118

THE COLLIERY GUARDIAN

January 21, 1916.

pressor is the number of cubic feet of free air compressed
per kilowatt input to the motor. These expressions are,
however, unsuitable when compressors operating under
dissimilar conditions are compared, and it is better to
state the over-all efficiency as the ratio of the power
available in the air delivered to that supplied to the
engine or motor driving the compressor.

There is <a natural tendency, on the part of manu-
facturers ■ whose machines may show a specially good
intermediate efficiency, to make the most of that
particular feature; but each stage in the process of com-
pression should be equally good, if a high over-all
efficiency is to be obtained. A high volumetric efficiency,
for instance, does not necessarily imply a high efficiency
compressor. As a matter of fact, a low volumetric
efficiency may indicate liberal ports and valves, and
consequently very small pressure losses in the com-
pressor, resulting in a high over-all efficiency.

When compressors were bought and sold on the
“ swept ” volume of the air cylinders, and not on the
actual quantity of air compressed, the volumetric
efficiency was a matter of great importance to the pur-
chaser. Nowadays, compressor contracts are usually
based on compressor output, and, provided that the
compressor yields its full specified output of air, the
volumetric efficiency is of little interest to the user. So
long as the purchaser obtains the specified output of
air compressed to the required pressure by the con-
sumption of not more than the specified amount of
steam or electrical energy, he need not trouble himself
about the volumetric efficiency, or any of the other
various efficiencies which are so often quoted.

The over-all efficiency of an electrically-driven com-
pressor is the ratio of the isothermal horse-power in the
air discharged to the electrical horse-power input to the
motor. A good modern direct-coupled high-speed
motor-driven compressor will show an over-all efficiency
of from 60 to 65 per cent. The higher over-all efficiency
in the case of the electrically-driven compressor is
explained by the fact that the conversion of electrical
energy into work is .(in our present state of knowledge)
a much more efficient process than the conversion of the
heat energy in steam into work.

Air Control.

The economical working of a compressor is affected to
a large extent by the means adopted for controlling the
air output. The most economical method is to vary the
speed of the compressor in proportion to' the demand for
air.

To effect this, an air governor, working in conjunction
with a centrifugal speed governor, controls the speed of
a steam-driven air compressor. For air pressures up to
the full working pressure, the compressor is under the
control of the centrifugal governor. When the working
air pressure is exceeded, the air governor takes control,
varying the speed in proportion to the demand for air.
It operates in the following manner :—•

When the air pressure is below the working pressure,
a relay valve is held in its inner position by springs at
the end of a lever. In this position the air pressure is
shut off from the cylinder, which is open to the atmo-
sphere through the exhaust port, and the speed of the
compressor is under the control of the centrifugal
governor. Excess air pressure lifts the relay valve
against the resistance of the springs, shutting the
exhaust port, and admitting pressure behind the plunger.
The latter moves outwards, depressing the near end of
a rocking lever, and raising at the other end the vertical
rod attached to the steam throttle valve, thus reducing
the supply of steam, to the engine. Since the lower
ends of the springs are connected to the. rocking lever,
the movement of that lever increases immediately the
tension on them, the relay valve is forced back to a
middle position over the cylinder port, and the gear
remains stable in the new position. On any further rise
in air pressure the operation is repeated, and the steam
supply is further reduced.

A falling air pressure has the reverse effect. The lower
air pressure under the relay valve is insufficient to main-
tain it in the middle position, and under the pressure of
the springs it , moves inwards, allowing some of the air
behind the plunger to escape into the atmosphere. The
plunger is thereupon moved inwards by a main control
spring relieving the tension on the relay valve springs,
and allowing the relay valve to centre itself again. With
a variation in the air pressure of less than 5 per cent.,
the range of speed obtainable is about 5 to 1. Without
the relay principle embodied in this design the speed
control would be intermittent, that is, the compressor
would run either at full speed or at the minimum speed.
Although a common arrangement, it is not so efficient
as the one just described.,

Variable speed can also be secured in connection with
compressors driven by direct-current motors, and to a
limited extent in the case of alternating-current motors;
but, owing to the increased cost of the motor, variable
speed is not usual for electrically-driven compressors of
large size. Automatic stopping and starting devices are
occasionally used for the smaller sizes, but, in general,
electrically-driven compressors are so devised as to run
at constant speed.

A common form of unloading device for constant
speed compressors is the automatically-operated air
inlet throttle valve. When the action is intermittent,
this is the simplest and most efficient arrangement.
When the inlet valve is designed to throttle the air
gradually, there is a considerable waste of power at light
loads, and the temperature of the air discharged is liable
to become excessive.

With an intermittent automatic air inlet valve, under
normal conditions the inlet valve is full open, and there is
no waste of power due to wiredrawing the air. Excess air
pressure shuts the valve completely, and the no-load
losses are practically confined to the unloaded friction

of the compressor. Since no air is being drawn in, there
is no churning of the’air through the delivery or suction
valves.

Where automatic stopping and starting gears are used,
the air pressure at the discharge side of the compressor
should be released before the motor is started again.
This necessitates a non-return valve in the air main
close to the compressor, and a relief valve automatically
operated by the motor starting gear.

Owing to the inertia of the flywheel and moving parts
at full speed, the full load torque is proportional to the
mean pressures in the air cylinders; but, when starting,
the flywheel effect is absent, and the resistance to rota-
tion is that due to the maximum discharge pressures on
the air pistons.

A compressor of a capacity of 450 cu. ft., driven by a
440-volt direct-current motor, recently tested by the
author, required about twice the full load current at
starting when the motor had to start up with the full
air pressure on the compressor outlet valves. When the
air pressure at the discharge was released, the starting
current fell to less than 50 per cent, of the full load
value. Other tests carried out on the same compressor
have demonstrated that relief of the pressure at the dis-
charge is much more effective in reducing the starting
torque than shutting off the air at the inlet.

Leakage Tests.

The maintenance of a high standard of economy
involves occasional tests for piston and air valve leakage.
When indicators are available, internal air leakages can
be detected readily by the shape of the diagram, or by
comparing the apparent volume of free air drawn in with
that shown at the discharge. The temperature and
pressure of the air entering and leaving the cylinder
must be known in order to calculate the amount of air
that is passing.

A standing test will often reveal leakages too small
to affect the running intercooler pressure. To carry out
a standing test, stop the compressor, leaving it connected
to the full air pressure in the receiver, and observe the
discharge and intercooler pressure gauges. When the
air valves are in a good state of adjustment, the pressure
in the receiver and intercooler will remain stationary,
or fall only very slowly. Leakage past the second-stage
valve will cause the intercooler pressure to rise; on the
other hand, leaky first-stage valves allow the air to
escape from the intercooler, and the pressure to fall.

If there are indicator cocks fitted to the air cylinders,
the leakage can often be traced to the offending valve or
valves by opening the cocks in succession, starting at
the first stage. When the air valves .are in good order,
the flow of air from the open cock will be quite feeble,
not sufficient to extinguish a candle held in the stream
6 in. from the cock.

Air Pipe Work.

Special consideration should be given to the arrange-
ment and proportions of the pipework close to the
compressor. Owing to the pulsating motion of the air
flow, surges can be set up which affect the efficiency of
the compressor. If, for instance, the pipe on the inlet
side is of the critical length in proportion to its diameter,
air surges will be set up which, if their period syn-
chronises with the beat of the compressor, may become
so much intensified as to increase greatly the flow of air
into the compressor. The increased volume of the air
passing is, of course, accompanied by an increase in
the power required, and, when the compressor is
electrically driven, this may cause an overload on the
motor. In one instance, where the motor operating the
compressor was considerably overloaded, it was found
that elimination of the inlet pipe resulted in a drop of
no less than 40 horse-power in the power required to
operate the compressor. When the inlet pipe has
happened to be of suitable proportions, the air flow has
been so stimulated that volumetric efficiencies in excess
of 100 per cent, have been demonstrated.

The effect of these air surges in the inlet pipe has been
found to increase the output by 15*6 per cent., and the
power required by 17 per cent. This effect is not, of
course, peculiar to high-speed air compressors; cards
showing the same characteristics have been obtained
from slow-speed compressors. The surge can be elimi-
nated by a considerable increase of the bore of the
suction pipe close to the compressor. From 10 to 15 ft.
of piping, of an area large enough to reduce the air
speed to about 10 ft. per second, is effective in most
cases.

The increased output obtained when the air surge
occurs on the suction side is some set off for the addi-
tional power required; but when the surge is on the dis-
charge side, there is no similar compensation, and the
extra work due to the surge is wholly wasted. Surges
on the discharge side can be greatly minimised by
increasing the pipe capacity to the compressor.

These surging effects are naturally of greater import-
ance the lower the delivery pressure is at which the
compressor works; and in the case of a compressor
operating at a pressure of 5| lb. per sq. in., such surging
effects occurring in the inlet and discharge pipes have
been known to increase the power required to operate
the compressor by as much as 30 per cent.

It might be supposed that surging effects would be
absent in the case of rotary or turbo-compressors, but
such compressors are specially liable to be adversely
affected in this way. With turbo-compressors, as the
output is reduced, a point is reached (generally about
five-eighths load) at which the air begins to surge to and
fro in the compressor, causing shocks, noisy working,
and very high air temperatures. This tendency to surg-
ing is responsible sometimes for serious difficulties in
running turbo-compressors in parallel delivery into a
common main. The remedy usually applied is to keep
the output above the critical load by discharging part of
the air direct to the atmosphere. To blow the surplus
air to waste is undoubtedly the most inefficient method

of controlling the output, and one that would not be
entertained in the case of a reciprocating compressor.

Straining Arrangements for Air.

It is important—at any rate, if there is much dust
about—that some adequate arrangements should be
made for filtering the air, and the expense involved in
so doing will prove to be amply repaid by the improved
results derived from the air compressing plant and tools.
At a colliery there is almost invariably a great deal of
dust, and in an extreme case it has been found that, not
only have the airways into the compressor become
choked with a caked composition of coal dust and oil,
but that the clearance spaces in the air cylinders have
been entirely filled up, causing the compressor' to knock.
The dirt also accumulates on the valves, thus increasing
their weight, and therefore the risk of breakage.

The inter-cooler efficiency is also impaired, with a
resulting increase in the air temperatures. Undoubt-
edly a considerable proportion of the dirt is intercepted
in the receiver, but some is certainly carried on into the
tools, increasing their wear and tear and liability to
derangement. On the whole, there is no doubt what-
ever that an air filter is a paying proposition. In some
cases where it has not been considered advisable to
incur the expense of such a filter, the inlet pipe has
been removed, and the compressor arranged to draw its
air supply from the compressor house, thereby avoiding
to some extent the dust; but this course is not advisable,
as the noise arising from the air inlet is very unpleasant,
to say nothing of the fact that the air is, as a rule,
being drawn from the compressor room at a higher
temperature than that outside.

The arrangements for the filtering of the air must be
adequate for the purpose. It is useless to set up, as
occasionally supplied, a frame of wire gauze very little
larger than the air inlet pipe on the compressor. Such
a sheet of gauze, if of fine enough mesh to filter effec-
tively the air, very quickly becomes choked with dirt,
and may go to pieces and pass into the compressor.

There are several fairly well-known patented systems
of air filtration, which, however, need not be described
here. They have the advantage of being compact, but
the only way of removing the dirt from some of them
is to send the strainer cloths away to be cleaned, which
is not, as a rule, a very practicable, process.

One simple and comparatively inexpensive straining
arrangement was developed on the Rand, where there is
a vast quantity of dust; and this form has become almost
standard in that area. The straining material is of the
best cocoanut matting supported by wire netting, and
is enclosed in wooden-frames, which are so arranged as
to form a box, into which the inlet pipe of the compressor
is introduced.

The frames are secured by bolts and fly-nuts, and are
thus easily removable when.it becomes necessary to clean
them, a spare frame being substituted. Experience has
shown that the area of matting should be about 50 to 60
times the sectional area of the inlet pipe. The dimen-
sions of the strainer vary, of course, with the size of the
compressor, and may either consist of a box carried on
the wall of the compressor house and suitably protected
against the weather, or may be of much larger
dimensions. Two strainer houses, each having 32 of
these strainer frames, are in use, in connection with
four high-speed compressors, each of a capacity of
6,500 cu. ft., at the New Modderfontein gold mine on
the Rand. The houses stand apart from the main build-
ing, and the air is carried into the compressor house by
culverts under the floor, into which the inlet pipes of
the compressor dip down. Each of the four compressors
is operated by a synchronous motor of 1,000 horse-power,
running at a speed of 163 revolutions per minute. A
special characteristic of these compressors is that, influ-
enced doubtless by previous experience, the mining
engineer specified that the intercoolers should be so
designed that the whole of the tubes could be readily
lifted out and dropped into a soda tank, in which they
might be boiled for the removal of the grease. Reports
received after more than two years’ operation showed,
however, that very little dirt or grease had accumulated.
Although each compressor was fitted with 40 light-plate
air valves (160 in all), none of them had become broken
after two years’ running. The mining engineer con-
sidered that this, as also the cleanliness of the,.com-
pressors and intercoolers, was attributable largely to the
excellence of the arrangements made for cleaning the air,
and, further, to the very small quantity of oil used for
internal lubrication of the air cylinders. Less than a
pint of oil is used in each of these large compressors for
a continuous run of 24 hours. About three-quarters of
this is used in the air cylinders and the rest in the crank
chamber.

In another installation of two electrically-driven air
compressors, each of a capacity of 1,500 cu. ft. working
at a pressure of 1001b., the air filtering arrangement
consists of frames such as those described above, with
a roof to keep off the rain, the air inlet pipe being carried
underground. If necessary, in the interests of the
motor, the pressure can be relieved from the discharge
pipe before starting up by closing the 6-in. sluice valve
and by opening the safety valve, gear being provided for
the purpose.

The motor is directly connected to the compressor
without the interposition of a flexible coupling. ' Experi-
ence has shown that flexible couplings for compressor
driving are liable to prove very troublesome, causing
much more trouble, in fact, than they are ever likely
to prevent. At the same time, it is very necessary
that the motor shaft should be of ample dimensions, as
very heavy stresses are liable to occur if torsional oscilla-
tions are set up by the interaction of the heavy revolving
weights constituted by the flywheel and the rotor. For
this reason, a single-bearing motor is to be recommended
whenever possible. The length of shaft necessary to
accommodate the extra bearing between the rotor * and
the rotor-coupling is a weak feature in the two-bearing
motor. With the single-bearing motor there is no neces-