under or in front of the benches, with every chance for a good circulation of air. "Header" coils are better than "return-bend" coils for this purpose. Mr. Baldwin's rule may be given the following form: Let H = heat-units transferred per hour, T temperature inside the greenhouse, t = - temperature outside, S= sq. ft. of glass surface; then H = 1.5S(T — t) × 60 ÷ 48 = 1.875S(T1). Mr. Wolff's coefficient K for single skylights would give H = 1.118S(T − t).

=

Heating a Greenhouse by Hot Water.-W. M. Mackay, of the Richardson & Boynton Co., in a lecture before the Master Plumbers' Association, N. Y., 1889, says: I find that while greenhouses were formerly heated by 4-inch and 3-inch cast-iron pipe, on account of the large body of water which they contained, and the supposition that they gave better satisfaction and a more even temperature, florists of long experience who have tried 4-inch and 3-inch cast-iron pipe, and also 2-inch wrought-iron pipe for a number of years in heating their greenhouses by hot water, and who have also tried steam-heat, tell me that they get better satisfaction, greater economy, and are able to maintain a more even temperature with 2inch wrought-iron pipe and hot water than by any other system they have used. They attribute this result principally to the fact that this size pipe contains less water and on this account the heat can be raised and lowered quicker than by any other arrangement of pipes, and a more uniform temperature maintained than by steam or any other system.

HOT-WATER HEATING.
(Nason Mfg. Co.)

There are two distinct forms or modifications of hot-water apparatus, depending upon the temperature of the water.

In the first or open-tank system the water is never above 212° temperature, and rarely above 200°. This method always gives satisfaction where the surface is sufficiently liberal, but in making it so its cost is considerably greater than that for a steam-heating apparatus.

In the second method, sometimes called (erroneously) high-pressure hotwater heating, or the closed-system apparatus, the tank is closed. If it is provided with a safety-valve set at 10 lbs. it is practically as safe as the opentank system.

Law of Velocity of Flow.-The motive power of the circulation in a hot-water apparatus is the difference between the specific gravities of the water in the ascending and the descending pipes. This effective pressure is very smail, and is equal to about one grain for each foot in height for each degree difference between the pipes; thus, with a height of 12" in "up" pipe, and a difference between the temperatures of the up and down pipes of 8°, the difference in their specific gravities is equal to 8.16 grains on each square inch of the section of return-pipe, and the velocity of the circulation is proportioned to these differences in temperature and height.

=

To Calculate Velocity of Flow.-Thus, with a height of ascending pipe equal to 10' and a difference in temperatures of the flow and return pipes of 8°, the difference in their specific gravities will equal 81.6 grains, or 7000.01166 lbs., or X 2.31 (feet of water in one pound): .0269 ft., and by the law of falling bodies the velocity will be equal to 8.0269 = 1.312 ft. per second, or X 60=78.7 ft. per minute. In this calculation the effect of friction is entirely omitted. Considerable deduction must be made on this account. Even in apparatus where length of pipe is not great, and with pipes of larger areas and with few bends or angles, a large deduction for friction must be made from the theoretical velocity, while in large and complex apparatus with small head, the velocity is so much reduced by friction that sometimes as much as from 50% to 90% must be deducted to obtain the true rate of circulation.

Main flow-pipes from the heater, from which branches may be taken, are to be preferred to the practice of taking off nearly as many pipes from the heater as there are radiators to supply.

It is not necessary that the main flow and return pipes should equal in capacity that of all their branches. The hottest water will seek the highest level, while gravity will cause an even distribution of the heated water if the surface is properly proportioned.

It is good practice to reduce the size of the vertical mains as they ascend, say at the rate of one size for each floor.

As with steam, so with hot water. the pipes must be unconfined to allow

for expansion of the pipes consequent on having their temperatures in creased.

An expansion tank is required to keep the apparatus filled with water, which latter expands 1/24 of its bulk on being heated from 40° to 212°, and the cistern must have capacity to hold certainly this increased bulk. It is recommended that the supply cistern be placed on level with or above the highest pipes of the apparatus, in order to receive the air which collects in the mains and radiators, and capable of holding at least 1/20 of the water in the entire apparatus.

Approximate Proportions of Radiating-surfaces to Cubic Capacities of Space to be Heated.

One Square Foot of Ra- In Dwellings, In Halls, Stores,

diating-surface will heat with

School-rooms,

Offices, etc.

Lofts, Facto-
ries, etc.

In Churches, Large Auditoriums, etc.

Diameter of Main and Branch Pipes and square feet of coil surface they will supply, in a low-pressure hot-water apparatus (212°) for direct or indirect radiation, when coils are at different altitudes for direct radiation or in the lower story for indirect radiation:

Direct Radiation. Height of Coil above Bottom of Boiler,

in feet.

0

sq. ft. sq. ft. sq. ft. sq. ft. sq. ft. sq. ft. sq. ft. sq. ft. sq. ft. sq.

57

65

68

98

1099 1132 1166 1202 1241 1283 1436 1478 1520 1571 1621 1654 1817 1871 1927 1988 2052 2120 2185 2244 2309 2376 2454 2531 2574 3140 3228 3341 3424 3552 3648 3763 4276 4396 4528 4664 4808 4964 5132

1327 1374 1425

1480

1733 1795 1861

1933

2193 2272
2713

2350

2415

2805 2907

3019

8

11

13

16

5580 5744 5912 6080 6284 6484 6616 6932 7180 7444 7068 7268 7484 7708 7952 8208 8482 8774 9088 9424 10 8740 8976 9236 9516 9816 10124 10296 10852 11220 11628 10559 10860 11180 11519 11879 12262 12666 13108 13576 14078 12 12560 1291213364 13696 14208 14592 15052 15588 16144 16736 14748 15169 15615 16090 1659117126 17697 18307 18961 19633 14 17104 175841810918656 19232 19856 20528 21232 21984 22800 15 19634 20195 20789 21419 22089 22801 23561 24373 25244 26179 22320 22978 23643 24320 25136 25936 26464 27728 28720 29776

7735

9780

12076

14620

17376

20420 23680

27168 30928

The best forms of hot-water-heating boilers are proportioned about as follows:

1

boiler-
grate-

66

1 sq. ft. of grate-surface to about 40 sq. ft. of boiler-surface.

5

radiating-surface.

66

Rules for Hot-water Heating.-J. L. Saunders (Heating and Ventilation. Dec 15, 1894) gives the following: Allow 1 sq. ft. of radiating surface for every 3 ft. of glass surface, and 1 sq. ft. for every 30 sq. ft. of wall surface, also 1 sq. ft. for the following numbers of cubic feet of space in the several cases mentioned.

In dwelling-houses: Libraries and dining-rooms, first floor.. 35 to 40 cu. ft.
Reception halls, first floor....
Stair halls,

Chambers above, "

40 to

50"

66

40 to

50 to

To find the necessary amount of indirect radiation required to heat a room: Find the required amount of direct radiation according to the foregoing method and add 50%. This if wrought-iron pipe coil surface is used; if castiron pin indirect-stack surface is used it is advisable to add from 70% to 80%. Sizes of hot-air flues, cold-air ducts, and registers for indirect work.Hot-air flues, first floor: Make the net internal area of the flue equal to 34 sq. in. to every square foot of radiating surface in the indirect stack. Hotair flues, second floor: Make the net internal area of the flue equal to 5% sq. in. to every square foot of radiating surface in the indirect stack.

Cold-air ducts, first floor: Make the net internal area of the duct equal to 5 sq. in to every square foot of radiating surface in the indirect stack. Cold air ducts, second floor: Make the net internal area of the duct equal to 1⁄2 sq. in. to every square foot of radiating surface in the indirect stack. Hot-air registers should have their net area equal in full to the area of the hot-air flues. Multiply the length by the width of the register in inches; of the product is the net area of register.

Arrangement of Mains for Hot-water Heating. (W. M. Mackay, Lecture before Master Plumbers' Assoc., N. Y., 1889)-There are two different systems of mains in general use, either of which, if properly placed, will give good satisfaction. One is the taking of a single large-flow main from the heater to supply all the radiators on the several floors, with a corresponding return main of the same size. The other is the taking of a number of 2-inch wrought-iron mains from the heater, with the same number of return mains of the same s ze, branching off to the severai radiators or coils with 14-inch or 1-inch pipe, according to the size of the radiator or coil. A 2 inch main will supply three 14-inch or four 1 inch branches, and these branches should be taken from the top of the horizontal main with a nipple and elbow, except in special cases where it is found necessary to retard the flow of water to the near radiator, for the purpose of assisting the circulation in the far radiator; in this case the branch is taken from the side of the horizontal main. The flow and return mains are usually run side by side, suspended from the basement ceiling, and should have a gradual ascent from the heater to the radiators of at least 1 inch in 10 feet. It is customary, and an advantage where 2-inch mains are used, to reduce the size of the main at every point where a branch is taken off.

The single or large main system is best adapted for large buildings; but there is a limit as to size of main which it is not wise to go beyond-gener ally 6-inch, except in special cases.

The proper area of cold-air pipe necessary for 100 square feet of indirect radiation in hot-water heating is 75 square inches, while the hot air pipe should have at least 100 square inches of area. There should be a damper in the cold-air pipe for the purpose of controlling the amount of air admitted to the radiator, depending on the severity of the weather.

THE BLOWER SYSTEM OF HEATING AND
VENTILATING.

The system provides for the use of a fan or blower which takes its supply of fresh air from the outside of the building to be heated, forces it over steam coils, located either centrally or divided up into a number of independent groups, and then into the several ducts or flues leading to the various rooms. The movement of the warmed air is positive, and the delivery of the air to the various points of supply is certain and entirely independent of atmospheric conditions. For engines, fans, and steam-coils used with the blower system, see page 519.

Experiments with Radiators of 60 sq. ft. of Surface. (Mech. News, Dec., 1893.)-After having determined the volume and temperature of the warm air passing through the flues and radiators from natural causes, a fan was applied to each flue, forcing in air, and new sets of measurements were made. The results showed that more than two and onethird times as much air was warmed with the fans in use, and the falling off in the temperature of this greatly increased air-volume was only about 12.6%. The condensation of steam in the radiators with the forced-air circulation also was only 66% greater than with natural-air draught. One of the several sets of test figures obtained is as follows:

Cubic feet of air per minute.

Condensation of steam per minute in ounces
Steam pressure in radiator, pounds...
Temperature of air after leaving radiator.

Natural ForcedDraught

air

in Flue. Circulation.

66 "before passing through radiator. Amount of radiating surface in square feet... Size of flue in both cases

There was probably an error in the determination of the volume of air in these tests, as appears from the following calculation. (W. K.) Assume that 1 lb. of steam in condensing from 9 lbs. pressure and cooling to the temperature at which the water may have been discharged from the radiator gave up 1000 heat-units, or 62.5 h. u. per ounce; that the air weighed .076 lb. per cubic foot, and that its specific heat is .238. We have

Natural Forced Draught. Draught. = 731 1225 H.U. 1399**

Heat given up by steam, ounces x 62.5... Heat received by air, cu. ft. x .076 x diff. of tem. x .238 = 673 Or, in the case of forced draught the air received 14% more heat than the steam gave out, which is impossible. Taking the heat given up by the steam as the correct measure of the work done by the radiator, the temperature of the steam at 237°, and the average temperature of the air in the case of natural draught at 102° and in the other case at 93°, we have for the temperature difference in the two cases 135° and 144° respectively; dividing these into the heat-units we find that each square foot of radiating surface transmitted 5.4 heat-units per hour per degree of difference of temperature, in the case of natural draught, and 8.5 heat-units in the case of forced draught (= 8.5 x 1440 = 1224 heat-units per square foot of surface).

In the Women's Homoeopathic Hospital in Philadelphia, 2000 feet of one-inch pipe heats 250,000 cubic feet of space, ventilating as well; this equals one square foot of pipe surface for about 350 cubic feet of space, or Jess than 3 square feet for 1000 cubic feet. The fan is located in a separate building about 100 feet from the hospital, and the air, after being heated to about 135°, is conveyed through an underground brick duct with a loss of only five or six degrees in cold weather. (H. I. Snell, Trans. A. S. M. E .ix. 106. Heating a Building to 70° F. Inside when the Outside Temperature is Zero.-It is customary in some contracts for heating to guarantee that the apparatus will heat the interior of the building to 70° in zero weather. As it may not be practicable to obtain zero weather for the purpose of a test, it may be difficult to prove the performance of the guarantee. E. E. Macgovern, in Engineering Record, Feb. 3, 1894, gives a calculation tending to show that a test may be made in weather of a higher temperature than zero, if the heat of the interior is raised above 70°. The higher the temperature of the rooms the lower is the efficiency of the radieting-surface, since the efficiency depends upon the difference between the

temperature inside of the radiator and the temperature of the room. He concludes that a heating apparatus sufficient to heat a given building to 70 in zero weather with a given pressure of steam will be found to heat the same building, steam-pressure constant, to 110° at 60°, 95° at 50°, 82° at 40°, and 74° at 32°, outside temperature. The accuracy of these figures, however has not been tested by experiment.

The following solution of the question is proposed by the author. It gives results quite different from those of Mr. Macgovern, but, like them, lacks experimental confirmation.

Let S sq. ft. of surface of the steam or hot-water radiator;

W
Ts

sq. ft. of surface of exposed walls, windows, etc.;

temp. of the steam or hot water, T1 = temp. of inside of building or room, To = temp. of outside of building or room;

a = heat-units transmitted per sq. ft. of surface of radiator per hour per degree of difference of temperature;

b = average heat-units transmitted per sq. ft. of walls per hour, per degree of difference of temperature, including allowance for ventilation.

It is assumed that within the range of temperatures considered Newton's law of cooling holds good, viz., that it is proportional to the difference of temperature between the two sides of the radiating-surface.

Heating by Electricity.-If the electric currents are generated by a dynamo driven by a steam-engine, electric heating will prove very expensive, since the steam-engine wastes in the exhaust-steam and by radiation about 90% of the heat-units supplied to it. In direct steam-heating, with a good boiler and properly covered supply-pipes, we can utilize about 60% of the total heat value of the fuel. One pound of coal, with a heating value of 13,000 heat-units, would supply to the radiators about 13,000 × 60 7800 heat-units. In electric heating, suppose we have a first-class condensingengine developing 1 H.P. for every 2 lbs. of coal burned per hour. This would be equivalent to 1,980,000 ft.-lbs. 778 2545 heat-units, or 1272 heat-units for 1 lb. of coal. The friction of the engine and of the dynamo and the loss by electric leakage, and by heat radiation from the conducting wires, might reduce the heat-units delivered as electric current to the electric radiator, and these converted into heat to 50% of this, or only 636 heatunits, or less than one twelfth of that delivered to the steam-radiators in direct steam-heating. Electric heating, therefore, will prove uneconomical unless the electric current is derived from water or wind power, which would otherwise be wasted. (See Electrical Engineering.)