How to Generate Steam Energy from Furnace Waste Heat

Technical progress takes place in two directions: the improvement of methods, affecting the quality of the product; and increase in the economy of operations, affecting its cost. In the iron-industry, practice is pretty well settled, and revolutionary changes of method are, for the present, not to be expected. Hence, in this industry, the principal endeavor is to reduce costs by the use of labor-saving inventions, such as the machines brought into use during the last decade. But it is not alone necessary to reduce the labor-cost; complete economy requires also a saving in raw material. Under this head we may rank the coal employed as metallurgical fuel; and it follows that progress includes the economic use of fuel, so managed as not to injure quality of product. In the present paper it will be shown how, by suitable utilization of the heat from the furnace, a large portion of it may be made available for other purposes.
Such utilization of waste-gases is important in all industries which employ furnaces (especially heating-and smelting-furnaces); and in the first rank of these stand the iron-, glass-, and porcelain-manufactures. The following pages are devoted chiefly to the waste-gases of smelting-works, and especially to those reverberatory-furnace gases which are not combustible.

The object of metallurgical furnaces is either to fuse the material charged or to bring it to a high temperature required for further mechanical manipulation. Leaving out of consideration for the present the reverberatory fusion-furnaces, we have to do with the heating-furnaces employed in connection with rolls, hammers, or hydraulic presses. The temperature to which the material must be raised in such furnaces varies, according to quality and use of product, from 800° to 1,200° C., and, of course, the hearth itself must be still hotter, as will be also the gases coming from the hearth (except in the continuous heating-furnace). It follows that the efficiency of such a furnace with respect to the utilization of fuel must be very low, as will appear from the following statement of the matter.

If ta and te be the temperatures at the beginning and the end of the hearth, respectively, η1, the theoretical ratio of utilizable heat, will be η1 = ta — te/ta. For instance, if the initial temperature be 1,400° C., and the final temperature 1,000° C., the percentage of available heat will be 1,400 — 1000/1,400 = 0.29 or roughly, 30 per cent. This calculation makes no allowance for the great loss of heat by radiation and conduction, which will be discussed hereafter.

In the instance given, on the assumption that there was no suction of air through the doors, the combustion-gases have lost, on the way from fire-bridge to flue, say 30 per cent, of their original heat, and there should remain, therefore, some 70 per cent, of the heat generated on the grate, to be made available (if it be possible to cool the gases to 0° C., or to the temperature of the outer air, as the case may be).

The simplest way to utilize this heat would be to prolong the furnace, and move the charge towards the fire, so as to heat it on the counter-current principle. By this method it is possible in the continuous heating-furnace to operate the furnace itself at a lower temperature, thus securing additional fuel-economy. Of course it is not the difference between initial and final temperatures only, but, as already observed, all incidental losses of heat also, which must be considered in maintaining the efficiency of the furnace. Making allowance for the great losses by radiation, which depend upon the temperature of the furnace, one soon arrives at a practical limit of furnace-size, beyond which it is not advantageous to go. The lower the flame-temperature at the end, the greater the efficiency of the furnace.

As to further reduction of chimney-temperature, however, it must be inquired whether the draft may not be thereby diminished. Theoretically, the maximum draft is obtained at about 300° C. of gas-temperature. An example will show how the draft is affected by higher or lower temperatures.

Let us assume that the stack of such a furnace has a draft of 20 mm. water-column, one-half of which is used in overcoming the resistances in the furnace and flue, and on the grate, leaving 10 mm. to affect the speed of the escaping gases. In order to determine the volume or weight of the gases drawn per unit of time through a stack of say 1 sq. m. area, we will consider three cases, assuming the gas-temperature at 200°, 300°, and 400° C. respectively.

Under the conditions assumed, the friction in the stack itself being neglected, we have v = √2gh for the velocity of the gases entering the stack with an effective draft-pressure of 10 mm. water. Substituting for the water-column hw the corresponding air-column in millimeters, γ being the specific gravity of the chimney-gas and γw that of water, we have √2ghwγw/γ as the velocity of the chimney current. At the given temperatures, the specific gravities of the chimney-gases are approximately :

specific-gravities-of-chimney-gases

Calculated upon these figures, the respective velocities of the chimney-gases in the three cases will be roughly 16, 18, and 19 m. per second; and the weight of gas passing per second through the assumed cross-section of 1 sq. m. (no allowance being made for resistance in the stack) will be roughly 12 to 10 kg. at these three temperatures. It appears at first sight that the weight of gas actually moved by the stronger draft is not greater, but, on the contrary, somewhat smaller. But this contradiction is not real; for we have assumed the absence of frictional resistance in the stack, whereas, this resistance not only exists, but may be confidently assumed as greater for the cooler, and smaller for the hotter (i. e., lighter), gas- current. The safe practical conclusion is that, within certain limits, the weight of gas drawn through a chimney remains constant under moderate increase or diminution of temperature. It follows likewise, that a considerable cooling of the gases for utilization of their surplus remaining heat cannot injure the process of combustion, provided the draft in the chimney is strong enough to overcome resistances, and produce the necessary gas-velocity. In most cases, the chimney-draft has to be throttled anyhow; and in cases where blast-pressure (Unterwind) is used, there is no room for any fear of disadvantage through reduction of draft.

The temperature of the waste-gases, though varying with the purpose for which the furnace is operated, is usually high enough for the utilization here to be described. According to the method of calculation employed above, the available heat in gases leaving the furnace at 1,000° C. and entering the chimney at 300° C. would be 1,000 – 3,00/1,000 or 70 per cent.; that is, under the conditions assumed, about 70 per cent, of the heat of the gases leaving the furnace should be available for other purposes. As a general rule, however, this figure is not reached, because, for reasons of construction and imperative considerations of safety, the boiler which is to be heated by the waste-gases must not be directly a part of the furnace construction, and hence a fall of temperature will take place in the connecting main. Assuming in our example such a fall of 100° C., we have, as the available percentage of heat, 900 – 300/1,000 or, about 60 per cent.

It is evident at once from this discussion that the arrangement of furnace and boiler should be such as to invoice the smallest possible fall of temperature between the two. Long mains are therefore to be avoided as far as practicable, since they involve a considerable relative fall of temperature. Fortunately there are many constructions which, in most cases, satisfy the requirements above stated.

The total theoretically available heat-energy of the coal burned on the grate is

formulae-heat-energy-from-coal-burned

the flame-temperature being ta at the beginning and te at the end of the hearth; t’a initial, and t’e final, temperature of the waste-gases. And since (apart from losses between furnace and boiler) te = t’a, the formula becomes

heat-furnace-formula

In the example given, ta = 1,400° C.; te = t’a = 1,000° C.; and Fe = 300° C. Hence,

η = 1,400 – 3,00/1,400 = 0.79

whence it appears that by the addition to a heating-furnace of a boiler heated by waste-gases a total theoretical availability of nearly 80 per cent, would be realized, about 30 per cent, being contributed by the furnace and 50 by the boiler. The theoretically possible utilization of heat is thus essentially higher in the boiler than in the furnace therewith connected.

So far we have not considered the various losses in the furnace and at the boiler. But they must necessarily receive attention, since only so can the increase of economy produced by adding to a heating-furnace a suitably constructed waste-gas-heated boiler be placed in the right light.

The modern rolling-mill furnaces, directly fired or with half-gas firing, combine with a maximum of work a utilization of the heat of the gases amounting in certain favorable cases to perhaps 15 or even 20 per cent. Assuming for less productive furnaces an economic degree ( Wirkungsgrad) of only 10 per cent, (by which is meant the relation between the heat actually utilized in warming the ingot or bloom and the total heat generated on the grate), it appears that of the 30 per cent, available for this work (according to the calculation in our example) only one-third is really brought out. The remaining 20 per cent, of the total heat generated we must regard as lost through radiation and conduction, as well as through incomplete combustion.

The loss by radiation and conduction of a quantity of heat greater than the quantity utilized is an extreme but conceivable case, in view of the high temperatures occurring within the furnace. According to Stefan-Boltzmann, the transfer of heat by radiation varies as the fourth power of the absolute temperature of bodies concerned.

In a boiler of proper dimensions and construction, the conditions are essentially more favorable, since here only the loss by radiation makes itself felt to any considerable extent. Assuming a total boiler-loss of 15 per cent, of the theoretical boiler-efficiency, we shall have, notwithstanding, in our example, an actual realization of 85 per cent, of the available 60 per cent., or about 50 per cent. Thus, by the addition of a properly constructed boiler to a heating-furnace, the utilization of the fuel-energy may be raised from 10 per cent, to 60 per cent., whence it is again clear that such an addition is imperatively demanded in the interest of furnace-economy.

Thus far, we have left out of consideration the fuel and the conduct of combustion. But in the operation of the reverberatory, it is, as already observed, necessary not only to maintain the furnace-temperature required by the pieces to be heated, but also, in many cases (particularly in furnaces heating “ quality ’’ steels), to keep the composition of the gases in the furnace such as to exclude, as far as possible, any deleterious action upon the charge, which it must be our aim to protect, not only against overheating, but also against too great oxidation. The flame should carry as little excess of air as possible; so that it seems in many cases better to work with too little than with too much air. In the former case, there would be a loss of unconsumed gases; but this would ordinarily be smaller than the loss occasioned in the latter case by the burning of expensive material in the charge.

The amount of loss from incomplete combustion depends chiefly on the construction of the furnace and the manner of firing. Investigations of this question made by the present writer have given various results, so that (as might, indeed, have been expected after the foregoing statements) no generally applicable figures can be given. In one experiment, no less than 6 per cent. of CO was found in the waste- gases of a furnace heating large blooms. In the very great majority of cases the furnace could be run with, perhaps, from 2 to 3 per cent, of carbon monoxide in the waste-gases, so as to avoid excessive burning of the charge. This loss through imperfect combustion has been omitted from the foregoing discussion because, being so highly dependent upon various working-conditions, it cannot well be expressed in a general formula. That such losses are not exceptional, has been shown by Dr. M. Philips, who found in the waste-gases from a half-gas furnace more than 9 per cent, of carbon monoxide, and also a considerable amount of hydrogen.

Before considering in detail the several kinds of waste-gas boilers, we should inquire why the production of steam is relatively the simplest and easiest, and also commercially the most profitable, way of utilizing the heat of such gases. The other ways which have been proposed are: Preheating the charge; preheating the air for combustion; and preheating the boiler-feed, or superheating the steam, of boilers directly fired.

The preheating of the charge in continuous heating-furnaces has been mentioned already; and it has been shown that there is a practical limit, beyond which the furnace cannot be advantageously lengthened for this purpose.

The heat of the waste-gases is applied satisfactorily to the preheating of air in half-gas furnaces, and of air and gas in gas furnaces.

For advantageous preheating of the boiler-feed, the difference in temperature between the escaping gases and the water to be thus treated is too great. The intensive heating of an ordinary amount of water would absorb but a small quantity of heat, and consequently a large quantity of heat would not be utilized, unless much more water were preheated than is called for in such establishments.

It would be easy to superheat steam by means of the waste furnace-gases. Not to mention the fact that the economic result might leave much to be desired, this method of utilizing the hot gases is open to a serious objection, namely, that the operation of the furnace does not uniformly correspond with the consumption of steam. It might happen, for instance, when the engines of the works were using but little steam, that what they received would be too hot; or, on the other hand, when the several engines happened to be all working at once, and thus taking a large aggregate of steam, it might not be sufficiently superheated.

Yet this method is occasionally practiced. At one of the works of the U. S. Steel Corporation a large central superheater was installed, for raising to 240° C. the temperature of 110,000 kg. of steam per hour, at a pressure of 10 atm. This apparatus received the waste-gases from four ingot-heating furnaces and superheated the steam for four roll-trains, requiring a total of 8,000 h-p. Whether such an arrangement can satisfactorily utilize the heat of the chimney-gases may be doubted, especially in view of the fact that in superheating steam, there must be, to secure a satisfactory efficiency per unit of heating-surface, a very great difference of temperature between the heat-giving and the heat-receiving material. If the transfer of heat per square meter of heating-surface is large, then the temperature of the gases leaving the superheater will be high—and vice versa.

The heat of waste furnace-gases can be most economically utilized in the production of steam. Steam is used everywhere and always in metallurgical works. Even those which are run throughout by electric power, derived from gas-engines, are provided with steam-boiler plants at least as a reserve, or for use in special operations. By contributing to the general steam-conducting system the steam generated by the heat of waste-gases, the work, and therefore the fuel-consumption, of the directly-fired boilers is diminished. Moreover, the difference of temperature between the hot gases from the furnace and the water under steam-pressure in the boiler is about the same as between the flame and the water in a boiler directly fired, so that the transfer of heat will be as satisfactory in the former case as in the latter.

One objection may be raised. It is well known that the operations of furnace and boiler do not continuously correspond. The times of maximum production of steam by means of the furnace-gases are not always those of the greatest consumption of steam from the boiler. But this circumstance will very seldom be important enough to impair the economy of the method. For, on the one hand, steam is almost always in use somewhere in the plant, and, on the other hand, boilers containing a considerable quantity of water are quite capable of storing temporarily a considerable quantity of heat—especially when (as is almost always the case) the steam-pressure is below the normal.

It may be added that in modern construction the proportion of size between furnace and boiler is very good, so that the two can exist side by side, without making the boiler awkwardly large. A comparison of the heat which can be developed by 1 sq. m. of grate-surface with that which can be taken up by 1 sq. m. of boiler heating-surface, shows that the boiler need be only about as large as the furnace, or, in some cases, may be even smaller.

Selecting a Boiler System

The first boilers built for this purpose combined economy of ground-space in the works with the direct upward discharge of the spent gases. The vertical cylinder-boiler stood, as it were, in the enlarged lower portion of a chimney, the hot gases played around it up to a given height, and were there taken up, generally by two diametric sheet-metal flues, at the side of the boiler, and conveyed away, upward. The vertical cylinder-boilers had generally a diameter of from 1 to 1.2 m., and a total length of 12 or even 15 m.

This type, often employed in the ’60’s of the last century, had two advantages:

  1. it occupied, together with the surrounding masonry, only about 6 to 8 sq. m. of ground-space; and
  2. its chimney was combined with it—a welcome feature from the standpoint of furnace-management, because it assured the complete independence of each, furnace.

This form of boiler presents, however, such weighty disadvantages that its use is no longer seriously considered. A very questionable improvement, was the vertical flame-tube boiler of Hall.

Gradually, the consideration of ground-space gave way to that of safety, and horizontal boilers, which could be more easily shut off, were placed behind the furnaces. The so-called Bouilleur boilers (cylinders, with one or more boiling-tubes [Sieder]) were first used, and afterwards came fire-tube boilers. These types permitted, through the establishment of a direct connection between furnace and chimney, the operation of the furnace with the boilers cut out of the system; and thus assured a greater safety. The development of heating-surface easily followed, since wide limits were provided for it. The degree of utilization—in other words, the amount of steam power produced—per kilogram of furnace-coal was greatly advanced by the use of the fire-tubes; the boilers were much more easily served and cleaned than the vertical ones; in short, this would have been recognized as the ideal form for a waste-heat boiler, if it had not demanded such a disproportionate ground-space. While a simple puddling- or heating-furnace occupied, say, from 16 to 20 sq. m., the accompanying fire-tube or Bouillier boiler called for from 20 to 40 sq. m., or almost twice as much as the furnace itself.

The necessary ground-area is smallest when the boiler is placed over the furnace. If it occupies not more, or not much more, horizontal area than the furnace, the question of space retires completely into the background, and the furnace can be designed to suit the conditions of operation with a freedom previously unknown, since furnace and boiler can now be regarded as a coherent whole. This conception seems to have taken tangible form first in America, where water-tube boilers were placed above and behind puddling-furnaces. But in designing such overhead boilers, the special kind and operation of the furnace must be kept in mind. A furnace within which the highest temperatures prevail suffers from these, or through the consequent expansion of its refractory material, certain changes, which the strongest armature cannot prevent. These movements or changes of form must be absolutely kept away from the boilers above; otherwise, the position of the boilers would be altered in the course of time, with a certain lack of safety in operation as a natural result. Nor should it be forgotten that a furnace, according to the demands made upon it, will have to be rebuilt much sooner than a boiler, even when the latter is placed over it. Though furnace and boiler constitute an almost indivisible whole, they must be built separately, so that changes suffered by the furnace may not be communicated to the masonry of the boiler, and thus to the boiler itself. This requirement is satisfied by mounting the boiler upon independent cast- or wrought-iron columns, placed at a certain distance from the furnace-buckstays. The use of these boiler-columns as part of the armature is, for the reasons just given, not permissible. They can be in each case so grouped around the furnace as not to hinder its operation. In many instances the structure thus carrying the boiler may serve as a support for door-levers, etc.

The question of total height for furnace and boiler plays no part in modern plants. Even in extreme cases, the highest part of the boiler would be scarcely 7 m. above the furnace-floor; whereas the traveling-cranes in modern works run on a still higher level.

The relative position of boiler and furnace having been determined, the next question concerns the kind of boiler which will involve the least interference with furnace-operations, while possessing the maximum technical efficiency, and not requiring too much in the way of attention and maintenance. Of the various boilers claiming consideration, we may name the fire-tube, the combined fire- and smoke-tube, the water-tube, and the boilers of locomotive and loco-mobile form. All of these are more or less suitable for the purpose in view, and it will require special study in each case to decide which should be adopted.

One fundamental difference among them is, that the flame-tube and water-tube boilers require to be set in masonry, while the other types named can do without it, since only their interior surfaces are played upon by the hot gases. In consequence of this masonry setting, boilers of the first-mentioned types sometimes seem somewhat large and clumsy, as compared with the furnaces, whereas the boilers which have interior firing exclusively look much more graceful and small, as indeed they are, since they have, instead of the thick masonry setting, a sheet-metal case encircling them at a small distance. Behind this, the non-conducting material lies upon the mantle of the boiler. By reason of the smaller size of this type of boiler, its use affords some advantages in the better lighting of the works.

Waste Gas Boilers for Directly Fired Heating Furnaces

Fig. 1 shows in section the usual arrangement of a fire-tube boiler above the furnace, as executed by the Deutsche Huttenbaugesellschaft (the German Furnace-building Company). The hot gases from the furnace traverse first the fire tubes, then a superheating chamber, and then return along the floor and on both sides of the boiler, to be finally conveyed by a sheet-metal flue to the chimney.

A required cutting-off of the furnace from the boiler is effected by the slide s, which, in a properly constructed apparatus, could scarcely give occasion for disturbance. When the valve is closed, the hot gases pass from the furnace through the opening a, Fig. 1, into a masonry pier, under the sheet-metal flue, and through this to the chimney-flue, which is lined for a short distance with refractory material.

Where the gas or flame enters the fire-tube, the precaution should be taken of providing the first section—say from 1 to 1.5 m.—of the latter with a refractory lining, to avoid injury to the boiler from impinging jets of flame.

Experiments covering more than a month with a furnace of this kind, built for a rolling-mill, gave, according to the above-named furnace-building company, the following results:

Heating-value of the coal-mixture in its original condition…………………………….6,192 h. u.
Heating-value of the coal-mixture in air-dried condition……………………………….6,763 h. u.
Heating-surface of the boiler………………………………………………………………………..80 sq. m.
Water evaporated by 1 kg. of coal burned on the grate………………………………………2.71 kg.
Average evaporation of the boiler per hour for 1 sq. m. surface……………………………12 kg.
Coal-consumption for 100 kg. of pieces, charged cold……………………………………..10.1 kg.
Total weight of pieces charged per shift…………………………………………………….42,000 kg.

In order to determine separately the parts played by furnace and boiler in utilizing the heat from the coal, each must be considered by itself.

section-through-furnace-and-overhead-flame-tube-boiler

On the assumption of a complete combustion, there were generated about 68,000 h. u.,of which, in round numbers, 100 x 0.168 x 1,100 = 18,000 h. u. were transmitted to the pieces heated. The utilization of heat in the furnace was therefore about 26 per cent., while about 17,000 h. u., or 25 per cent., were consumed in the production of steam. The total heat-utilization was therefore about 51 per cent. These figures have approximate value only, since the temperature of the material as charged and as withdrawn was merely estimated.

If the calculation be pursued further, and 10 per cent, of CO2 be assumed in the gas at the end of the boiler (after complete combustion), it is found that there is a loss of about 19 per cent, in the gases escaping at about 300° C. from the boiler. Hence the loss by radiation, conduction, and unburned residues is 100 — 51 — 19 = 30 per cent.. That is to say, the loss by radiation, conduction, etc., for both furnace and boiler, is 30 per cent., of which about three-fourths is to be charged to the furnace, and one-fourth to the boiler.

joint-construction-of-furnace-with-waste-heat-boiler

The furnace in Fig. 1 having a long hearth per se, there is no discrepancy of size between furnace and boiler; but in the case of a smaller furnace, this boiler (a 2-fire-tube boiler, from 8 to 9 m. long) would seriously over-balance the furnace, or else the boiler would have to be shortened, with a considerable diminution of heating-surface. But in the presence of a large quantity of gas at high temperature, the heating-surface must be correspondingly large, in order to convey a large amount of heat to the boiler-water. This circumstance has led to the adoption of water-tube boilers, which permit for a given ground-area a great development of heating-surface, and are adapted to the highest steam-pressures.

Fig. 2 shows the combination of a furnace with a Durr water-tube boiler placed partly over and partly behind it. Since, in this instance, the available space was not sufficient for water-tubes of the normal length of 5 m., it was determined to use a boiler containing several sets of tubes, 3 m. long, and placed over one another, and, moreover, to provide for two passes only, so that the hot gases ascending in the first pass would flow along the superheating-pipes (which were built into the masonry) and the lower part of the boiler-mantle, and then return downward, around the bundle of water-pipes, to the chimney- flue. The possibility of cutting off the boiler from the furnace was, at least partly, secured in this arrangement by building into the wall under the first pass a Schlitz sliding-plate (a, Fig. 2), which, during the normal operation of the furnace, is protected by a layer of sand from the direct action of the fire. This does not effect a complete separation of furnace and boiler; but it permits the waste-gases, in case of need, to be drawn, under the first pass, directly into the chimney-main.

Thorough experiments with this boiler occupied 11 days, during which the furnace heated by day the pieces for the rolling-mill (which was not at that time running at night). During the night-shift, the furnace was merely kept warm. The measurements and results were as follows:

Grate-area of the furnace……………………………………………………..1.9 sq. m.
Hearth-area (width, 1.8 m.)………………………………………………….8.2 sq. m.
Ratio of grate-area to heating-surface of boiler……………………………1:48
Coal-consumption per sq. m. grate-area for each hour of the working period, about………….200 kg.
Average evaporation per sq. m. and hour, as above…………………………………….13 kg.
Maximum evaporation per sq. m. and hour, as above…………………………………17 kg.
Average heating-value of the coal-mixture, about…………………………………..4,900 h. u.
Average steam-pressure………………………………………………………………………….5 atm.
Average steam temperature…………………………………………………………………….270° C.
Superheating…………………………………………………………………………………………112° C.
Water evaporated by 1 kg. coal………………………………………………………………..3.15 kg.
Corresponding heat utilized from 1 kg. coal in evaporation, about……………2,170 h. u.
Efficiency of the boiler, about…………………………………………………………….44 per cent.

As the above figures show, the conditions of operation with this type of boiler-construction were very favorable to a high efficiency. But the union of furnace- and boiler-masonry had a bad effect, causing, in the latter, cracks, dislocations and bulges, which increased the frequency of necessary repairs. Moreover, the longitudinal buckstays of the furnace left much to be desired.

Fig. 8 shows a Durr boiler, with 120 sq. m. of boiler heating-surface, and 24 sq. m. of heating-surface in the superheater, the whole supported, upon a separate frame of rolled iron, above a large ingot-heating furnace, the hearth-area of which is about 10 sq. m.

waste-heat-boiler

These boilers, which have been running satisfactorily for about six years, produce per hour per square meter of heating-surface about 11 kg. of steam at 280° C. The furnace-gases enter through two openings (a, a, Fig. 3) to the boiler, go in three passes over the water-tubes, and then escape into the atmosphere through a sheet-metal pipe placed between two boilers and carried by the boiler-masonry. Any necessary shutting-off of boiler from furnace is made possible by a fire-proof chimney-flue under the floor of the works, with which the openings a, a, are connected, and which is ordinarily closed by a slide, so as to permit the normal circulation of the gases to the boiler.

water-heat-boiler-durr-type-over-small-heating-furnace

Fig. 4 represents a Steinmuller boiler, with 82.5 sq. m. of heating-surface, and provision for auxiliary direct firing, carried upon a frame over a small heating-furnace. According to the builders, this boiler, heated with the waste-gases only, produces up to 15 kg. of steam per square meter of heating-surface, per hour. With the aid of the auxiliary grate, the product is from 20 to 25 kg. This combination of gas-heat with direct firing is no doubt due to the desire to get as much steam as possible out of the boiler, incidentally utilizing the heat of the waste-gases, and also diminishing the cost of the installalion per unit of steam-production. But the objection cannot be disregarded, that a separately located and separately heated boiler costs considerably more for attendance than one which constitutes a unit in a battery.

garbe waste-heat boiler with reverberatory furnace

The steeply inclined or vertical tubular boilers (Steilrohrkessel), so much used at present, are well adapted for the utilization of waste-gases from furnaces with working-doors on the side, because in such cases, the boiler can be built directly to the rear end of the furnace— the requirement of floor-space being very small. Fig. 5 shows a Garbe boiler with 120 sq. m. of heating-surface, attached to a reverberatory furnace, as built by the Dusseldorf-Ratinger tube-boiler works. The boiler shown in Fig. 5 was built without a superheater; but there would be no difficulty in adding one, as, indeed, the same builders have done in other cases.

To meet the most frequent requirement, namely, that the hook of the crane must be brought as near as possible to the furnace-door, so as to facilitate the handling of the heavy ingots, it seems advisable that the outer edge of the boiler over the furnace should at least not project beyond the furnace-armature. Since, for water-tube and fire-tube boilers, the surrounding masonry occupies a large part of the available width, attention is naturally drawn to the saving of space by the use of boilers which do not need masonry. Such tubular boilers approach the locomotive, locomobile, and marine types.

In locomotive boilers, the furnace-flue enters the fire-box from below, so that the heating-gases go directly into the fire-box, and thence into the boiling-tubes. In accordance with the nature of the locomotive boiler, the gases pass through it in one direction, and escape through a sheet-metal chimney, which branches from the smoke chamber above the firing-door. In this case, the boiler can be shut off from the furnace only when there is a second flue, leading from the main furnace-flue to a reserve chimney-flue. During the cleaning or repair of the boiler, this reserve conduit conveys the gases to a chimney, which may serve several furnaces, if necessary. But all this makes the design costly and complicated. Moreover, it must not be forgotten that the locomotive boiler, by reason of the great frictional resistance encountered by the gases in passing through the usually small heating-tubes, requires a strong chimney draft. It is therefore advisable, in order to avoid this evil, to choose heating- tubes of larger diameter, sacrificing thereby a few square meters of heating-surface.

For welding-furnaces with waste-gases of very high temperature, the locomotive type of boilers is not well adapted, since the pointed flame from the furnace strikes directly upon the tube-walls, and might easily make them leak. This type is likewise unsuitable where only a natural chimney draft is employed, because, when the furnace-doors are opened, the entrance of cold air tends to produce leakiness of the tube-ends.

Fig. 6 shows a boiler once designed by the present writer to be heated by furnace-gases. It is a variant of the marine boiler. A short vertical flue (usually 2 m. long) leads the gases from the furnace- flue into a corrugated tube, b, lying along one side in the boiler. The first 0.6 m. of this tube are lined with refractory material. The gases pass through the vertical flue a and the tube b, into a reversing chamber, c, in which a superheater, d, may be advantageously placed; and from this chamber they pass back through the boiler in the small boiling-tubes, e, which, like the corrugated tube, lie along one side. Finally they are received, at the end at which they entered, in a smoke-chamber f, of thin sheet-metal, and are conveyed at a greatly reduced temperature through the vertical flue g to the subterranean chimney-flue h. To make all parts of the boiler accessible, the superheating coils are placed in the reversing-chamber opposite the corrugated pipe, so that when it is necessary to tighten up the smoke- pipes, or clean them from soot, they can be easily reached from both sides after the corresponding doors have been opened. The inside of the corrugated tube itself is accessible through the door k, which has a peep-hole, i.

A direct connection of the main flue, a, with the flues g and h is established by breaking through the thin wall m separating a from g. When this is done, the furnace-gases will be drawn through g and h into the chimney, without touching the boiler at all. The introduction of valves, etc., for this purpose was avoided, because they would not have endured the high temperature which, at one point in the main flue, often melts the refractory lining. It should be added that, in later constructions of this type, the boiler rests upon a grating of beams, carried by six wrought-iron columns. Since the temperatures in the corrugated tubes are much higher than in the small tubes near it, care must of course be taken that the latter do not become loosened in the tube-walls through expansion and contraction. This consideration affects the design of the corrugated tube and the length adopted for the small tubes. Boilers of this type have been erected in considerable numbers at various works, and have run satisfactorily for several years.

Fig. 7 shows a modification of the type, in which the gases finally escape through a sheet-metal chimney over the boiler. A direct connection between furnace and chimney can be easily effected in this case, as in the preceding. It is only necessary to line both the smoke- chamber and the chimney with refractory material, and then to provide either a sliding cut-off or a removable separating wall, as shown in Fig. 6.

Experiments conducted upon a boiler about like that of Fig. 6, through 45 days (90 shifts) of the running of the heating-furnace on small forgings, gave the following results:

Grate-area of the heating-furnace……………………………………………………………..1.2 sq. m.
Hearth-area (width, 1.3 m.)………………………………………………………………………4.0 sq. m.
Distance from mid-hearth to beginning of fire-tubes…………………………………….6.5 m.
Length of connecting flue between furnace and fire-tubes……………………………..2.5 m.
Total length of furnace……………………………………………………………………………….6.2 m.
Total width of furnace………………………………………………………………………………..1.8 m.
Heating-surface of boiler……………………………………………………………………………61.0 sq. m.
Heating-surface of superheater in smoke-chamber………………………………………..5.6 sq. m.
Proportion of furnace grate-area to boiler heating-surface………………………………….1:51
Total length of boiler………………………………………………………………………………………5.25 m.
Diameter of boiler……………………………………………………………………………………………1.7 m.
Diameter of flame-tube………………………………………………………………………………..0.7 to 0.8 m.
Consumption of coal in 45 days (9 shifts)……………………………………………………….234 met. tons.
Consumption of coal per hour, about…………………………………………………………………217 kg.
Consumption of coal per hour per sq. m. grate-surface, about……………………………..181 kg.
Consumption of coal per hour per sq. m. boiler heating-surface, about……………….3.56 kg.
Amount of water evaporated during experiments……………………………………………594 met. tons.
Amount of water evaporated per hour…………………………………………………………………..550 kg.
Amount of water evaporated per hour per sq. m. boiler heating-surface, about…………….9.0 kg.
Amount of water evaporated by 1 kg. coal……………………………………………………………….2.54 kg.
Average steam-pressure……………………………………………………………………………………….5.7 atm.
Average temperature of superheated steam…………………………………………………………….310° C.
Superheating…………………………………………………………………………………………………………..148° C.
Temperature of boiler feed-water, about………………………………………………………………………20° C.
Quantity of heat transmitted to 1 kg. superheated steam, about……………………………………710 h. u.
Quantity of heat consumed in steam-production from combustion of 1 kg. coal…………….1,800 h. u.
Average heating-value of coal consumed………………………………………………………………………5,100 h. u.
Hence, proportion of heating-value utilized in making steam, about…………………………………35 per cent.
Temperature of waste-gases escaping from boiler, about…………………………………………………320° C.

In another experiment, lasting only through three shifts, an evaporation of 2.85 kg. water per kilogram of coal was reached. This would be a utilization of nearly 40 per cent. It should be observed that these experiments were made upon a boiler somewhat too small for the furnace, on which account the results are not particularly remarkable. Most of the boilers afterwards erected of this or similar type had about 80 sq. m. of boiler heating-surface, and from 10 to 12 sq. m. of heating-surface in the superheater, with which dimensions a steam-temperature of 400° C. was often attained, under favorable conditions of furnace-operation.

Among the chief advantages of this type of boiler are its small space-requirements, easy adaptation, and total freedom from masonry.

In case of a great fall in temperature but a comparatively small quantity of heat, which would not warrant the building of a boiler, the heat of the escaping gases can be utilized, if the locality be suitable, and the need sufficient, in a superheater, placed behind or (very easily) over the furnace. The steam-temperature can then be regulated in a highly simple, though not economical, fashion, by admit-

water-heat-boiler

waste-heat-boilers-2

ting air into the superheater through two holes in the masonry. This strongly cools the hot gases, and correspondingly lowers the steam-temperature.

As a general rule, however, when the heating surface has been properly determined, such special means of regulation are not required, since the quantity rather than the temperature of the chimney-gases is likely to vary.

That, moreover, the utilization of the hot gases in preheaters may be advisable even when their temperature is considerably below 600° C., is shown in Fig. 8, which represents such an arrangement, applied to the waste-gases from the boiler-plant, as well as from four sheet-heating furnaces. This plan works satisfactorily, and up to expectation.

An inquiry addressed to several works using hot waste-gases under boilers established the following:

preheaters-connected-with-sheet-heating-furnaces

Precise tests of efficiency had been made almost nowhere, so that only partial reports of the results of practice accumulated in the course of time were available. But these reports were more valuable for the expert operator than the figures obtained by any special tests. They showed in general, that with puddling-furnaces, from 2.5 to 3.5 kg. of steam per kilogram of coal were generated under boilers having from 40 to 80 sq. m. of heating-surface—the quantity of steam produced per square millimeter per hour being from 10 to 15 kg.

With the gases from welding-furnaces, boilers having up to 130 sq. m. of heating-surface produced from 7 to 12 kg. of steam per square meter per hour, representing from 3.5 to 5.5 kg. of steam per kilogram of coal. Of course these figures are subject to great variation, according to the amount of material heated in furnace per unit of time, the nature of this material (ingot-iron or wrought-iron), and its condition (hot or cold) when charged. Upon the production of steam per square meter of heating-surface per hour, the size of the boiler also has an effect.

In the operation of heating-furnaces for hammers, and especially also in furnaces heating thin sheets, it is advisable to introduce fresh air (preferably preheated) into the gases before or in the boiler, in order to effect the combustion of still unburned gas, and thus increase the steam-production.

A similar use of the hot waste-gases from zinc-furnaces has been introduced very recently, and promises to receive a rapid extension, the results so far having been very satisfactory, indicating an increase of from 15 to 20 per cent, in the utilization of the coal. In the cement- industry, the recently adopted revolving tube-furnaces permit the use of boilers, located behind them, for this purpose. The glass-manufacture has long utilized in the same way the waste-heat of its furnaces, even installing, for the preheating of water, apparatus which employs only the radiant heat of the furnace-arch.

The foregoing discussion shows that the utilization of the hot chimney-gases from directly-fired reverberatory furnaces is in very many cases practicable, and in not a few cases highly economical.

Utilization of Waste Heat from Regenerative Furnaces

This department of the subject is incomparably more important than the preceding, because the number and the size of the Siemens-Martin furnaces are so much greater than those of the heating-furnaces for rolling-mills and hammer-forges described above.

In regenerative furnaces, there is already a partial utilization of waste heat for preheating gas and air ; but its principal purpose is to raise the combustion-temperature in the hearth by raising the temperature of the entering gas and air; the recovery of heat thus effected is not very great, compared with the amount which still escapes unused; and it is therefore worth while to inquire whether a saving of further heat, which would otherwise escape to the chimney, can be effected without injury to the operation of the Martin furnaces.

The quantity of heat escaping at present, for the most part without hindrance, from these furnaces is very large—about 30 per cent, of the total heat employed (according to Professor Mayer’s experiments, 29 per cent.; according to Springorum, 32 per cent. These figures come from well-managed works, so that 30 per cent, as an average is not too high).

To these losses, due to the high temperature of the escaping gases, is to be added another, caused by radiation in the furnaces and chambers, which may be estimated at 40 per cent. From these figures we may deduce approximately the economic effectiveness of such furnaces.

Pfoser in Achern has attempted to utilize in part the radiant heat of a furnace arch, as shown in Fig. 9, which represents such an arrangement for preheating water, placed over the arch of a glass-furnace. This preheater has about 10 sq. m. of heating-surface, and consists of a coil of pipe, over the furnace-arch, through which the water passes. With glass-furnaces, where this method of preheating has been considerably used, it has been found to transfer about 300 to 500 h. u. per square meter of heating-surface. This cannot be called a very great success, but it makes a beginning; and it remains to be seen whether a better way of recovering the radiant heat may not soon be attained. Meanwhile, the placing of a coil full of water on

preheater-for-water-utilizing-the-radiated-heat-of-the-arch-of-a-glass-furnace

the top of a furnace-arch is rather a ticklish business; and many a manager would not risk it. Much easier and simpler is the recovery of heat from the waste-gases, as described below.

The average temperature of the escaping gases of the open-hearth, according to numerous determinations at various works, is about 700° C., or from 200° to 300° below that of the heating-furnace gases considered in the earlier parts of this paper. Hence we have to reckon with a somewhat smaller possible recovery, inasmuch as the transfer of heat through heating-surfaces will be smaller. Yet such a recovery is not only practicable but profitable, because the quantities of gas upon which it is carried out are almost always very large.

Prof. Dr. Mayer found in his experiments upon the heat-economy of the Martin furnace that daring the production of 1,000 kg. of steel 667,000 h. u. escape in the hot waste-gases. The investigations of Mackenzie on an entirely different and differently operated furnace gave a loss of 1,000,000 h. u. In both cases the temperature of the waste gases was 700° C.

Of course it would be impossible to recover in any way the whole of this heat. After treatment, the gases would always escape at last at some temperature or other ; and, moreover, every transfer of heat involves some loss.

To be perfectly safe, and to reach values which can be, with certainty, verified in practice, we shall use the somewhat unfavorable figures given by Mayer.

But first we must consider certain characteristic conditions of the Martin process itself, in order to avoid the great mistake of proposing any method which would react injuriously upon that process.

The first requirement is that for a given heating surface the boiler should be as small as practicable, in order, on one hand, to permit its installation where available space is limited, and, on the other hand, to reduce to a minimum the loss through radiation and conduction.

Furthermore, the construction and operation of the boiler should be, if possible, such that no interruption for the purpose of cleaning it will be required during the entire furnace-campaign of from 250 to 350 days. This requirement can be met by the use of pure water (surface-condenser water), and by such a construction as will permit outside cleaning of the boiler during operation.

The construction should also be such that occasional explosions, occurring when valves are reversed, will not injure the boiler. This condition is fulfilled by making the boiler-structure of a large number of single elements, with large open areas between them, so that the pressure from an explosion may pass through without doing harm. Moreover, these single elements must be able to resist occasional gas-puffs.

For this, as well as the first-named reason, boilers holding a large amount of water are out of the question.

Finally, the boiler, which is placed in the chimney-flue of the Martin furnace, must not present too great a resistance to the passage of the waste-gases. The reason for this requirement is not alone that the draft might be injured, and the chimney might prove inadequate; for the chimney does not play so very important a role when artificial draft is at hand. The chief reason is, that the pressure from a possible explosion should pass between the single members of the boiler-structure, without injury to the boiler.

The obvious requirements, that the boiler should not require too much attendance and maintenance; that safety of operation should be secured; and that the cost of installation should not be too great —need only be mentioned in passing.

All these requirements together can be met by a boiler, the elements of which are made of water-tubes; for, in the first place, such boilers combine large heating-surface with small bulk, and can be operated in a given case with small contents of water; secondly, the outside cleaning of such heating-surfaces is easily performed, if necessary, while the boiler is running; thirdly, the third and even the fourth requirement above stated can be satisfied by a suitable arrangement of the tubes; and fourthly and lastly, such heating-surfaces are comparatively inexpensive, and, when design and shopwork are good and solid, are highly safe against interior and exterior pressure.

In a boiler for the utilization of the hot chimney-gases of Martin furnaces in the production of steam, we distinguish the boiler proper, the superheater, and the preheater.

It is well to place the superheater before the boiler, or at least at the beginning of the draft, and the preheater behind the boiler. In calculating the necessary heating-surfaces, it seems best to begin with the preheater, then to calculate for the superheater, and to end with the calculation for the boiler. In my book, Die Abhitzkessel, I have given such calculations for a 30-ton Martin furnace, assuming an hourly steam-production of 2,000 kg. The several heating-surfaces were: preheater, 125 sq.m.; superheater (250° C. steam-temperature), 10 sq. m.; boiler, 200 sq. m.

But since, after utilization, the gases would be so cool that the chimney-draft could not overcome the increased resistances, it is self-evident that such utilization is practicable only when natural draft is not relied upon, but the cooled gases are moved by an exhaust-fan.

This would require in the case named a theoretical expenditure of 6.4 h-p. But in view of the comparatively low efficiency of fans of this size I have assumed for safety as required to draw off the chimney-gases, 40 h-p.—which would represent a fan efficiency of less than 20 per cent.

The calculation of profit from such an installation, costing 36,000 marks, is as follows :

Average steel production per hour………………………………………5,166 kg.
Steam-production from 340 sq. m. aggregate heating-surfaces of attached boiler, in terms of steam at 10 atm. and 250° C……………………………………………………………..2,000 kg.
Work of fan per hour…………………………………………………………40 h-p.
Cost of plant, including fan……………………………………………36,000 marks.

Assuming the works to contain six furnaces, of which five are constantly in operation, we shall have :

cost-of-the-boiler-plant

Since in the five units 10,000 kg. of steam are produced hourly, or 240,000 kg. daily, or, in round numbers, 77,000 metric tons per annum, the cost of this steam is about 1.20 marks per 1,000 kg.

If we now consider the work done with this steam through a steam-turbine, we find that at the rate (including all losses) of 7 kg. of steam per kilowatt-hour, the 10,000 kg. produced hourly will generate about 1,400 kw-hr., of which (150 kw-hr. being deducted for the driving of fans, etc.) 1,250 kw-hr. per hour, or 30,000 per day, or say 9,000,000 per year, remain available, at a cost of 40,000 marks for steam-generation and 40,000 for the turbine-plant, or 80,000 marks total annual cost. In other words this electrical energy is obtained for less than 1 pfennig per kilowatt-hour.

No other way of producing electrical energy can show a lower cost. Even water-power, favorably located and, unfailing in adequate supply, is seldom so cheap, if the amortization of the plant be provided for at about the rate assumed in the foregoing example.

It must be noted also, that the generation of electric energy in the works—that is, in the place where it is to be used—is independent of the season or weather. A given production of steel gives a corresponding, and (apart from minute variations) almost uniform quantity of energy in this form.

It is praiseworthy to try to make use of the water-powers of which so many are still neglected; but the falls of temperature which occur before our eyes deserve our attention likewise, since they are easier to appropriate, possess more constant strength, do not need long lines of transmission, and can therefore be made available at smaller expense of installation. Moreover, the utilization of these great temperature-falls in large steel-works recovers enormous quantities of heat, otherwise wasted.

Further results of calculation are graphically shown in Fig. 10.

Thus far, only the theoretical determination of the heating-surface areas and the probable costs and profits have been considered. The records of actual practice with waste-heat boilers attached to Martin furnaces are likewise interesting.

In the Duisburg steel-works of the Phoenix Co., live such boilers were erected last year behind Martin furnaces, and put in operation; and thorough tests were made under the direction of the Imperial

heating-surface and costing interest

German Steel Works. From the report of these tests by Chief Engineer J. Schreiber, which was published this year in Stahl und Eisen, I present here only the average figures of a series of experiments, lasting through 20 shifts, upon a boiler erected behind a 50-ton furnace (No. 4).

Average temperature of boiler feed-water, about……………………………………….20° C.
Average temperature of gases entering boiler, about…………………………………700° C.
Average temperature of gases leaving boiler, about…………………………………..350° C.
Temperature-fall of gases in boiler, about………………………………………………..350° C.
Average steam-pressure in boiler…………………………………………………………….7.5 atm.
Consumption of current in operating fan per hour………………………………………57 kw.
Product of steel during experiment…………………………………………………….2,000 met. tons.
Product of steel per hour……………………………………………………………………….8.3 met. tons.
Coal-consumption per ton of steel made………………………………………………………234 kg.
Evaporation per sq. m. heating-surface (without preheating)…………………………….7 kg.
Steam-production per ton of coal………………………………………………………………..1,842 kg.
Steel-production per ton of coal…………………………………………………………………….449 kg.
Heating-surface of boiler…………………………………………………………………………..500 sq. m.

graphically-comparison-with-cost

The great economy of such a utilization of waste heat appears very clearly from this statement, especially if it be considered that, even in this first undertaking of the kind, not the least injurious effect of the boiler upon the operation of the furnace was noticeable.

Finally, Fig. 11 exhibits graphically a comparison between the cost of 100 kg. of steam produced directly by coal combustion and the corresponding cost of production by means of the waste heat in Martin furnace-gases. In this diagram, the horizontal lines give the cost of 100 kg. of steam from direct combustion, according to the varying prices of coal, while the curve shows the cost of the same quantity of steam produced from waste-gases of different temperatures, as shown along the bottom line. The dotted horizontal line next above shows the approximate proportion of the total cost required in the use of directly-heated boilers to cover interest, amortization, and labor. The costs are given on the left, in marks and decimals, the interval being 0.20 mark.

The intersections of the horizontal lines by the curve mark the limit of economy in utilizing the waste-gases. It will be seen that this limit is extended in proportion as the temperature of the waste-gases or the price of coal is increased.