When steam engines were first introduced, they were commonly applied to work pumps or mills which had been previously wrought by horses. It was, therefore, convenient, and indeed necessary, in the first instance, to be able to express the performance of these machines by reference to the effects of animal power, to which manufacturers, miners, and others, had been long accustomed. When an engine, therefore, was capable of performing the same work in a given time, as any given number of horses of average strength usually performed, it was said to be an engine of so many horses' power. This term was long used with much vagueness and uncertainty: at length, as the use of steam engines became more extended, it was apparent that confusion and inconvenience would ensue, if some fixed and definite meaning were not assigned to it, so that the engineers and others should clearly understand each other in expressing the powers of these machines. The term horse-power had so long been in use, that it was obviously convenient to retain it. It was only necessary to agree upon some standard by which it might be defined. The performance of a horse of average strength, working for eight hours a day, was, therefore, selected as a standard or unit of steam-engine power. Smeaton estimated the amount of mechanical effect which the animal could produce at 22,916 pounds, raised one foot per minute; Desaguiliers makes it 27,500 pounds, raised through the same height. Messrs. Bolton and Watt caused experiments to be made with the strong horses used in the breweries in London; and from the result of these they assigned 33,000 pounds raised one foot per minute, as the value of a horse's power: this is, accordingly, the estimate now generally adopted; and, when an engine is said to be of so many horses' power, it is meant that, when in good working order and properly managed, it is capable of overcoming a resistance equivalent to so many times 33,000 pounds raised one foot per minute. Thus, an engine of ten-horse-power would be capable of raising 330,000 pounds one foot per minute.
As the same quantity of water converted into steam will always produce the same mechanical effect, whatever be the density of the steam produced from it, and at whatever rate the evaporation may proceed, it is evident that the power of a steam engine will depend on two circumstances: first, the rate at which the boiler with its appendages is capable of evaporating water; and, secondly, the rate at which the engine is capable of consuming the steam by its work. We shall consider these two circumstances separately.
The rate at which the boiler produces steam will depend upon the rate at which heat can be transmitted from the fire to the water which it contains. Now this heat is transmitted in two ways: either by the direct action of the fire radiating heat against the surface of the boiler; or by the flame, and heated air which escapes from the fire, passing through the flues, as already explained. The surface of the boiler exposed to the direct radiation of the fire is technically called fire surface; and that which takes heat from the flame and air, on its way to the chimney, is called flue surface. Of these the most efficient in the generation of steam is the former. In stationary boilers, used for condensing engines, where magnitude and weight are matters of little importance, it has been found that the greatest effect has been produced in general by allowing four and a half square feet of fire surface, and four and a half square feet of flue surface, for every horse-power. By means of this quantity of fire and flue surface, a cubic foot of water per hour may be evaporated.
It has been already shown that the total power exerted by a cubic inch of water, converted into steam, will be equivalent to 2160 pounds raised one foot. A cubic foot of water consists of 1728 cubic inches, and the power produced by its evaporation will therefore be found by multiplying 2160 by 1728; the product, 3,732,480, expresses the number of pounds' weight which the evaporation of a cubic foot of water would raise one foot high, supposing that its entire mechanical force were rendered available: but to suppose this in practice, would be to suppose the machine, through the medium of which it is worked, moved without any power being expended upon its own parts. It would be, in fact, supposing all its moving parts to be free from friction and other causes of resistance. To form a practical estimate, then, of the real quantity of available mechanical power obtained from the evaporation of a given quantity of water, it will be necessary to inquire what quantity of this power is intercepted by the engine through which it is transmitted. In different forms of steam engine—indeed, we may say in every individual steam engine—the amount thus lost is different; nevertheless, an approximate estimate may be obtained, sufficiently exact to form the basis of a general conclusion.
Let us consider, then, severally, the means by which mechanical power is intercepted by the engine.
First, The steam must flow from the boiler into the cylinder to work the piston; it passes necessarily through pipes more or less contracted, and is, therefore, subject to friction as well as cooling in its passage.
Second, Force is lost by the radiation of heat from the cylinder and its appendages.
Third, The friction of the piston in the cylinder must be overcome.
Fourth, Loss of steam takes place by leakage.
Fifth, Force is expended in expelling the steam after having worked the piston.