cost of air compression in any given case. Diameters of from 4 to 6 in. for the mains are large enough for any ordinary mining practice. Up to a length of 3,000 ft., a 4-in. pipe will carry 480 cu. ft. per min. of free air compressed to 82 lb., with a loss of 2 lb. pressure. This volume of air will run four 3-in. drills. Under the same conditions, a 6-in. pipe, 5,000 ft. long, will carry 1,100 cu. ft. per min. of free air, or enough for 10 drills.

A mistake is often made by putting in branch pipes of too small a diameter. For a distance of, say, 100 ft., a 1-in. pipe is small enough for a single drill, though a 1-in. pipe is frequently used. While it is, of course, admissible to increase the velocity of flow in short branches considerably beyond 20 ft. per sec., extremes should be avoided. To run a 3-in. drill from a 1-in. pipe 100 ft. long, will require a velocity of flow of about 55 ft. per sec., causing a loss of 10 lb. pressure.

The piping for conveying compressed air may be of cast or wrought iron. If of wrought iron, as is customary, the lengths are connected either by sleeve couplings or by cast-iron flanges into which the ends of the pipe are screwed or expanded. Sleeve couplings are used for all except the large sizes. The smaller sizes, up to 1 in., are butt-welded, while all from 1 in. up are lapwelded, to insure the necessary strength. Wrought-iron, spiral-seam, riveted, or spiral-weld steel tubing is sometimes used. It is made in lengths of 20 ft. or less. For convenience of transport in remote regions, rolled sheets in short lengths may be had. They are punched around the edges, ready for riveting, and are packed closely-four, six, or more sheets in a bundle.

All joints in air mains and branches should be carefully made. Air leaks are more expensive than steam leaks because of the losses suffered when compressing the air. The pipe may be tested from time to time by allowing the air at full pressure to remain in the pipe long enough to observe the gauge. A leak should be traced and stopped immediately. When putting together screw joints, care should be taken that none of the white lead or other cementing material is forced into the pipe; this would cause obstruction and increase the friction loss. Also, each length as put in place should be cleaned thoroughly of all foreign substances that may have lodged inside. To render the piping readily accessible for inspection and stoppage of leaks, it should, if buried, be carried in boxes sunk just below the surface of the ground; or, if underground, it should be supported upon brackets along the sides of the mine workings. Low points in pipe lines, which would form pockets for the accumulation of entrained water, should be avoided, as they obstruct the passage of the air. In long pipe lines, where a uniform grade is impracticable, provision may be made near the end for blowing out the water at intervals, when the air is to be used for pumps, hoists, or other stationary engines.

For long lengths of piping, expansion joints are required, particularly when on the surface. They are not often necessary underground, as the temperature is usually nearly constant, except in shafts, or where there may be considerable variations of temperature between summer and winter.

LOSSES IN THE TRANSMISSION OF COMPRESSED AIR To obtain compressed air, an engine drives a compressor, which forces air into a reservoir; the air under pressure is led through pipes to the air engine, and is there used after the manner of steam. The resulting power is frequently a small percentage of the power expended. In a large number of cases the losses are due to poor designing, and are not chargeable as faults of the system or even to poor workmanship.

The losses are chargeable, first, to friction of the compressor. This will amount ordinarily to 15% or 20%, and can be helped by good workmanship, but cannot probably be reduced below 10%. Second, a loss is occasioned by pumping the air of the engine room, rather than air drawn from a cooler place; this loss varies with the season, and amounts to from 3% to 10% and can all be saved. The third loss, or series of losses, is caused by insufficient supply, difficult discharge, defective cooling arrangements, poor lubrication, and a host of other causes, in the compression cylinder. The fourth loss is found in the pipe, it varies with every different situation, and is subject to somewhat complex influences. The fifth loss is chargeable to a fall of temperature in the cylinder of the air engine. Losses arising from leaks are often serious, but the remedy is too evident to require demonstration; no leak can be so small as not to require immediate attention. An attendant who is careless about packings and hose couplings will permit losses for which no amount of engineering skill can compensate.

It is possible to realize 100% efficiency in the air engine, leaving friction out of our consideration, only when the expansion of the air and the changes of its temperature in the expanding or air-engine cylinder are precisely the reverse of the changes that have taken place during the compression of the air in the compressing cylinder; but these conditions can never be realized. The air during compression becomes heated, and during expansion it becomes cold. If the air immediately after compression, before the loss of any heat, was used in an air engine and there perfectly expanded back to atmospheric pressure, it would, on being exhausted, have the same temperature ít had before compression, and its efficiency would be 100%.

But the loss of heat after compression and before use cannot be prevented, as the air is exposed to such very large radiating surfaces in the reservoir and pipes, on its passage to the air engine. The heat that escapes in this way, did, while in the compressing cylinder, add much to the resistance of the air to compression, and as it is sure to escape, at some time, either in reservoir or pipes, the best plan is to remove it as fast as possible from the cylinder and thus remove one element of resistance. Hence, compressors are almost universally provided with cooling attachments more or less perfect in their action, the aim being to secure isothermal compression, or compression having equal temperature throughout.

If air compressed isothermally is used with perfect expansion and the fall of temperature during expansion is prevented, 100% efficiency will be obtained. But air will grow cold when expanded in an engine, hence warming attachments have the same economic place on an air engine that cooling attachments have on an air compressor. In fact, attachments of this kind are found in large and permanently located engines, but their use on most of the engines for mine work is dispensed with, and the engines expand the air adiabatically, or without receiving heat.

The practical engineer, therefore, has to deal with nearly isothermal compression, and nearly adiabatic expansion, and must also consider that the air in reservoirs and pipes becomes of the same temperature as surrounding objects. Consideration must also be had for the friction of the compressor and the air engine. For the pressure of 60 lb., which is that most commonly used, the decrease in resistance to compression secured by the cooling attachments is almost exactly equaled by the friction of the compressor. Hence it is safe, when calculating the efficiency of the air engine, to consider the compressor as being without cooling attachments, and also as working without friction. The results of such calculations will be too high efficiencies for light pressures, which are little used; about correct for medium pressures, which are commonly employed; and too low for higher pressures, and will thus have the advantage of not being overestimated. This result is occasioned by the fact that, owing to the slight heat in compressing low pressures of air, the saving of power by the cooling attachments is not equal to the friction of the machine, but at high pressures, on account of the great heat, the cooling attachments are of great value and save very much more power than friction consumes.

In expanding engines, the expansion never falls as low as the adiabatic law would indicate, owing to a number of reasons, but if the expansion is considered as adiabatic, an error in calculations caused thereby will be on the safe side and the actual power will exceed the calculated power. Therefore, the compressor and engine may be considered as following the adiabatic law of compression and expansion, and as working without friction.

With this view of the case, the efficiency of an air engine, working with perfect expansion, stated in percentages of the power required to operate the compressor, can be placed as here shown for the various pressures above the atmosphere. As the efficiencies for the lower pressures are very much greater than for the high pressures, the conclusion is almost irresistible that to secure economical results air engines should be designed to run with light pressures.

In the foregoing the pipe friction has been entirely neglected. A pressure of 2.9 lb. is credited with an efficiency of 94.85%; but, if the air were conveyed through a pipe, and the length of the pipe and the velocity of flow were such that 2.9 lb. pressure was lost in friction, the efficiency of the air, instead of being 94.85%, would be absolutely zero. It is the power that can be obtained from the air, after it has passed the pipe and lost a part of its pressure by friction, that must be considered when the efficiency of an apparatus is given. The foregoing table of efficiencies with a loss of 2.9 lb. in the pipe, now gives different values for the efficiencies at the various pressures.

It will be noticed that the light pressures have lost most by the pipe friction, 2.9 lb. having lost 100%; 14.7 lb. 11%, and 88.2 lb. only a trifle over onehalf of 1%. It is also seen now 14.7 lb. is apparently the economical pressure to use. But a further careful analysis of the subject shows, that when the loss in the pipe is 2.9 lb., then 20.5 lb. is the most economical pressure to use, and that the efficiency is 71%. But 2.9 lb. is a very small loss between compressor and air engine, and cases are extremely exceptional where the friction of valves, pipes, elbows, ports, etc. does not far exceed this. Yet, with these conditions, which are very difficult to fill, 20.5 lb. is the lightest pressure that should probably ever be used for conveying power, and 71% is an efficiency scarcely to be obtained.

Continuing the investigation and taking examples where the pipe friction amounts to 5.8 lb., it is found that the following efficiencies correspond to the stated pressure:

As friction increases, or, in other words, when more air is used and greater demands are made on the carrying capacity of the pipe, the pressure must be greatly increased to attain the most economical results. If the demands are such as to increase the friction and loss in pipe to 14.7 lb., the air of 14.7 lb. pressure at the compressor is entirely useless at the air engine. The table will therefore stand thus:

It is to be noticed that 88.2 lb. pressure has lost only about 3% of its efficiency by reason of as high a friction as 14.7 lb., while the efficiency of the lower pressures has been greatly affected. As the friction increases the most efficient, and consequently, most economical, pressure increases. In fact, for any given friction in a pipe, the pressure at the compressor must not be carried below a certain limit. The following table gives the lowest pressures that should be used at the compressor, with varying amounts of friction in the pipe:

So long as the friction of the pipe equals the amounts there given, an efficiency greater than the corresponding sums stated in the table cannot be expected. In a case that corresponds to any of those cited in the table, the efficiency can be increased only by reducing the friction. An increase in the size of pipe will reduce friction by reason of the lower velocity of flow required for the same amount of air. But many situations will not admit of large pipes being employed, owing to considerations of economy outside of the question of fuel or prime motor capacity.

An increase of pressure will decrease the bulk of air passing in the pipe, and in that proportion will decrease its velocity. This will decrease the loss by friction, and, as far as that goes, a gain is obtained, but there is a new loss, and that is the diminishing efficiencies of increasing pressures. Yet as each cubic foot of air is at a higher pressure, and, therefore, carries more power, as many cubic feet will not be needed for the same work. It is obvious that with so many sources of gain or loss the question of selecting the proper pressure is not to be decided hastily.

As an illustration of the combined effect of these different elements, a very common case will be taken. The compressor makes 102 rev. per min., pressure is 52.8 lb., loss in pipe is 14.7 lb., machine in mine running at 38.2 lb., efficiency is 55.73%. As long as the friction of the pipe amounts to 14.7 lb., 52.8 lb. is the best pressure and 55.73% the greatest efficiency, but friction may be reduced by reducing the bulk of air passing through the pipe and if the cylinder of the air engine is reduced until it requires 47 lb. pressure to do the same work as before; the friction of pipe will then drop to 11.7 lb. The pressure on the compressor will rise to 58.8 lb., its number of revolutions will fall to 100, and the resulting efficiency will be 57.22%.

Another change of pressure on compressor to 64.7 lb. will decrease its revolutions to 93, friction to 8.8 lb., and its efficiency will rise to 57.94%. If the pressure is increased to 73.5 lb., there will be only 84 revolutions of compressor, 5.8 lb. loss in pipe, and an efficiency of 57.73%. In this last case the efficiency begins to fall off a little, and higher pressures will show less efficiency; but, in comparison with the first example, the same work will be done with a trifle less power and with a decrease of nearly 20% in the speed of the compressor.

Other common examples can be shown where an increase of pressure would result in wonderful increase in efficiency and economy. There are many cases where light pressures and high velocity in the pipe will convey a given power with greater economy than higher air pressures and lower speed of flow through the pipe. But these cases arise mostly when the higher air pressures become very much greater than are at present in common use. Therefore, when estimating the efficiency of the complete outfit, it is found that the pipe and the pressure are very important elements, and must be determined with care and skill to secure the most satisfactory results. As the volume and power of air vary with its pressure, the size and consequent cost of compressor for a certain work will also be affected by the pressure. To plan an outfit for a mine, due regard must be had to the cost of fuel or prime motor power, and also to the cost of compressor, pipes, and machinery, as the saving in one is often secured by a sacrifice in the other.

Next to determining the size of pipe, the skilful engineer has need of further care in the proper position of reservoirs, branches, drains, and other attachments, as only by the exercise of good judgment in this can satisfactory working be secured. The fact that, on account of the diminished density of the atmosphere at high altitudes, air compressors do not give the same results as at sea level, should also be taken into consideration when a compressor is to be installed in a mountainous region.

LOSS OF PRESSURE, IN POUNDS PER SQUARE INCH, BY
FLOW OF AIR IN PIPES 1,000 FT. LONG

Friction of Air in Pipes.-Air in its passage through pipes is subject to friction in the same manner as water or any other fluid; therefore, the pressure at the compressor must be greater than at the point of consumption, in order to overcome this resistance. The power that is needed to produce the extra

.0173

343

434

.0143 537 680

.0762

687

864

.0640 1,073 1,359

.1805

1,030

1,303

.1455 1,610 2,039

.3141 1,373

1,736

.2562 2,146 2,719

.4910 1,717 2,171

.3934 2,683 3,399

.5423 3,220 4,079