnut sizes of anthracite coal, even allowing for the lower heat value per pound of the coke. This is, of course, largely due to the fact that coke is produced from bituminous coal. in the process of gas manufacturing and bituminous is invariably cheaper than anthracite. Although impressed by the material saving in cost, many amateur firemen are skeptical of the satisfactory behavior of coke in ordinary furnaces and feel that they could not maintain comfortable conditions in their homes. They know that coke burns rapidly, unless the draft is very light; and they believe that it would be impossible to produce reasonably uniform warmth in a residence without more frequent attention than they can give. The 30-hour continuous test reported in detail below was made to study the peculiarities and suitability of coke as a fuel in a small standard-type boiler designed for anthracite coal. Broken coke, furnished by the New Haven Gas Light Co., was used. This is about equivalent in size to a mixture of stove and nut anthracite. The test was made on a cast-iron sectional boiler which forms part of the house heating boiler test plant at the Mason Laboratory of Mechanical Engineering of the Sheffield Scientific School, Yale University. VARIATION IN WEIGHT OF COKE. Broken coke is sold by the New Haven Gas Light Company, at retail, by the bushel, fifty bushels being considered the equivalent of a ton. As coke ab sorbs moisture freely selling by weight would be unsatisfactory and might lead to misunderstandings with customers. The bulk of a ton will vary 10% or more, depending on the moisture content. In this particular test the coke averaged 47.84 lbs. per bushel, and the analysis gave 8.35% moisture and a heat value of 12,670 B.T.U. per pound dry. Fifty bushels, therefore, contained about 2,200 lbs. of dry coke. If anthracite coal, has a heat value of 13,000 B.T.U. per pound as purchased, fifty bushels of coke are equivalent in potential heat value to 2,140 lbs. of coal. As a matter of interest and to furnish a basis for comparison a test of this same boiler with anthracite stove coal has been reported in adjoining columns to the coke test. RESULTS OF COKE TEST. The results of the coke test may be summarized briefly: 1. The overall efficiency of the boiler with coke is slightly higher than with coal; which is partly due to the smaller size of the coke, partly due to its burning more completely, and partly due to the lower output at which the boiler was operating. The coke requires more frequent attention to check holes developing in the fire bed. These holes are not large, but the air leaking through them would lower the efficiency somewhat if they were allowed to remain. 2. The coke charges in this test were about all the boiler could accommodate without losing steam pressure for too long a period to be satisfactory. With a firing of 175 lbs. pressure was lost for 20 minutes, on the average; when 225 lbs. were fired no pressure was shown on the gage for 40 minutes. With stove coal 250 lbs. can be fired at one time with no longer loss of pressure than for 175 lbs. of coke. Results are secured, of course, when operating at an output equal to at least 75% of rating. 3. From a financial standpoint heating with coke is about 20% less expen sive than with stove anthracite, and corresponds closely to the cost with pea anthracite. This latter fuel cannot be burned in many residence boilers, however, on account of insufficient draft. At the present prices, $7.75 for 50 bushels of coke and $9.75 for a ton of stove anthracite, the costs of fuel for serving 1,000 sq. ft. of radiation one hour would be $0.1045 and $0.145, respectively, which increases the relative saving of coke. 4. A coke fire in a coal furnace would not satisfactorily warm a house unless it received attention morning, noon, and night in ordinary winter weather, and every three hours in extreme weather. This statement is based on the supposition that the furnace is suited to the load it has to carry; i.e., is not very much oversize. The coke would require nearly twice the bin capacity and would be somewhat dustier in the cellar than coal. Offsetting these disadvantages is a sav Data and Results of Evaporative Test of a House-Heating Boiler. 1. Test of cast-iron, sectional, heating boiler located at Mason Laboratory to determine characteristics of coke fuel, conducted by D. B. Prentice. 2. Kind of furnace: hand-fired, rocking grate. Depth, bottom of fire door to grate, 10 in. Mean height, grate to bottom of section above, 18 in. Fuel capacity at one charge on kindling fire, 225 lbs. coke. 3. Grate surface: width, 22.5 in.; length, 43 in.; Area, 6.75 sq. ft. 4. Water-heating surface: Direct, 24.13 sq. ft.; indirect, 78.1 sq. ft. total, 102.23 sq. ft. 5. Superheating surface, sq. ft. 17.4. 6. Date. Coke test, Dec. 23-24, 1914. Coal test, April, 22-23, 1913. 7. Duration. Coke test, 29.83 hrs. Coal test, 33.67 hrs. 8. Kind and size of fuel, Broken gas coke; anthracite stove coal. AVERAGE PRESSURES, TEMPERATURES, ETC. Coke Coal a. The ash from ashpit was less than by analysis in the coke test probably because the fire was not shaken down to the starting thickness when the test was closed. b. Approximately half of the coal test was made firing charges of 150 lbs. each, with charges of 200 lbs. each. c. There was some tendency for the coke to clinker, but the clinkers were easily broken when a new charge was fired and they did not seem to affect the fire seriously. d. The ash from the coke was very light and blew off from the shovel and out of the ash-barrel easily. A Simple Method of Figuring the Economy of Humidity Under all ordinary conditions where ventilation is carried on to any extent the drop in temperature allowable by artificial humidifying is not enough to conserve sufficient heat on the radiation end to evaporate the required amount of water necessary for the humidifying process. This is particularly true in any case where artificial ventilation or large air leakage is present. To make a statement that humidifying economizes the cost of heating is erroneous in all but a few very exceptional cases. In comparing costs for a general statement on any matter it must be remembered that average conditions should be considered and average results compared. Now for residence heating it is generally agreed that without humidifying at all the inside. air will have a relative humidity (at about 35° F.-average outside winter temperature) of about 40%. The writer has also verified this by actual experiment. Now with 35° outside temperature, 40% inside relative humidity and, say 70° inside temperature we have average conditions attained in residences during winter weather. It is not believed that many would advocate raising the inside relative humidity to over 60%. If this is taken as a desirable limit the increase in humidity at 70° could not exceed 60% minus 40% or 20%. 8 grains x 20% = 1.6 grains per cubic foot to be added to the air. According to the "Comfort Zone" chart of Dr. Vernon Hill 70° F. and 40% relative humidity comes just a trifle above the center line of the "zone of comfort" as defined by thou sands of experiments. Therefore, at With 70° inside and 35° outside, With an average outside temperature during the winter of 35° the average hourly efficiency of 1 sq. ft. of radiator surface would be 250 x 35/70 or 125 B.T.U. This is accomplished by turning the steam in the radiator on and off during mild weather, making a maximum efficiency when "on" of 250 B.T.U. per square foot per hour and an efficiency of zero when "off," the average efficiency being 125 B.T.U., as above stated. Then a saving of 14 1/3% under average conditions would mean a saving of 125 x 0.14 1/3, or 18 B.T.U. per square foot of radiator surface. As the living rooms of a residence seldom have an air change of more than once per hour it can easily be computed how many cubic feet of air in the room can be humidified by this saving from 40% to 60% relative humidity. Each cubic foot under such conditions, as already shown, requires the addition and evaporization of 1.6 grains of water or 1.6/7000 pounds of water which, to evaporate, requires about 1000 B.T.U. per pound, or 1.6/7000 x 1000 = 1.6/7 B.T.U. per cubic foot. Since we have 18 B.T.U. for every square foot of radiation each square foot will take care of 18 1.6/7 18 x 7/1.6 = 80 cu. ft. approximately of contents. From this it can be seen that if the ratio of radiator surface to the cubic contents is about 1 to 80 the saving would about equal the loss at one air change per hour, but if the cubic contents exceed this or if the ventilation carried on in cubic feet per hour exceeds this one air change there will be a loss; if, on the other hand the cubic contents is less than 80 cu. ft. to each square foot of radiator surface then there will be a saving. Such ratios, however, are rare and would only apply to small unventilated rooms with comparatively large heat losses. Typical Details For Hot Water, Steam and Hot Blast Heating By T. W. REYNOLDS While the general run of engineering practice follows well-known grooves, there are nevertheless many individual schemes or "kinks" that have been found to work out successfully in practice. Some of these are presented herewith, taken at random from many years of personal experience. Fig. 1 is a detail of the connections as made for an installation of two expansion tanks. The tanks in this particular case are 30 in. in diameter by 54 in. long and are equipped with a water gauge at one end and a manhole at the other. The tanks are connected by a 4-in. equalizing pipe from the center of which is extended a 3-in. vent to the atmosphere and a 3-in. overflow to the nearest rain-water conductor. It is best, however, to extend the overflow to a point over a slop sink so that it may drip openly and disclose any leakage into the heating system from the automatic water feeder. A 1-in. circulating pipe should be connected to the bottom of the tank and used also as an air vent from the risers. The cold water supply is automatically fed as required through the automatic device which is shown equipped with an equalizing pipe and by-pass, all of 1 in. pipe. The expansion pipe is 22-in. in size. In Fig. 2 is shown a good design for a trench curb and covers for use in piping work. Two types of cover are shown. Those of wrought iron should be of the checker pattern. These covers are of 4-in. wrought-iron plate, either with turned-down edges or reinforced with 1-in. x 1-in. x 8-in. angle irons, as shown. The latter cover should have rivets with flush heads. The 2-in. x 3"vent to atmosphere Expansion Tanks I circulating pipe FIG. I. TYPICAL EXPANSION TANK CONNECTIONS. |