1911 Encyclopædia Britannica/Refrigerating and Ice-making

REFRIGERATING and ICE-MAKING. “Refrigeration” (from Lat. frigus, frost) is the cooling of a body by the transfer of a portion of its heat to another and therefore a cooler body. For ordinary temperatures it is performed directly with water as the cooling agent, especially when well water, which usually has a temperature of from 52° to 55° F., can be obtained. There are, however, an increasingly large number of cases in which temperatures below that of any available natural cooling agent are required, and in these it is necessary to resort to machines which are capable of producing the required cooling effect by taking in heat at low temperatures and rejecting it at temperatures somewhat above that of the natural cooling agent, which for obvious reasons is generally water. The function of a refrigerating machine, therefore, is to take in heat at a low temperature and reject it at a higher one.

This involves the expenditure of a quantity of work W, the amount in any particular case being found by the equation W = Q$2$ − Q$1$, where W is the work, expressed by its equivalent in British thermal units; Q$2$ the quantity of heat, also in B.Ther.U., given out at the higher temperature T$2$; and Q i the heat taken in at the lower temperature T$1$. It is evident that the discharged heat Q$2$ is equal to the abstracted heat Q ', plus the work expended, seeing that the work W, which causes the rise in temperature from T$1$ to T$2$, is the thermal equivalent of the energy actually expended in raising the temperature to the level at which it is rejected. The relation then between the work expended and the actual cooling work performed denotes the efficiency of the process, and this is expressed by Q$1$/(Q$2$-Q$1$); but as in a perfect refrigerating machine it is understood that the whole of the heat Q$1$ is taken in at the absolute temperature T$1$, and the whole of the heat Q$2$, is rejected at the absolute temperature T$2$, the heat quantities are proportional to the temperatures, and the expression T$1$/(T$2$ -T,) gives the ideal coefficient of performance for any stated temperature range, whatever working substance is used. These coefficients for a number of cases met with in practice are given in the following table.

They show that in all cases the heat abstracted exceeds by many times the heat expended. As an instance, when heat is taken in at 0° and rejected at 70°, a perfect refrigerating machine would abstract 6·6 times as much heat as the equivalent of the energy to be applied. If, however, the heat is to be rejected at 100°, then the coefficient is reduced to 4·6.

By examining Table I. it will be seen how important it is to reduce the temperature range as much as possible, in order to obtain the most economical results. No actual refrigerating machine does, in fact, take in heat at the exact temperature of the body to be cooled, and reject it at the exact temperature of the cooling water, but, for economy in working, it is of great importance that the differences should be as small as possible.

There are two distinct classes of machines used for refrigerating and ice-making. In the first refrigeration is produced by the expansion of atmospheric air, and in the second by the evaporation of a more or less volatile liquid.

Compressed-air Machines.—A compressed-air refrigerating machine consists in its simplest form of three essential parts—a compressor, a compressed-air cooler, and an expansion cylinder. It is shown diagrammatically in fig. 1 in connexion with a chamber which it is keeping cool. The compressor draws in air from the room and compresses it, the work expended in compression being almost entirely converted into heat. The compressed air, leaving the compressor at the temperature T$1$, passes through the cooler, where it is cooled by means of water, and is then admitted to the expansion cylinder, where it is expanded to atmospheric pressure, performing work on the piston. The heat equivalent of the mechanical work performed on the piston is abstracted from the air, which is discharged at the temperature T$2$.

. 1.—Compressed-Air Refrigerating Machine.

This temperature T$2$ is necessarily very much below the temperature to be maintained in the room, because the cooling effect is produced by transferring heat from the room or its contents to the air, which is thereby heated. The rise in temperature of the air is, in fact, the measure of the cooling effect produced. If such a machine could be constructed with reasonable mechanical efficiency to compress the air to a temperature but slightly above that of the cooling water, and to expand the air to a temperature but slightly below that required to be maintained in the room, we should of course get a result approximating in efficiency somewhat nearly to the figures given in Table I. Unfortunately, however, such results cannot be obtained in practice, because the extreme lightness of the air and its very small heat capacity (which at constant pressure is. 2379) would necessitate the employment of a great volume, with extremely large and mechanically inefficient cylinders and apparatus. A pound of air, representing about cub. ft., if raised 10° F. will only take up about 2·4 B.T.U. Consequently, to make such a machine mechanically successful a comparatively small weight of air must be used, and the temperature difference increased; in other words, the air must be discharged at a temperature very much below that to be maintained in the room.

This theory of working is founded on the Carnot cycle for a perfect heat motor, a perfect refrigerating machine being simply a reversed heat motor. Another theory involves the use of the Stirling regenerator, which was proposed in connexion with the Stirling heat engine (see AIR Engines). The air machine invented by Dr. A. Kirk in 1862, and described by him in a paper on the " Mechanical Production of Cold " (Proc. Inst. C.E., xxxvii., 18 74, 2 44), is simply a reversed Stirling air engine, the air working in a closed cycle instead of being actually discharged into the room to be cooled, as is the usual practice with ordinary compressed air machines. Kirk's machine was used commercially with success on a fairly large scale, chiefly for ice-making, and it is recorded that it produced about of ice for  of coal. In 1868 J. Davy Postle read a paper before the Royal Society of Victoria, suggesting the conveyance of meat on board ship in a frozen state by means of refrigerated air, and in 1869 he showed by experiment how it could be done; but his apparatus was not commercially developed. In 1877 a compressed-air machine was designed by J. J. Coleman of Glasgow, and in the early part of 1879 one of his machines was fitted on board the Anchor liner " Circassia," which successfully brought a cargo of chilled beef from America-the first imported by the aid of refrigerating machinery, ice having been previously used. The first successful cargo of frozen mutton from Australia was also brought by a Bell-Coleman machine in 1879. In the Bell-Coleman machine the air was cooled during compression by means of an injection of water, and further by being brought into contact with a shower of water. Another, perhaps the principal, feature was the interchanger, an apparatus whereby the compressed air was further cooled before expansion by means of the comparatively cold air from the room in its passage to the compressor, the same air being used over and over again. The object of this interchanger was not only to cool the compressed air before expansion, but to condense part of the moisture in it, so reducing the quantity of ice or snow produced during expansion. A full description of the machine may be found in a paper on “Air Refrigerating Machinery” by J. J. Coleman (Proc. Inst. C.E. lxviii., 1882). At the present time the Bell-Coleman machine has practically ceased to exist. In such compressed-air machines as are now made there is no injection of water during compression, and the compressed air is cooled in a surface cooler, not by actual mixture with a shower of cold water. Further, though the interchanger is still used by some makers, it has been found by experience that, with properly constructed valves and passages in the expansion cylinder, there is no trouble from the formation of snow, when, as is the general practice, the same air is used over and over again, the compressor taking its supply from the insulated room. So far as the air discharged from the expansion cylinder is concerned, its humidity is precisely the same so long as its temperature and pressure are the same, inasmuch as when discharged from the expansion cylinder it is always in a saturated condition for that temperature and pressure.

The ideal coefficient of performance is about 1, but the actual coefficient will be about 8 i after allowing for the losses incidental to working. In practice the air is compressed to about per square inch above the atmosphere, its temperature rising to about 300° F. The compressed air then passes through coolers in which it is cooled to within about 5° of the initial temperature of the cooling water, and is deprived of a portion of its moisture, after which it is admitted into the expansion cylinder and expanded nearly to atmospheric pressure. The thermal equivalent of the power exerted on the piston is taken from the air, which, with cooling water at 60° F. and after allowing for friction and other losses, is discharged at a temperature of 60° to 80° below zero F. according to the size of the machine. The pistons of the compression and expansion cylinders are connected to the same crankshaft, and the difference between the power expended in compression and that restored in expansion, plus the friction of the machine, is supplied by means of a steam engine coupled to the crankshaft, or by any other source of power. For marine purposes two complete machines are frequently mounted on one bed-plate and worked either together or separately.

In some machines used in the United States the cold air is not discharged into the rooms but is worked in a closed cycle, the rooms being cooled by means of overhead pipes through which the cold expanded air passes on its way back to the compressor.

Liquid Machines.—Machines of the second class may conveniently be divided into three types: (a) Those in which there is no recovery of the refrigerating agent, water being the agent employed; they will be dealt with as "Vacuum machines." (b) Those in which the agent is recovered by means of mechanical compression; they are termed " Compression machines." (c) Those in which the agent is recovered by means of absorption by a liquid; they are known as "Absorption machines." In the first class, since the refrigerating liquid is itself rejected, the only agent cheap enough to be employed is water. The boiling point of water varies with pressure; thus at one atmosphere or 14·7 lb per square inch it is 212° F., whereas at a pressure of per square inch it is 32°, and at lower pressures there is a still further fall in temperature. This property is made use of in vacuum machines. Water at ordinary temperature, say 60°, is placed in an air-tight glass or insulated vessel, and when the pressure is reduced by means of a vacuum pump it begins to boil, the heat necessary for evaporation being taken from the water itself. The pressure being still further reduced, the temperature is gradually lowered until the freezing-point is reached and ice formed, when about one-sixth of the original volume has been evaporated.

The earliest machine of this kind appears to have been made in 1755 by Dr. William Cullen, who produced the vacuum by means of a pump alone. In 1810 Sir John Leslie combined with the air pump a vessel containing strong sulphuric acid for absorbing the vapour from the air, and is said to have succeeded in producing 1 to of ice in a single operation. E. C. Carré later adopted the same principle. In 1878 F. Windhausen patented a vacuum machine for producing ice in large quantities, and in 1881 one of these machines, said to be capable of making about 12 tons of ice per day, was put to work in London. The installation was fully described by Carl Pieper (Trans. Soc. of Engineers, 1882, p. 145) and by Dr. John Hopkinson (Journal of Soc. of Arts, 1882, vol. xxxi. p. 20). The process, however, not being successful from a commercial point of view, was abandoned. At the present time vacuum machines are only employed for domestic purposes. The hand apparatus invented by H. A. Fleuss consists of a vacuum pump capable of reducing the air pressure to a fraction of a millimetre, the suction pipe of which is connected first with a vessel containing sulphuric acid, and second with the vessel containing the water to be frozen. Both these vessels are mounted on a rocking base, so that the acid can be thoroughly agitated while the machine is being worked. As soon as the pump has sufficiently exhausted the air from the vessel containing the water, vapour is rapidly given off and is absorbed by the acid until sufficient heat has been abstracted to bring about the desired reduction in temperature, the acid becoming heated by the absorption of water vapour, while the water freezes. The small Fleuss machine will produce about of ice in one operation of 20 minutes. Iced water in a carafe for drinking purposes can be produced in about three minutes. The acid vessel holds of acid, and nearly  of ice can be made for each  of acid before the acid has become too weak to do further duty. Another machine, which can be easily worked by a boy, will produce 20 to of ice in one hour, and is perhaps the largest size practicable with this method of freezing. The temperature attainable depends on the strength and condition of the sulphuric acid; ordinarily it can be reduced to zero F., and temperatures 20° lower have frequently been obtained.

Though prior to 1834 several suggestions had been made with regard to the production of ice and the cooling of liquids by the evaporation of a more volatile liquid than water, the first machine actually constructed and put to work was made by John Hague in that year from the designs of Jacob Perkins (Journal of Soc. of Arts, 1882, vol. xxxi. p. 77). This machine, though never used commercially, is the parent of all modern compression machines. Perkins in his patent specification states that the volatile fluid is by preference ether. In 1856 and 1857 James Harrison of Geelong, Victoria, patented a machine embodying the same principle as that of Perkins, but worked out in a much more complete and practical manner. It is stated that these machines were first made in New South Wales in 1859, but the first Harrison machine adopted successfully for industrial purposes in England was applied in the year 1861 for cooling oil in order to extract the paraffin. In Harrison's machine the agent used was ether (C 2 H 5) 2 O. Improvements were made by Siebe & Company of London, and a considerable number of ether machines both for ice-making and refrigerating purposes were supplied by that firm and others up to the year 1880. In 1870 the subject of refrigeration was investigated by Professor Carl Linde of Munich, who was the first to consider the question from a thermodynamic point of view. He dealt with the coefficient of performance as a common basis of comparison for all machines, and showed that the compression vapour machine more nearly reached the theoretic maximum than any other (Bayerisches Industrie and Gewerbeblatt, 1870 and 1871). Linde also examined the physical properties of various liquids, and, after making trials with methylic ether in 1872, built his first ammonia compression machine in 1873. Since then the ammonia compression machine has been most widely adopted, though the carbonic acid machine, also compression, which was first made in 1880 from Linde's designs, is now used to a considerable extent, especially on board ship.

Regulating Valve

. 2. - Vapour Compression Machine.

A diagram of a vapour compression machine is shown in fig. 2.. There are three principal parts, a refrigerator or evaporator, a compression pump, and a condenser. The refrigerator, which, Refrigerator Condenser T, Compressor consists of a coil or series of coils, is connected to the suction side of the pump, and the delivery from the pump is connected to the condenser, which is generally of somewhat similar construction to the refrigerator. The condenser and refrigerator are connected by a pipe in which is a valve named the regulator. Outside the refrigerator coils is the air, brine or other substance to be cooled, and outside the condenser is the cooling medium, which, as previously stated, is generally water. The refrigerating liquid (ether, sulphur dioxide, anhydrous ammonia, or carbonic acid) passes from the bottom of the condenser through the regulating valve into the refrigerator in a continuous stream. The pressure in the refrigerator being reduced by the pump and maintained at such a degree as to give the required boiling-point, which is of course always lower than the temperature outside the coils, heat passes from the substance outside, through the coil surfaces, and is taken up by the entering liquid, which is converted into vapour at the temperature T i. The vapours thus generated are drawn into the pump, compressed, and discharged into the condenser at the temperature T2, which is somewhat above that of the cooling water. Heat is transferred from the compressed vapour to the cooling water and the vapour is converted into a liquid, which collects at the bottom and returns by the regulating valve into the refrigerator. As heat is both taken in and discharged at constant temperature during the change in physical state of the agent, a vapour compression machine must approach the ideal much more nearly than a compressed-air machine, in which there is no such change.

This will be seen by taking as an example a case in which the cold room is to be kept at 10° F., the cooling water being at 60°. Under these conditions, the actual evaporating temperature T 1, in a well-constructed ammonia compression machine, after allowing for the differences necessary for the exchange of heat, would be about 5° below zero, and the discharge temperature T would be about 75°. An ideal machine, working between 5° below zero and 75° above, has a coefficient of about 5·7, or nearly six times that of an ideal compressed-air machine of usual construction performing the same useful cooling work.

A vapour compression machine does not, however, work precisely in the reversed Carnot cycle, inasmuch as the fall in temperature between the condenser and the refrigerator is not produced, nor is it attempted to be produced, by the adiabatic expansion of the agent, but results from the evaporation of a portion of the liquid itself. In other words, the liquid-refrigerating agent enters the refrigerator at the condenser temperature and introduces heat which has to be taken up by the evaporating liquid before any useful refrigerating effect can be performed. The extent of this loss is determined by the relation between the liquid heat and the latent heat of vaporization at the refrigerator temperature. If r represents the latent heat of the vapour, and q$1$ and q1 the amounts of heat contained in the liquid at the respective temperatures of T$1$ and T$2$, then the loss from the heat carried from the condenser into the refrigerator is shown by (q$2$−q$1$)/r and the useful refrigerating effect produced in the refrigerator is r−(q$2$−q$1$). Assuming, as in the previous example, that T$2$ is 75° F., and that T$1$ is 5° below zero, the results for various refrigerating agents are as follows:—

The results show that the loss is least in the case of anhydrous ammonia and greatest in the case of carbonic acid. At higher condenser temperatures the results are even much more favourable to ammonia. As the critical temperature (88·4° F.) of carbonic acid is approached, the value of r becomes less and less and the refrigerating effect is much reduced. When the critical point is reached the value of r disappears altogether, and a carbonic-acid machine is then dependent for its refrigerating effect on the reduction in temperature produced by the internal work performed in expanding the gaseous carbonic acid from the condenser pressure to that in the refrigerator. The abstraction of heat does not then take place at constant temperature. The expanded vapour enters the refrigerator at a temperature below that of the substance to be cooled, and whatever cooling effect is produced is brought about by the superheating of the vapour, the result being that above the critical point of carbonic acid the difference T2-T2 is increased and the efficiency of the machine is reduced. The critical temperature of anhydrous ammonia is about 266° F., which is never approached in the ordinary working of refrigerating machines. Some of the principal physical properties of sulphurous acid, anhydrous ammonia, and carbonic cold are given in

The action of a vapour compression machine is shown in fig. 3. Liquid at the condenser temperature being introduced into the refrigerator through the regulating valve, a small portion evaporates and reduces the remaining liquid to the temperature T$2$. This is shown by the curve AB, and is the useless work represented by the expression (q$1$−q$2$)/r. Evaporation then continues at the constant temperature T, abstracting heat from the substance outside the refrigerator as shown by the line BC. The vapour is then compressed along the line CD to the temperature T$1$, when, by the action of the cooling water in the condenser, heat is abstracted at constant temperature and the vapour condensed along the line DA.

In a compression machine the refrigerator is usually a series of iron or steel coils surrounded by the air. brine or other substance it is desired to cool. One end (generally the bottom) of the coils is connected to the liquid pipe from the condenser and the other end o to the suction of the compressor.

FIG. 3. - Action of Vapour Compression Machine.

A Liquid from the condenser is admitted to the coils through an adjustable regulating valve, and by taking heat from the substance outside is evaporated, the vapour being continually drawn off by the compressor and discharged under increased pressure into the condenser. The condenser is constructed of coils like the refrigerator, the cooling water being contained in a tank; frequently, however, a series of open coils is employed, the cooling water falling over the coils into a collecting tray below, and this form is perhaps the most convenient for ordinary use as it affords great facilities for inspection and painting. The compressor may be driven by a steam engine or in any other convenient manner. The pressure in the condenser varies according to the temperature of the cooling water, and that in the refrigerator is dependent upon the temperature to which the outside substance is cooled. In an ammonia machine copper and copper alloys must be avoided, but for carbonic acid they are not objectionable.

The compression of ammonia is sometimes carried out on what is known as the Linde or " vet " system, and sometimes on the " dry " system. When wet compression is used the regulating valve is opened to such an extent that a little more liquid is passed than can be evaporated in the refrigerator. This liquid enters the compressor with the vapour, and is evaporated there, the heat taken up preventing the rise in temperature during compression which would otherwise take place. The compressed vapour is discharged at a temperature but little above that of the cooling water. With dry compression, vapour alone is drawn into the compressor, and the temperature rises to as much as 180 or 200 degrees. Wet compression theoretically is not quite so efficient as dry compression, but it possesses practical advantages in keeping the working parts of the compressor cool, and it also greatly facilitates the regulation of the liquid, and ensures the full duty of the machine being continuously performed. Very exact comparative trials have been made by Professor M. Schroeter and others with compression machines using sulphur dioxide and ammonia. The results are published in Vergleichende Versuche an Kdltemaschinen, by Schroeter, Munich, 1890, and in Nos. 32 and 51 of Bayerisches Industrie and Gewerbeblatt, 1892. Some of the results obtained by Schroeter in 1893 with an ordinary brine cooling machine on the Linde ammonia system are given in Table VI.: -

The principle of the absorption process is chemical or physical rather than mechanical; it depends on the fact that many Absorp- vapours of low boiling-point are readily absorbed in water, and can be separated again by the application of heat. In its simplest form an absorption machine consists of two iron vessels connected together by a bent pipe. One of these contains a mixture of ammonia and water, which on the application of heat gives off a mixed vapour containing a large proportion of ammonia, a liquid containing but little ammonia being left behind. In the second vessel, which is placed in cold water, the vapour rich in ammonia is condensed under pressure. To produce refrigeration the operation is reversed. On allowing the weak liquor to cool to normal temperature, it becomes greedy of ammonia (at 60 F. at atmospheric pressure water will absorb about 760 times its own volume of ammonia vapour), and this produces an evaporation from the liquid in the vessel previously used as a condenser. This liquid, containing a large proportion of ammonia, gives off vapour at a low temperature, and therefore becomes a refrigerator abstracting heat from water or any surrounding body. When the ammonia is evaporated the operation as described must be again commenced. Such an apparatus is not much used now. Larger and more elaborate machines were made by F. P. E. Carre in France; but no very high degree of perfection was arrived at, owing to the impossibility of getting an anhydrous product of distillation. In 1867 Rees Reece, taking advantage of the fact that two vapours of different boiling-points, when mixed, can be separated by means of fractional condensation, brought out an absorption machine in which the distillate was very nearly anhydrous. By means of vessels termed the analyser and the rectifier, the bulk of the water was condensed at a comparatively high temperature and run back to the generator, while the ammonia passed into a condenser, and there assumed the liquid form under the pressure produced by the heat in the generator and the cooling action of water circulating outside the condenser tubes.


 * . 4.

Fig. 4 is a diagram of an absorption apparatus. The ammonia vapour given off in the refrigerator is absorbed by a cold weak solution of ammonia and water in the absorber, and the strong liquor is pumped back into the generator through an interchanger through which also the weak hot liquor from the generator passes on its way to the absorber. In this way the strong liquor is heated before it enters the generator, and the weak liquor is cooled Generator before it enters the absorber. The generator being heated by means of a steam coil, ammonia vapour is driven off at such a pressure as to cause its condensation in the condenser. From the con denser it passes into the refrigerator through a regulating valve in the usual manner. The process is continuous, and is identical with that of the compression machine, with the exception of the return from the temperature T 1 to the temperature T2, which is brought about by the direct application of heat instead of by means of mechanical compression. With the same temperature range, however, the same amount of heat has to be acquired in both cases, though from the nature of the process the actual amount of heat demanded from the steam is much greater in the absorption system than in the compression. This is chiefly due to the fact that in the former the heat of vaporization acquired in the refrigerator is rejected in the absorber, so that the whole heat of vaporization has to be supplied again by the steam in the generator. In the latter the vapour passes direct from the refrigerator to the pump, and power has to be expended merely in raising the temperature to a sufficient degree to enable condensation to occur at the temperature of the cooling water. On the other hand, a great advantage is gained in the absorption machine by using the direct heat of the steam, without first converting it into mechanical work, for in this way its latent heat of vaporization can be utilized by condensing the steam in the coils and letting it escape in the form of water. Each pound of steam can thus be made to give up some 950 units of heat; while in a good steam engine only about 200 units are utilized in the steam cylinder per pound of steam, and in addition allowance has to be made for mechanical inefficiency. In the absorption machine the cooling water has to take up about twice as much heat as in the compression system, owing to the ammonia being twice liquefied - namely, once in the absorber and once in the condenser. It is usual to pass the cooling water first through the condenser and then through the absorber.

The absorption machine is not so economical as the compression; but an actual comparison between the two systems is difficult to make. Information on this head is given in papers read by Dr. Linde and by Professor J. A. Ewing before the Society of Arts (Journal of the Society of Arts, vol. xlii., 18 94, p. 322, and Howard Lectures, January, February and March 1897).

An absorption apparatus as applied to the cooling of liquids consist s s of a generator containing coils to which steam is supplied at suitable pressure, an analyser, a rectifier, a condenser either of the submerged or open type, a refrigerator in which the nearly anhydrous ammonia obtained in the condenser is allowed to evaporate, an absorber through which the weak liquor from the generator continually flows and absorbs the anhydrous vapour produced in the refrigerator, and a pump for forcing the strong liquor produced in the absorber back through an economizer into the analyser where, meeting with steam from the generator, the ammonia gas is again driven off, the process being thus carried on continuously. Sometimes an additional vessel is employed for heating liquor by means of the exhaust steam from the engine driving the ammonia pump. Absorption machines are also made without a pump for returning the strong liquor to the generator. In these cases they work intermittently. In some machines the same vessel is used alternately as a generator and absorber, while in others, in order to minimize the loss of time, two vessels are provided which can be used alternately as generators and absorbers.

Applications.—Apart from the economical working of the machine itself, whatever system may be adopted, it is of importance that cold once produced should not be wasted, and it is therefore necessary to use some form of insulation to protect the vessels in which liquids are being cooled, or the rooms of ships' holds in which the freezing or storage processes are being carried on. This insulation generally consists of materials such as charcoal, silicate cotton, granulated cork, small pumice, hair-felt, sawdust, &c., held between layers of wood or brick, and forming a more or less heat-tight box. There is no recognized standard of insulation. For a cold store to be erected inside a brick or stone building, and to be maintained at an internal temperature of from 18° to 20° F., a usual plan is shown in fig. 5. The same insulation is used for the floors and


 * . 5.—Insulation of a Cold Store.

ceilings, except that the wearing surface of the floor is generally made thicker than the inside lining of the sides. Should the walls or floor be damp, waterproof paper is added. Granulated cork has practically the same insulating properties as silicate cotton, and the same thicknesses may be used. About 10 in. of flake charcoal and vegetable silica, or 11 of small pumice, are required to give the same protection as 7 in. of good silicate cotton. Cork bricks made of compressed granulated cork are frequently used, a thickness of about 5 in. giving the same protection as 7 in. of silicate cotton. The walls and ceilings are finished off with a smooth coating of hard cement and the floors are protected by cement or asphalt, according to the nature of the traffic on them. For lager-beer cellars and fermenting rooms, for bacon-curing cellars, and for similar purposes, brick walls with single or double air spaces are used, and sometimes a space filled with silicate cotton or other insulating material. In Australia and New Zealand pumice, which is found in enormous quantities in the latter country, takes the place of charcoal and silicate cotton. In Canada air spaces are largely used either alone or in combination with silicate cotton or planer shavings. The air spaces, two or three in number, are formed between two layers of tongued and grooved wood, and the total thickness of the insulation is about the same as when silicate cotton alone is used. On board ship charcoal has been almost entirely employed, but silicate cotton and granulated cork are sometimes used. The material is either placed directly up to the skin of the vessel, and kept in place by a double lining of wood inside, in which case a thickness of about To in. is used depending upon the depth of the frames, or it is placed between two layers of wood, with an air space next the skin, in which case about 6 in. of flake charcoal is generally sufficient for the insulation of the holds, though for deck-houses and other parts exposed to the sun the thickness must be greater. A layer of sheet zinc or tin has frequently to be used as protection from rats. Given a certain allowable heat transmission, the principal points to be considered in connexion with insulation are, first cost, durability, weight and space occupied, the two last named being specially important factors on board ship. No exact rules can be laid down, as the conditions vary so greatly; and though experiments have been made to determine the actual heat conduction of various materials per unit of surface, thickness and temperature difference, the experience of actual practice is at present the only accepted guide.

With compressed-air machines which discharge the cold air direct into the insulated room or hold, a snow box is provided close to the outlet of the expansion cylinder to catch the snow and congealed oil. The air is distributed by means of wood air trunks with openings controlled by slides, and similar trunks are pro vided in connexion with the suction of the compresser to conduct the air back to the machine. With liquid machines of the compression and absorption system, the rooms are either cooled by means of cold pipes or surfaces placed in them, or by a circulation of air cooled in an apparatus separated from the rooms. The cold pipes may be direct-expansion pipes in which the liquid evaporates, or they may be pipes or walls through which circulates an uncongealable brine previously cooled to the desired temperature. The pipes are placed on the ceilings or sides according to circumstances, but they must be arranged so as to induce a circulation of air throughout the compartment and ensure every part being cooled. With what is termed the air circulation system the air is generally circulated by means of a fan, being drawn from the rooms through ducts, passed over a cooler, and returned again to the rooms by other ducts. In some coolers the cooling surfaces consist of direct-expansion pipes placed in clusters of convenient form; in others brine pipes are used; in others there is a shower of cold brine, and in some cases combinations of cold pipes and brine showers. Whether pipes in the rooms or air circulation give the best results is to some extent a matter of opinion, but at the present time the tendency is decidedly in favour of air circulation, at any rate for general cold storage purposes. Whichever system be adopted, it is important for economical reasons that ample cooling surface be allowed, and that all surfaces be kept clean and active, to make the difference between the temperature of the evaporating liquid and the rooms as small as possible. Small surfaces reduce first cost, but involve higher working expenses by decreasing the value of T$2$/(T$1$−T$2$), and thus demanding more energy, and consequently more fuel, to effect the given result than if larger surfaces were employed.

. 6.—General Arrangement of an Ice Factory.

The general arrangement of an ice factory for producing can ice is shown in fig. 6. The water to be frozen is contained in galvanized or terned steel moulds suspended in a tank filled to the proper level with brine maintained at the desired temperature. The moulds are frequently arranged in frames, so that by means of an overhead crane one complete row is lifted at a time. When the water is frozen the moulds are dipped in a tank containing warm water, and on being tipped the blocks of ice fall out. Ordinary water contains air, and ice made from it is generally opaque, due to the inclusion of numerous small air-bubbles. To produce clear ice the water must be agitated during the freezing process, or previously boiled to get rid of the air. Distilled water is frequently used, as well as the water produced by the condensation of the steam from the engine, which of course must be thoroughly purified and filtered. It should be noted, however, that with an icemaking plant of moderate size and a steam-engine of good construction the weight of steam used will not neatly equal the weight of ice produced, so that the difference must be made up either by distillation, which is a costly process, or by ordinary water. Can ice is usually made in blocks weighing 56, 112 or, and from 4 to 8 in. thick. For cell ice ordinary water is used, agitated during freezing. The cells are flat and constructed of galvanized iron, so as to form a hollow space of about 2 in. in width, through which cold brine is circulated by a pump. They are placed vertically in a tank, the distance between them being from 8 to 14 in., according to the thickness of the ice to be produced. The tank is filled with water, which is kept in agitation by means of a reciprocating paddle or piston; in this way the air escapes, and with proper care a block of great transparency is produced. To thaw it off, warm brine is circulated through the cells. A usual size for cell ice is 4 ft. by 3 ft. by 1 ft. mean thickness, the weight being about 6 cwt. If perfectly transparent ice is required, the two sides of the block are not allowed to join up, and it is then called plate ice, which is often made in very large blocks, afterwards divided by saws or steam cutters. In such cases the evaporation of the ammonia or other refrigerating liquid frequently takes place in the cells themselves, brine being dispensed with. With a well-constructed can ice-plant of say 25 tons capacity per day, from 15 to 16 tons of ice should be made in Great Britain to a ton of best steam coal. For cell and plate ice the production is considerably below this, and the first cost of the plant is much greater than that for can ice.

. 7.—Cold Stores. Fig. 7 shows an arrangement of cold storage on land, refrigerated on the air circulation system. The insulated rooms, on two floors, are approached by corridors, so as to exclude external air, which if allowed to enter would deposit moisture upon the cold goods. The air cooler is placed at the end, and the air is distributed by means of wood ducts furnished with slides for regulating the temperature of the rooms, which are insulated according to the method shown in fig. 5. In some cases, instead of the entrance being at the sides or ends, it is at the top, all goods being raised to the top floor in lifts and lowered by lifts into the rooms. With good machinery the cost of raising is not great, and is probably equalled by the saving in refrigeration, since the rooms hold the heavy cold air as a glass holds water.

Large passenger vessels and yachts are now generally fitted with refrigerating machinery for preserving provisions, cooling water and wine, and making ice. Usually two insulated compartments are provided, one for frozen meats at about 20° F., and one for vegetables, &c., at about 40°. They have a capacity of from 1500 to 3000 cub. ft. or more, according to the number of passengers carried, and they are generally cooled by means of brine pipes, though direct expansion and air circulation are sometimes adopted. A passenger vessel requires from 2 to 4 cwt. of ice per day. On battleships and cruisers the British Admiralty use small compressed-air machines for ice-making, and larger machines, generally on the carbonic-acid system, for cooling the magazines. A modern frozen meat-carrying vessel will accommodate as much as 120,000 carcases, partly sheep and partly lambs, requiring a hold capacity of about 300,000 cub. ft. In some vessels both fore and aft holds and 'tween decks are insulated. Lloyd's Committee now issue certificates for refrigerating installations, if constructed according to their rules, and most modern cargo-carrying vessels have their refrigerating machinery classed at Lloyd's. In the meat trade between the River Plate, the United States, Canada and Great Britain, ammonia or carbonic acid machines are now exclusively used, but for the Australian and New Zealand frozen meat trade compressed-air machines are still employed to a small extent. The holds of meat-carrying vessels are refrigerated either by cold air circulation or by brine pipes.

Though the adoption of refrigerating and ice-making machinery for industrial purposes practically dates from the year 1880, the manufacture of these machines has already assumed very great proportions; indeed, in no branch of mechanical engineering, with the exception of electrical machinery, has there been so remarkable a development in recent years. The sphere of application is extending year by year. The cooling of residential and public buildings in hot countries, though attempted in a few cases in the United States and elsewhere, is yet practically untouched, the manufacture of ice and the preservation of perishable foods (apart from the frozen and chilled meat trades) have in many countries hardly received serious consideration, but in breweries, dairies, margarine works and many other industries there is a large and increasing field for refrigerating and ice-making machinery. A recent application is in the cooling and drying of the air blast for blast furnaces. Though this matter had been discussed for some years, it was only in 1904 that the first plant was put to work at Pittsburg.

For further information reference may be made to the following: Siebel, Compend. of Mechanical Refrigeration (Chicago); Redwood, Theoretical and Practical Ammonia Refrigeration (New York); Stephansky, Practical Running of an Ice and Refrigerating Plant (Boston); Ledoux, Ice-Making Machines (New York); Wallis-Taylor, Refrigerating and Ice-Making Machines (London) Ritchie Leask, Refrigerating Machinery (London); De Volson Wood, Thermodynamics, Heat Motors and Refrigerating Machinery (New York); Linde, Kdlteerzeugungsmaschine Lexikon der gesamten Technik; Behrend, Eis and Kälteerzeugungsmaschinen (Halle); De Marchena, Kompressions Kältemaschinen (Halle); Theodore Koller, Die Kälteindustrie (Vienna); Voorhees, Indicating the Refrigerating Machine (Chicago); Norman Selfe, Machinery for Refrigeration (Chicago); Hans Lorenz, Modern Refrigerating Machinery (London); Lehnert, Moderne Kältetechnik (Leipzig); L. Marchis, Production et utilisation du froid (Paris); C. Heinel, Bau and Betrieb von Kältemaschinen Anlagen (Oldenburg); R. Stetefeld, Eis and Kalteerzeugungs-Maschinen (Stuttgart).