860

THE COLLIERY GUARDIAN.

October 25, 1918.

better insulating material but not so easily handled in
the tubes.

As regards insulating, this should be carried out with
strips of Empire cloth, which is, however, usually
supplied in sheets, and therefore must be cut into strips
in such a manner that the texture of the fabric runs
diagonally to the strip. If a square sheet is cut and
folded as shown in fig. 8, A and B, and then rolled up
as at C, it may be chopped through at points indicated
at the dotted lines, with a large knife under a few light
blows of a mallet.

The strips when unrolled will have the texture as
desired, which will enable the insulation to be tightly
applied without tearing. Too much tension, however,
should not be applied, as the strips stretch considerably
and, if taken too far, causes the varnish surface to
fracture, thus lowering the insulating qualities of the
tape.

The insulation should be applied over the whole coil
and right on to the tubes and up to the stator iron, so
that the coil will be proof against dust. Finally, the
coil should receive one close layer of white cotton tape,
which should be afterwards treated with shellac
varnish.

Assuming that there are two more tubes to wind on
each side of those already wound, there will still be a

B

Fig. 6.

Fig. 7.

Fig. 8.

long length of wire left, and this will be wound into the
next pair of tubes in a similar manner to that already
described.

At the point where the wire leaves the first coil before
entering the second pair of tubes, it should.be insulated
for a few inches with Empire cloth, this being the only
insulation on the short length of wire leaving the inside
coil and entering the outer coil to be wound. This
point is difficult to insulate and should receive special
care.

After all the sections of the coil have been completed
and finally insulated with white cotton tape, they
should be taped together up to the point where the coil
begins to bend to enter the tubes. Several coats of
shellac varnish should then be applied, and the coil may
be connected up to the remainder of the windings.

Former-Wound Coils.

It is unnecessary to say much about these, as they
are usually replaced without difficulty. The principal
point, however, is to see that the fibre wedges are a good
fit and driven well home. If the wedges are of fibre
and are not perfectly dry, they tend to contract with
the heat of a machine, and may become loose and
actually get out of the slots. This is the principal
objection to former-wound coils, and in spite of the
greater difficulty of repairing those wound by hand,
the latter are mechanically stronger and less liable to
damage.

Rotor Repairs.

Rotor windings are usually of the “bar-wound” type,
excepting with some small machines, when they are
wound with wire in a similar manner to that already
described. The variety of machines in existence makes
it difficult to describe the operations of a repair to a
rotor, as the methods of arriving at the same end are
so numerous.

There are, besides those rotors wound with wire, two
methods of fixing rotor bars—(1) t> insert the bars
and hold them in place with fibre wedges; (2)' to insert
the bars in pressphan tubes, and afterwards bend the
inserted end to obtain the pitch of the windings.

Provided that an efficient form of wedging is used,
the former method is easier, as in the latter case the
tubes in which the bars are placed must be cut, in order to
move a single bar. Consequently, the new bar must be
inserted with no tube, as, in order to remove the bottom
bar to provide a new tube, a considerable portion of the
winding would have to be dismantled. Spare bars of
such machines are usually well insulated, and if given
a coat of varnish before being inserted do not, as a
rule, give any trouble. In cases where it is considered
necessary to dry out freshly wound coils before being
put into service, one of the methods described in our
issue of October 11 will meet the case.

Damage to Lens Mines.—M. Lebrun, French Minister
of Blockade, and M. Loucheur, Minister of Munitions,
proceeded recently to Lens, accompanied by MM.
Verier and Basly, the president and vice-president re-
spectively of the Committee on Mines. The whole town
has been razed. The mines are flooded, and the process
of emptying them will require nearly two years. All
needful measures have been taken for putting them in
working order, but no coal can be extracted before the
lapse of from 18 months to two years. It will be at least
five years before the normal output is reached. At the
Courrieres mines the Germans, before leaving, blew up
all the working machinery and pushed their systematic
destruction to the utmost limits.

MOISTURE AND SPONTANEOUS
HEATING OF STORED COAL.*

By S. H. Katz and H. C. Porter.

Because of the uncertainty of the effect of moisture
and its seeming importance among the conditions that
affect spontaneous combustion in stored coal, the work
described in this paper was undertaken by the Bureau
of Mines. The experiments showed that a coal from
Illinois oxidised faster when dry, whereas a sample of
Pittsburg doal oxidised faster when moist. These
discordant results and similarly inconsistent results
from the observations of different investigators may be
coupled with the fact that, under actual conditions of
storing, both coal and air always contain a considerable
amount of moisture. This has led to the conclusion
that moisture in stored coal is not of practical import-
ance as far as its chemical effects on oxidation are
concerned. However, water may produce physical
effects such as cooling through evaporation and reducing
porosity by being held by capillarity in finely broken
coal, and acting in these ways water may have far-
reaching effects on spontaneous combustion.

Method of Experiment.

Of two equal samples of the same coal, one was
thoroughly desiccated and then placed in a closed

bottle filled with air and containing a tube of
phosphorus pentoxide to maintain complete dryness.
The other sample, which contained moisture, was put in
a similar bottle with a tube of water to keep the
atmosphere saturated with watQr vapour. The bottles
were placed side by side in a thermostat. With a few
exceptions the thermostat operated only during the
hours of work when the constant temperature was
necessary to obtain the data. At night and on Sundays
the temperature of the bath varied with that of the
room.

As the oxygen was absorbed by the coal the pressure
in the bottles fell. At regular intervals measurements
were made, from which the pressures in the bottles
were calculated. After 100 days the gas in contact
with the coal was analysed and the changes of the
various gaseous constituents computed.

Description of the Apparatus.

The apparatus shown in the drawing comprises an
albarine stone tank; a motor and connectedreducing
gears for operating the stirrer; a switchboard for
temperature control; and Haldane burettes enclosed
in water jackets. The glass tubes, which con-
nected to the burettes, lead to the submerged
parts of the apparatus in the bath. The ther-
mostat, of the electrically - heated and toluene-
regulated type, is operated with a high degree of
constancy. (A thermometer of the Beckman type
showed a variation in temperature of not more than one
thousandth of a degree from 25 degs. when the room
temperature was 20 degs. Cent.) In the part of the
apparatus used for the dry coal the volume of the gas
in communication with the coal was 4,564 c.c. In the
apparatus used for the moist coal the corresponding
volume was 4,519 c.c.

Method of Determining Pressures.

By means of the gas formula V! = P2 V2 and
from measurements made as described below the
pressure of the gas over the coal was determined.
When the experiment started, the manometer was level,
the burette contained 20,000 c.c. of air, the temperature
of the thermostat was 25 degs. Cent., according to the
thermometer used, and the air in all parts of the
apparatus was at the recorded barometric pressure.
Openings to the outside air were then closed. To
determine the pressure of the atmosphere over the coal
at 25 degs. Cent, at any time thereafter the following
operations were performed :—

The manometer first was made level; 20,000 c.c. (V1)
of gas at the bottle pressure (PJ, which was to be
determined, was then fixed in the burette; next the
volume (V2) of this gas at the original known pressure
(P2) was determined by means of the manometer and
trapped gas; from these data and with the gas equation
the pressure of the gas over the coal was calculated.

Changes in the temperature of the room would tend
to influence the determinations of the pressure by
causing temperature changes of the gas in the burettes
and exposed tubing. However, calculation showed that
a change of 10 degs. Cent, from normal room tempera-
ture would change the determined pressure by only
1 part in 20,000 parts, a negligible effect.

In one experiment the coal used was from the Majestic
Mine at Du Quoin, Perry County, Ill., working the
No. 6 or Herrin coal, the sample for oxidation in dry air

* From United States Bureau of Mines, Technical Paper
172.

being desiccated in a vacuum with phosphorus pentoxide,
whilst that for oxidation in the presence of moisture
was kept in a small Mason jar with rubber gasket and
air-tight cover.

The analysis of the air-dry coal was as follows :—
Moisture 3T7 per cent., volatile matter 35’27, fixed
carbon 53 20, ash 8 36. and sulphur 0 82 per cent. ; heat
value 6,985 calories.

Starting the Oxidation.

The first experiment differed from succeeding ones in
that the coal was put into the bottles before they were
in place in the thermostat. The bottles were then fixed in
the thermostat, connections were made, the thermostat
filled with water, heat°d to 25 degs. Cent, and kept so
for one hour to establish equilibrium of temperature
and water vapour before connections to the outside air
were closed. During these operations the coal was
exposed to the air for three hours before the actual
experiment began.

As indicated by extrapolating from the subsequent
da La, the moist coal did not absorb enough oxygen to
affect appreciably the results afterwards obtained. But
the desiccated coal absorbed oxygen rapidly when first
exposed. It seems probable that this absorption may
have been enough to cause a measurable decrease in the
oxidation afterwards recorded, as compared with the
oxidation in the moist sample during the same time.

Experimental Results.

Several determinations of pressure were made daily
for a period of 100 days. Analyses of the gas in contact
with the coal at the end of the peiiod of 100 days were
made with the Haldane apparatus. Separate determina-
tions of carbon monoxide and methane were made by
the method of iodine pentoxide and cupric oxide com-
bustion.

The rate of oxidation of the moist sample was at
first slow, in so far as decrease of pressure indicates
oxidation. The rate gradually increased thereafter
until it reached a maximum at about the twenty-fifth
day of the experiment. From that time there was a
small and gradual decrease in the rate un il the experi-
ment ended.

In marked contrast to this, the dry sample was
oxidised at a greater rate in the first week. During this
period the rate declined rapidly; thereafter there was
a more gradual decrease in the rate until the end.

Throughout the whole experiment the dry coal
absorbed oxygen at a greater rate than the moist.
Oxidation of the uhdried sample while the dry coal was
being desiccated cannot account for this ; for if the
coal had absorbed all the oxygen in the 400 c.c. of air
to which it was exposed while in the Mason jar, it would
still have absorbed insufficient oxygen to account for
the difference found in the experiment.

In the second experiment the coal used was from the
Bureau of Mines experimental mine at Brucetown, Pa.,
in the Pittsburg bed. The sample for oxidation in dry
air was desiccated in vacuo by phosphorus pentoxide.
The sample for oxidation in saturated air was kept in an
atmosphere of 99 8 per cent, nitrogen.

Analysis of the air-dry coal gave :—Moisture, 1'68 per
cent.; volatile matter, 37’65; fixed carbon, 58'02; ash,
2’65 ; sulphur, 0'70 per cent.

The residue was ground and screened. Parts passing
an 80-mesh sieve and lodging on a 100-mesh sieve were
put in a small Mason jar with tight cover. The coal
completely filled the jar. Two hours time was required
for these operations.

In starting the oxidation the thermostat was com-
pletely arranged for the experiment. The temperature
had stood at 25 degs. Cent, for more than an hour.
First, the desiccated sample of coal was put into
its bottle in the thermostat. Fifteen minutes later the
undried sample was put into the thermostat. Twenty
minutes later the apparatus was closed to the outside
air and the experiment started.

As in the case of the Illinois coal, the Pittsburg
coal which had been kept in vacuo during the drying
process showed a very rapid absorption during the first
few hours of exposure to the air. Thereafter the rate
of absorption gradually decreased till after a week,
when it became quite uniform and continued without
marked change till the end of the 100-day period.

The decline in the rate of oxidation in the moist
sample was more uniform during the entire period of
the experiment. Compared with the rate of oxidation
of the dry sample, that of the moist was greater after
the first five days had passed.

In the third experiment, the coal used was a large
lump from the lot of Pittsburg coal used in experiment
two. During the period of preparation both samples
were kept in an atmosphere of nitrogen.

Analysis of the air-dried coal gave moisture 1'74;
volatile matter, 39’01; fixed carbon, 56'33; ash, 2'92;
and sulphur, 0'79 per cent.

The residue was ground and the parts passing an 80-
mesh screen and lodging on a 100-mesh screen were
used. Two samples of 100 grammes each were taken.
One was put into a desiccator containing water and
the other into a desiccator containing phosphorus
pentoxide. Both desiccators were alternately evacuated
and refilled several times with 99'8 per cent, nitrogen.
Four hours after the lump was crushed the two samples
were in the atmosphere of nitrogen.

In starting the experiment the thermostat, arranged
for experiment, had stood at a temperature of 25 degs.
Cent, for 16 hours.

First the sample of coal in equilibrium with water
vapour was put into the bottle in the thermostat. Ten
minutes later the desiccated sample was put into the
thermostat. Thirty-five minutes thereafter both
samples in the thermostat were simultaneously closed
from the outside air.

On the first and third days of the experiment the rate
of oxidation of the dry sample was greater than that of
the moist coal. On the second, fourth, fifth days, and
all times thereafter the moist coal oxidised more
rapidly.

Comparison of the rates of oxidation of the dry
samples used in the three experiments showed excep-
tionally high rates of oxidation during the first few