Steam Heating and Ventilation/Chapter III


Chapter III. Steam-heating Apparatus.

Boilers.—The questions of boiler design, construction, setting, etc., involve so many considerations requiring careful scientific study, that the entire problem is omitted from this work and the reader is referred to the valuable treatises devoted exclusively to the subject. There is, however, one kind of boiler which should be given some special mention in this place. This is the self-contained cast-iron sectional boiler which is used only for small, low-pressure gravity-heating systems in residences. These boilers are built up in sections of various shapes bolted together, the joints being kept tight merely by the pressure caused by the bolts. On account of their material, as well as the method of construction, they are not adapted for anything but very low pressures. For the larger plants in which cast-iron boilers are used, it is better to install two small ones than to attempt to use one large one, as it gives better economy in operation, and there is less danger of accident to small boilers than to those with large castings.

In selecting these boilers it should always be borne in mind that the demands of commercial competition cause them to be very generally overrated by their makers, so that in choosing sizes from catalogues it is advisable to make considerable reduction from the rated capacities. Mr. James R. Willett, in a pamphlet on the heating and ventilating of residences, gives the accompanying table for the sizes of cast-iron boilers. Where boilers of the kinds used for steam power are employed in simple heating plants of large size of which there are to-day but few—there is an old—rule for proportioning the size, to the effect that there must be 1 square foot of boiler surface to 10 square feet of radiating surface. This rule when applied to a boiler with a well-designed furnace, stack and setting should be very conservative. It is much better to estimate the steam consumption of the entire plant (see Chapter V., page 60 for steam consumption of radiation) and calculate the boiler requirements according to the rules of boiler design. The designs of the cast-iron heating boilers are innumerable and can only be selected by the careful judgment of the engineer. The chief points to be considered are efficiency of heating surface and capacity of grate in proportion to surface, strength of parts, tightness of joints, ease of cleaning and effectiveness of circulation—there should be no "dead ends" where the water is not kept in circulation by the heat of the fire.

Willett's Table of Cast-iron Boiler Capacities.
Radiators; Area of boiler Radiators; Area of boiler
total sq. ft. grate in sq. in. total sq. ft. grate in sq. in.
400 500 1,600 1,420
500 580 1,800 1,560
600 650 2,000 1,700
700 740 2,250 1,880
800 820 2,500 2,020
900 890 2,750 2,230
1,000 970 3,000 2,400
1,200 1,120 3,500 2,770
1,400 1,270 4,000 3,120

There might also be classed under the head of steam-heating apparatus the various traps, automatic receivers and other special appliances which are used especially in connection with exhaust heating for returning the water of condensation to the boiler, and which were described in the preceding chapter. But besides these the only apparatus pertaining to a heating system without mechanical ventilation are the radiators, and in the design of the system the utmost consideration should be given to their selection and proportioning.

Fig. 9     Fig. 10     Fig. 11

One, Two and Three Column Radiators.

Radiators.—Radiators are made either of cast or wrought iron and are classified according to kind of heating for which they are used into direct, direct-indirect, and indirect radiators. A few years ago wrought-iron pipe made up with bends and headers into coils of various sizes and shapes was used very largely for radiators, and especially for indirect radiators; but on account of their greater economy of construction, cast-iron radiators are rapidly supplanting the wrought-iron pipes for all purposes. To-day wrought-iron pipe coils are used in direct heating where very large radiators are required spread over a large area of wall surface, such as in factory rooms or warehouses, where a series of long pipes connected into headers at each end are run along the walls, either over or under the windows, preferably under. Pipe coils are also used for large indirect radiators, but in this case, as a rule, only in connection with ventilating fans, as will be described in subsequent chapters.

Fig. 12     Fig. 13     Fig. 14

Flue Radiators.

The styles and kinds of cast-iron radiators are innumerable. They are made in sections or loops, which are fastened together by pipe nipples of various kinds, with a paper or thin metal gasket between the faced surfaces of the joint. These nipples are sometimes threaded and screwed up tight by special wrenches, but what is known as the push-nipple is extensively used. These nipples are not threaded but are turned to a close fit with the holes in the loops, at the joint, which are bored out perfectly true, and they are driven tight by pressure with jacks or presses made for the purpose.

Cast-iron radiators are classified according to the kind of surface, into plain-surface and extension-surface radiators, and according to their style of construction into open and flue radiators, and of the open one, two, three and four column type, according to the formation of the loops. The different classifications will be better understood from inspection of the accompanying illustrations. The extension-surface radiators, Figures 14 and 18, as the name implies, have extensions of various kinds in the form of ridges or pins cast on to the otherwise plain surface, and are used principally for indirect radiators. The flue radiators, Figure 16, are used extensively for direct-indirect radiation, but for such purpose are provided with shields and provision of some kind for connection with the outer air. Flue radiators, such as shown in Figures 12 and 13, are sometimes spoken of as veiled-surface radiators. Most low radiators of considerable width in proportion to the height are of the flue type. The radiator that has been used for direct radiation much more widely than any other is the two-column plain-surface radiator, such as is shown in Figure 10, but the numerous other forms are being more and more extensively introduced.

Most radiators for steam heating have all the loops connected by only one steam passage through the bottom, but some are also connected at the top as well. Such radiators are adapted to steam heating from hot-water heating practice, as they are the only kind that can be used in the latter, and are hence known as the hot-water type. See Figure 20. Almost all low and wide steam radiators are made in this way, and by some authorities this construction is preferred in all kinds. Further discussion of this subject will be taken up later.

Fig. 15     Fig. 16     Fig. 17

Direct-Indirect Radiators.

Measuring radiators.—Radiators are universally rated by the number of square feet of surface which they contain. This, at the present time, is for many reasons a very arbitrary method and holds chiefly for want of a better. The main difficulty lies in the great difference in the value of the various kinds of surfaces, no distinction being made in such rating between plain, extension or veiled surface. The variation is enhanced also by the fact that radiators are to a large extent overrated, especially in the less common sizes and styles; and owing to the difficulty of accurately measuring the surface, this fact is very generally overlooked. A number of methods have been proposed and tried for measuring the surface of radiators which are made in ornamental design and with all kinds of irregular surfaces. In the course of a large experience with radiators of all kinds, the author tried many different methods and finally devised one which he has found comparatively simple and very reliable. By this method all irregular surfaced are measured by covering them with very thin flexible paper which must be carefully turned and folded into all irregularities of the surface. After being thus fitted, the paper should be rubbed by blackened fingers. They are generally sufficiently soiled for the purpose from handling the radiators. In this way when the paper is spread out, the part that was folded under can be readily distinguished, and the actual area of the surface can be determined by measuring the blackened parts with a planimeter. In lieu of a planimeter, thin cross-section paper can be used and the areas determined by counting the small squares. In measuring up a radiator loop, it is best to divide the surface by thin lines of white chalk or paint and measure each division separately. The parts that have a uniform cross-section, as the columns of most direct radiators, can be measured by determining the actual circumference of the surface by fitting a paper around it and multiplying this by the length of the column. It was once objected to this method that it does not take into account the effect of the raised ornamentation on a radiator. This is not the case, or at least any ornament that it does not take into account would not increase the total surface to an appreciable degree. Measurements of radiators by this method by different observers acting independently have been found to check within less than 1 per cent.

Fig. 18 Indirect Radiator.

Fig. 19 Indirect Radiator.


Action of radiators.—Before proceeding further in the discussion of radiators, it may be well to consider some of the principles which govern their operation in practical use. The fundamental principle of their operation is undoubtedly the axiomatic theory that there is a universal tendency toward the equalization of temperature; in other words, that hot bodies give up their heat to the colder ones which surround them. In general this is accomplished by three different processes, namely: conduction, convection and radiation, the word radiation being used here in the special sense of radiant heat. These may best be defined by illustrations. When one part of a rigid body is in contact with a warmer body heat passes or flows from the latter through the former to its cold portions as long as there is any difference of temperature. This is conduction of heat, and the rate of flow under the same conditions of temperature varies as a factor known as the heat conductivity of the body concerned. The heat conductivity of fluids, liquids and gases is very low, practically zero, but heat is transferred in them by convection. The particles in direct contact with the source of heat are heated above the temperature of the rest, and an increase in temperature of any liquid or gas invariably decreases its specific gravity and causes an upward current of the hot particles, which brings the colder particles in turn in contact with the hot body, thus maintaining a circulation which tends to raise the temperature of the entire mass. Radiation or radiant heat is entirely different from either of these in its action. It is a wave motion and travels through air and all transparent bodies with the velocity of light without heating them, and only appears as sensible heat when its course is interrupted by opaque bodies by which it is absorbed.[1]

Fig. 20 Hot-Water and Steam Pipe Loops.

A radiator gives out its heat to its surroundings by radiation to the walls and objects and by convection to the air. What proportion is given out in each way it is difficult to measure in any case and depends principally on the construction of the radiator and the way it is set, but also more or less upon the conditions of temperature, nature of surface, etc. Péclet, the great French physicist, in the middle of the century, fully investigated the laws of radiant heat as well as those of convection in still air. The formulas which he deduced are applicable only in a limited degree to radiator practice. His investigations, as well as those of others since that time, showed that for a single iron pipe in still air[2] under conditions of temperature which prevail in radiator practice, the heat given off as radiant heat is just about equal to that given out by convection.[3]

But radiators are invariably built of clusters of pipes or surfaces, and as radiant heat travels only in straight lines and perpendicular to the surface of its source, a large proportion of the surface is wasted so far as radiant heat is concerned, due to what may be called the mutual interception of the rays. In the ordinary one-column cast-iron radiator, the proportion of surface from which no radiant heat takes place is nearly 20 per cent., in the two-column, 45 to 55 per cent. and in the three-column, 55 to 65. Assuming that the radiant heat amounts to one-half the total, the reduction of heat emitted would be one-half of these percentages. Another fact which further reduces the radiant heat is that radiators are usually set very close to a wall which becomes heated to a comparatively high degree and consequently radiates back a large portion of the heat to the wall side of the radiator. This is true to an extreme degree in the case of indirect radiators which are enclosed in boxes of wood or sheet metal and are not located in the room they are to warm. With these the heating is accomplished entirely by convection, while with direct radiators the radiant heat rarely amounts to 40 per cent., generally not over 30, and often in practice considerably less. For a radiator in any particular location the radiant heat is constant for the same conditions of temperature of the radiator and surrounding objects, and is independent of currents of air. This is by no means the case with the convected heat, which is increased greatly by a slight draft from any extraneous source. In this connection it is very remarkable what a great effect an almost imperceptible draft will have on the heat given out by a radiator. This is partly due to the lowering of the temperature of the air between the loops, but also to the fact that with the same temperatures any increase in velocity increases the amount of heat the air absorbs.

Radiator tests.—Numerous tests of radiators have been made since those of Mills, Richards and others in the early '70's, but there is a wide variation in the results obtained, due partly to the different kinds of radiators tested and partly to the different methods of testing. As yet, no standard means of testing radiators has been adopted. The steam radiator as a heat-using device is theoretically perfect; that is, all of the heat that is put into the radiator by the latent heat of the steam condensed is given out to the air and objects surrounding. Its efficiency is therefore 100 per cent. The question of practical efficiency is, therefore, more strictly speaking, only one of effectiveness of surface. That is, of two radiators under exactly the same conditions of temperature and surroundings, that one which has such an arrangement of its surfaces as to give out the most heat per square foot is the most effective, usually called the most efficient. In all tests of radiators, the heat given off is measured by connecting them so that the steam which condenses can be accurately weighed, its pressure, quality and temperature being determined at the same time. The results are generally reduced to British thermal units given off per square foot per hour per degree difference of temperature between the air of the room and the steam of the radiator.

Fig. 21 Prof. Carpenter's Arrangement for Testing Radiators.

Tests of radiators have been made in various ways by Mr. George H. Barrus, by Profs. Denton and Jacobus of the Stevens Institute, by Prof. R. C. Carpenter, of Sibley College, Cornell University, and by the author. The results of Prof. Carpenter's tests are published in detail in his valuable work on "Heating and Ventilation of Buildings. In these tests the radiators were located in separate compartments, 7xlO feet, built together in a large room, and as shown in Figures 21 and 22. In order to allow some circulation of air so that the temperature of air of the compartments might not get as high as that of the radiators, small openings were made in each at the bottom and top. In 1895 and 1896 the author had occasion to make comparative tests of a large number of radiators of various kinds and types, and the arrangement used by him for testing is shown in Figures 23 to 25. The two test rooms were built in the main floor of a large warehouse, and each was 15 feet by 11 feet 8 inches, and extended to the ceiling, 15 feet 5 inches high. The walls of the test rooms were built of matched and beaded pine and lined with lapped courses of heavy building paper. The warehouse room in which the test rooms were located was about 85 by 50 feet, with brick walls on both sides and wood and glass partitions at each end. Neither end of this large room, however, was open to the outside air and the side walls were party walls. Every effort was made to keep the air of the test rooms free from all drafts except those induced by the column of hot air rising from the radiators and to otherwise make the conditions as nearly as possible those of actual practice. To permit some circulation in the test room, so that the air would not get too hot, an opening 4 inches long and 18 inches high was cut in the front wall at the floor, and to prevent direct drafts on the radiators, those openings were surrounded inside by a wooden screen, 2 feet 8 inches high. The front partition was also opened 18 inches at the ceiling. The piping, as shown in Figure 25, was arranged so that the steam was supplied to the radiators on what is known as the one-pipe overhead system. Steam was supplied to the separator at a pressure of 2 or 3 pounds above atmosphere through a 2-inch pipe from a small heating boiler. The pressure carried at the boiler was slightly in excess of that on the separator, it being throttled at the latter. By this means and by leaving the drip on the separator slightly opened, the steam was supplied free from all entrained moisture. The piping, separator, etc., were all carefully covered. The heat given off was measured by drawing off the condensed steam from the drip pots into cold water and accurately weighing it.

Fig. 22 Prof. Carpenter's Arrangement of Each Radiator and Compartment Removed.

In practically all of the tests made in this plant, one radiator was used as a standard and all the others tested and compared with it. In each test the radiators were connected and a preliminary run made with open air-valves until the conditions became constant and uniform, and a test run made for two to three hours. The radiators were then interchanged and the test repeated. Even with these precautions it was only by exercising the greatest care that it was possible to obtain results which checked closely. It is very remarkable what a decided effect can be created by a very slight motion of the air from an external source. Opening a door at one end of the warehouse room, although, as before stated, these doors did not open outdoors, made a decided difference, and if a strong wind was blowing outside, the radiator to the windward side had some advantage, although no draft would be perceptible. However, with due care, the two tests for each comparison were made to check with fair accuracy.


Fig. 23 Front Elevation of Testing Rooms.

Fig. 24 Plan of Testing Rooms.

Fig. 25 Elevation of Piping. Radiator Tests.

The radiator used as the standard on these tests was an ordinary cast-iron two-column steam radiator, 38 inches high, with but little ornamentation. The writer believes that this is the only way of accurately testing radiators, and the adoption of any one definite make of radiator as a universal standard of comparison would do much to extend the knowledge of the comparative effect and value of radiators. Tests of radiators made in different ways or in different locations are valueless for accurate comparison. But all comparative tests made against the same standard if accurately and carefully carried out, could be compared in percentages of the heating effect of the standard used. The writer found that under the conditions in his testing plant the 38-inch two-column cast-iron radiator used as a standard gave out 1.60 British thermal units per square foot per hour per degree difference of temperature with an average steam temperature of 224 degrees Fahr., and average temperatures of the rooms of 76.5 degrees. The average difference of temperatures was 147.5 degrees.

This cofficient of 1.6 B. T. U. per square foot per hour per degree difference of temperature between the steam and air is somewhat lower than that which Prof. Carpenter obtained for a radiator of almost the same size and design. Assuming that the radiators were exactly alike, such variation as there was can be due to two causes: 1, a variation in the difference of temperature between the steam and the surrounding room; and 2, the mode of setting and the consequent freedom of air circulation around the radiator. In regard to the first cause, all tests of radiating surfaces from Péclet down show that the coefficient is greater, the greater the difference of temperature, and for extreme variations in the difference of temperature, the coefficient is very much greater than in the limits of ordinary radiator practice, with steam temperatures from 212 to 230 and mean air temperatures from 40 to 70; within which range the variation in the coefficient from this cause is less than 9 per cent. In regard to the second cause—the freedom of the air-circulation around the radiator—this is by far the chief cause of difference in action of radiators. Profs. Denton and Jacobus of the Stevens Institute of Technology made some comparative tests of radiators, published in The Engineering Record of September 8, 1894, with a plant very similar to that used by the author, except, besides having an opening at the top, there was in each of the test rooms an outside window, "which was opened a certain amount during the tests"—a dangerous way, the author believes, to test radiators with the expectation of obtaining checking results—although "a screen was placed between the radiators and windows to prevent direct drafts from striking the radiators." These tests were unquestionably carefully made and were checked by reversing the position of the radiators, so that the comparative results obtained may be taken as reliable; but the coefficients were in the neighborhood of 20 per cent. higher than that obtained by the writer on similar radiators, due entirely to the greater freedom of air-circulation from the open windows. This is a matter of great importance in practice and in consequence the results obtained in radiator tests depend largely upon the setting of the radiator. It is this fact also which makes a radiator to some extent automatic or self-adjustable. Take for example, an ordinary room say with a north exposure and one direct radiator set, as they generally are and always should be, under the window. On a moderately cold day with the thermometer outside at 20 degrees and the wind from the south or east, the radiator will be turned on the entire time. On another day with the thermometer outside at zero and the wind from the north, the radiator very probably keeps the room at the same mean temperature with practically the same temperature of steam. The reason is that in the latter case the cold air which leaks in through the walls and around the window casing, besides keeping the air immediately in contact with the radiator at a lower mean temperature, causes a much more rapid circulation around the radiator, with the result that it gives out more heat units per square foot. It is for this reason, too, that one may put down for a positive and infallible rule in radiator design that the radiator which has the most open space around its surfaces and the most interrupted exposure to the surrounding air will give out the most heat per square foot under the same conditions of setting. In compliance with this rule, other things being equal, narrow radiators are more effective than wide ones, and low ones than high ones; but the effect of width and height can more than be offset by a slight increase in the distance between the loops; for example, the author found that a four-column 38-inch radiator gave out exactly the same heat per square foot of measured surface as a two-column 38-inch radiator, because the former had a mean distance between the loops about 16/100 of an inch greater than the latter; also a 38-inch flue radiator was improved 7 per cent. in heating effect by separating the loops ¼ inch.

It is largely for this reason of freer circulation around the surface that the ordinary wall coils of 1 or l¼-inch wrought-iron pipe, which are extensively used in factories, are much more effective than cast-iron radiators. Some careful tests by Prof. M. E. Cooley of the University of Michigan showed that a single layer coil of horizontal pipes gives out 40 per cent. more heat per square foot than a two-column cast-iron radiator under the same conditions of setting.

It may be further stated in compliance with the same rule that the hot-water type of radiators is somewhat less effective than the steam type on account of the obstruction offered by the loop connection at the top, and also that flue radiators are less effective than the ordinary open type. Just how much less effective the flue types are depends upon the kind and design, but the wide low flue radiators, not considering extension-surface types, are sometimes over 23 per cent. less effective than the ordinary two-column radiators of the same height. It may also be stated that there is a greater difference between the high and low radiators of the flue type than of the open types. One make of two-column cast-iron radiator proved 6½ per cent. more effective in the 20-inch height than in the 38-inch, and another make nearly 10 per cent., while an 8-inch wide flue radiator showed over 20 per cent. improvement in the 20-inch over the 38-inch.

Condensation at different pressures.—Mr. W. J. Baldwin made some tests some years ago to determine the relative heating effect of radiators under different steam pressures, with the same conditions of setting. Taking the condensation at 1 pound steam pressure as 100, he found the condensation at other pressures to be represented by the following figures:

1 lb. pressure ... 100
5 lbs. pressure ... 108
10 lbs. pressure ... 118
15 lbs. pressure ... 126
20 lbs. pressure ... 134

Extension-surface radiators.—In regard to what is known as extension-surface radiators, they are generally made in the flue form, and numerous tests show they are never as effective per square foot of actual surface as are the radiators which have no extension surface. Profs. Denton and Jacobus made some tests to illustrate this point. They tested two forms of extension-surface flue radiators and then planed off the extensions and re-tested them. They found that with one radiator, which had 43 per cent. more surface in the first condition than in the second, gave only about 17 per cent. more heating effect. This, however, is not exactly a fair comparison on the grounds of extension surface alone, as the radiators had a much more effective proportioning of air-space design with the extensions planed off.

Surface of radiators.—There are other considerations which alter the effectiveness of a radiator besides the design and setting, notably the condition of the surface. Radiators are rarely used with the natural iron surface, but are painted in all possible ways. The nature of the surface has practically no effect on the convected heat, but a very decided effect on the radiant heat; but as the latter is generally less than 35 per cent, of the whole, the effect on the total heat emitted is not so marked. Furthermore, usually only the top and outside surface of one side is painted at all. In general the dark and lustreless paints are the most effective, and may even improve the heating power, while the bright shiny metallic paints may reduce the effect quite decidedly.

But few tests as to the effect of paints have been made. Prof. Carpenter found that two coats of black asphaltum increased the total heating effect by 6 per cent., two coats of white lead 9 per cent., rough bronzing about 6 per cent., while a coat of glossy white paint reduced it by 10 per cent., although the kind of radiator considered is not mentioned. The author found that two coats of ordinary "radiator japan paint" had but little, if any, effect, but in one case, on a 38-inch flue radiator, three coats of gold bronze reduced the heating power by over 12 per cent. This loss was probably due partly to the reduction in radiant heat from the polished surface and also to the fact that the convected heat was somewhat reduced by the heavy coating of paint acting as a non-conductor. This is doubtless sometimes the effect with old radiators which have been painted several times.

Location of radiators.—As before stated, the setting of the radiator has a decided effect on its heat-giving power, and no two conditions can be considered exactly alike in this regard. For direct radiators, the best place to set them is unquestionably under a window. The reason of this is that the greatest leaks of cool air from outside are around the window frames and the greatest loss of heat by radiation is from the glass. There is in consequence a decided downward current of cold air at the window which, if the radiator is on the opposite side of the room, rushes across the floor and is accelerated by the upward current of hot air from the radiator. Such a condition tends very decidedly to make cold currents along the floor. If the radiator be placed under the window the cold drafts are interrupted and the heat more diffused moreover, the upward current from the radiator, the downward current from the window and the leakage drafts, all combine to make a resultant draft of cold air against the side of the radiator which lowers the temperature between the loops and altogether tends to increase the effectiveness of the surfaces very considerably over what would be found in still air.

Direct radiators are often put in recesses under windows and low radiators are sometimes placed under window seats. Such settings, while highly desirable from an aesthetic consideration, decidedly change the effectiveness of the radiator both by shutting off the radiant heat and by lessening the free convection. From some rough experiments the author is led to believe that the ordinary marble top, which is often placed on radiators, will reduce the heating effect from 6 to 15 per cent., depending on the size, kind and height of the radiator; but this is not stated on the authority of a careful comparative test.


  1. Some bodies which are quite opaque to light waves are more or less transparent to heat waves, and vice versa, and nothing except the immaterial ether is absolutely transparent to them. The earth's atmosphere absorbs a considerable amount of the radiant heat from the sun.
  2. By "still air" in this sense is meant the air of a room in which there are no currents except those created by a column of hot air rising from the heated surface.
  3. For a complete account of Péclet's experiments and his results see "A Treatise on Heat" by Thos. Box; London: E. & T. N. Spon.