April 30, 1915.

THE COLLIERY GUARDIAN.

909

shaft is 24 in., and length of stroke 56 in. The diameter
of the drum is 14 ft., and the breadth of same 5 ft. The
diameter of winding engine cylinder at downcast shaft
is 22 in., length of stroke 48 in.; diameter of the drum,
lift., breadth 8| ft.

In the shaft bottom there are two sets of Riedler’s
steam pumps, which have a capacity of 160 cu. ft. per
minute, with 1,050 ft. of water head. At present, by
these pumps, 85 cu. ft. of waler per minute is pumped
out to the surface. Besides this, at the first landing
there are two turbine pumps (capacity, 50 ft. of water
per minute, and water head 250 ft.), and at
the second landing two turbine pumps (capacity
30 cu. ft. of water per minute, and water head 300 ft.),
and now 67 cu. ft. of water are pumped out to the
surface, making a total of 152 cu. ft. of water per
minute. Besides the above, electrical pumps are set up
in various parts underground.

At the upcast shaft there is a Rateau fan, the dimen-
sions of which are: Diameter, 12-2 ft.; breadth inlet
30 in., outlet 14 in.; 180 revolutions per minute. ' The
fan engine is of the single-cylinder non-condensing type,
the diameter of which is 18 in., length of stroke 29 in.,
and of 120-horse power. The capacity of this fan is
154,000 cu. ft. per minute. The water gauge is 3 in.

The fresh air (141,000 cu. ft. per minute) which enters
through the downcast shaft is divided into four principal
districts, namely, (1) Togo Rise; (2) Left Nogi kata; (3)
Togo oroshi; (4) Kuroki oroshi. The fresh air is split in
the first two districts, and joins again in Left Nogi kata,
after which it is again split to ventilate the working
faces at the dip side. The fresh air which descends
Togo oroshi is divided into two parts at Second New
Mata oroshi. The air which descends Kuroki oroshi is
split into two portions at the mouth of Second kata.

The Explosion and the Rescue Operations.

At 9.40 a.m. on Tuesday, December 15, 1914, a
great roar was heard, to the consternation of the porters
and overseers on the surface, who were engaged not far
from the shaft mouth. The concrete covering the circum-
ference of the shaft mouth was blown off by the force
of the explosion, and approach to the pit mputh was
cut off. In a moment both shafts emitted violently a
great deal of black smoke and dust, which continued
about 20 minutes. The destructive force of the explo-
sion was so great that the cage which was resting in the
shaft mouth was blown up to a height of 50 ft., striking
the girder of the headgear, whence it fell a distance of
10 ft., to stop there in an inclined position between the
guides. The safety fence surrounding the shaft mouth
was blown off entirely, and winding operations became
impossible until the cage could be repaired. The steel
plate casing covering the upcast shaft mouth, which was
constructed of |in. thick plate, riveted with fin. bolts
at a pitch of 2|in., was brown off in pieces. A steel
door, measuring 3 ft. 7in. by 8 ft. 2 in., constructed by
riveting 2 in. by 2 in. by fin. angle iron frame into
steel plate of sin. thickness, was also blown off by the
explosion. At the Kanada and Hokoku Colliery, which is
about two to four miles distant from this mine, a roaring
sound like distant thunder was heard. The engineer
immediately examined the fan, and found that it was
undamaged, but the fan drift, which leads the air from
the shaft to the fan, was broken at the arch. The steel
casing and steel door were blown off from the mouth of
the shaft. Accordingly, he closed the mouth of the
upcast with planks, and sealed it hermetically by cover-
ing them with clay. He also closed the fracture line of
fan drift with clay, and then commenced the revolution
of the Rateau fan. He further endeavoured to remove
the cage from the guides. In a short time he was able
to lower the repaired cage in the downcast. Before this
he lowered an electric lamp and telephone to the shaft
bottom. Although no message was sent over the tele-
phone, voices could be heard at the shaft mouth, indi-
cating that some men still remained at the shaft bottom.
Accordingly, the engineer quickly descended in the cage
at 3 p.m., and rescued 12 men, many of whom were
severely injured.

I arrived at the colliery at 7 p.m., and immediately
descended the shaft, accompanied by Mr. G. Yakushi
and Mr. M. Imai (who are the engineers of Fukuoka
Mine Inspection Office), Mr. I. Yoshizawa (acting-
director of Hojo Colliery), Mr. K. Ikeda (chief engineer
of the colliery), Mr. Yoshida, and others. At the bottom
a pitiable sight awaited us; men cried for help, and
dead bodies lay here and there. I proceeded to Togo
oroshi, Third kata (120yds. from the shaft bottom), but
as smoke and afterdamp filled the gallery, further
exploration was deemed to be very dangerous, and some
of the party returned to the shaft bottom. But Mr. I.
Yoshizawa, Mr. K. Ikeda, Mr. Yoshida, and others,
without accepting our advice, proceeded a little farther,
and at once were overcome by gas, falling into a
comatose state. Mr. K. Ikeda carried two men
(Yoshizawa and Yoshida) to the shaft bottom, and
quickly administered artificial respiration to them, by
which means Yoshizawa shortly recovered, but Yoshida
was not revived until an hour had elapsed. As the cage
was arrested by debris at the height of 18 ft. from the
shaft bottom, it was very troublesome and tedious work
to convey the injured men to the cage. Accordingly, I
called a doctor from the surface to the shaft bottom,
where many injured men were lying, and he was able
to give the necessary treatment, to them. But the
doctor, 'after having performed medical service for 10
men or more, himself became comatose. It became
clear that the fresh air was not yet circulating regularly
from the downcast shaft, the air having short-circuited
through the first and second landing. These landings,
as well as the pump station at the shaft bottom, were
closed up with planks, and by this means the fresh air
was enabled to enter into Togo oroshi., At this time it
was the opinion of the management that efforts should
first be made to explore the Left Nogi kata, in which
many miners seemed to be still alive. Having the above
object in view, the rescue party proceeded into the
interior of Left Nogi kata at about 10 p.m,, but the men

were all dead. As the bodies were not burned, and
death was caused only by suffocation, it seemed that
the flame of the explosion had not penetrated there. The
Riedler pump was broken by the explosion, and the
water accumulated gradually at the shaft bottom, until
its depth reached 4^- ft. Consequently a bridge was
constructed over the water to facilitate the removal of
the dead bodies. According to the requirements of the
case, the turbine pump (50 cu. ft. per minute) was started
in Togo oroshi, Right First kata, to drain away this
water into the goaf of Right Mata oroshi.

It took three days to examine and fully explore Left
Nogi kata. There 50 or more dead bodies were found.
After finishing the exploration of Left Nogi kata, it was
decided instantly to send the fresh air into the Right
Mata oroshi and other parts of the pit. Roof falls,
and blown off mine props, were extremely numerous,
and as it would have taken a long time to repair the
galleries completely, roads were made sufficient only to
pass over upon the fallen rocks ; and through such narrow
roads the dead bodies were carried. The roof falls in
New Togo oroshi were not great, hence the conveying
of bodies was very easy, but the flame of the explosion
seemed to be very intense, as the surface of the props
was burned.

It was evident that , the water which had accumu-
lated in the shaft bottom had been pumped out into the
goaf of Right Mata oroshi, flowing down gradually into
the dip side of the mine; and if this water had reached
the main level of the 15th kata, then the ventilation of
the Fourth oroshi quarter would be stopped. The
management, therefore, hastened the exploratory work,
and the work of searching and conveying out the dead
bodies was finished in perfect order. In this quarter
of the mine roof falls were extremely great, the height
frequently reaching 10 ft. or more. In the lower portion
of the Second Kuroki oroshi, 16th kata, it was not
possible, to search for the dead, on account of the accu-
mulated water, and steps were taken to drain away this
water, but it was found that the pump could not reach
this position owing to the obstruction of fallen rock, and
the road had to be opened for the -transportation of
pumps. Up to the present (January 10, 1915), the
number of dead bodies conveyed to the surface is 482.
The number of men rescued was 22, of whom two subse-
quently died. The number of dead bodies which still
remained in the mine under the water or under the
fallen rocks was 183. In addition, four rescuers were
killed by gas (CO), and one by falls.

Investigation of the Cause of the Explosion.

In investigating the cause of the explosion it was a
very difficult matter to examine minutely all parts of so
extensive a mine, the underground area being 39 acres,
and it was felt that catechism of the officials and miners
would have been a lengthy and inconclusive proceeding.
Hence I decided to employ the same method which was
applied in investigating the cause of the explosion in
the Futase Colliery, in February 1913. My method of
investigation consisted in examining the condition and
position of adhering coke scale, the state and situation
of blown-out timbers, rails, tubs, etc., as well as the
dead bodies, and by these means attempting to ascer-
tain the direction of the explosion wave. I have plotted
these directions with arrow marks in the underground
map (fig. 2). When the working face or the gallery from
which the explosion originated was clearly defined, it
was essential to enquire of the miners and officials who
were engaged in that face or gallery as to possible,
sources of fire.

The general principles upon which the author
proceeded may be briefly. outlined. In the initial
point of the explosion and its neighbourhood the
coked particles are projected directly upon the
exposed surfaces of props, caps, collars, and walls, parti-
cularly upon the upper portion towards the roof. This
coked scale faces the initial point, and may be as much
as half-an-inch thick, but on the reverse side of the
props or timbers (facing away from the origin) there
are usually only a few scattered particles of coke, unless
some reflex action has had effect. If the coal has a
well-defined coking property, the adhering coked particles
show the following appearance : When there .is abundant
dust accumulated, the coked particles usually form thick,
loosely cohering scale. The coking process of this scale
has been carried only so far as to render the individual
grains plastic enough to cohere, and lightly to stick to
the surfaces. A slight touch will detach and break up
the scale. Such a scale is dull in lustre, and presents a
granular appearance. It often contains pieces of shale
or fragments of rock and timber. Again, at the
point of origin, such a scale rarely adheres to the roof,
but usually coked scale is attached to the bottom side of
a collar. " In general, there are five kinds of coke
which are produced by a coal dust explosion, among
which the first and second kinds of coke are found at
the original point of explosion or in its vicinity.

In the first place, coked scale appears like a splash
of mud upon the surface of the walls, usually adheres to
the ribs, and to a less extent to the roof. It indicates
plenty of dust and a severe heat, but not much
movement.

Another distinct class of coke may be found near
the origin of explosion, or where there have been
secondary explosions. ' This coke is less friable, and
consists of caked dust by the melted bituminous matter.
It indicates an abundance of dust, but not much
movement.

As regards points distant from the point of origin,
when the flame of explosion proceeds through a gallery
from the origin, the air wave or a blast prior to the
flame dislodges fine coal dust from the roof, sides,
and floor of the gallery, and forms dense coal dust
clouds. . If the flame is supplied with such new dust
clouds, it propagates instantly, increases its velocity
and violence, and finally rushes out at the shaft mouth,
passing through every weak point of the mine. In
this case, as its velocity is great, no more coked scale

is found attached upon the facing surface of the
origin, but it adheres as a rule on the reverse side, or
lee side of the wind. The coke scale is uniformly
thin and bright. Now, the explosive wave is not pro-
duced by adding a superfluous amount of gas from
outside, but it arises from the fact that as the mine air
is heated by the explosion flame, its temperature
becomes very high, reaching 1000 to 2500 degs. Cent.,
its volume thus expanding three to seven times of
the original bulk. Accordingly, after the explosion
wave has swept away, the heated gas is gradually being
cooled by the surrounding rocks of the mine, and its
volume begins to contract, generating a vacuum in some
point of the underground workings. In accomplishing
this, the air moves in the reverse direction to the
explosion wave, producing a tolerably strong current.
When this reverse current of the air attains .a tolerable
velocity, it propels fine coal dust in its course, and
deposits the dust on the lee-side or on the reverse side
of the props, as in fig. 4.

If the coal cokes strongly, as in the case of the Yubari
or Futase coal, the grains of coke are large, and it is
easy to determine the coke with naked eye; but as the
coal of Hojo does not coke at all, and so its grain is
very fine, it is not easy to determine whether it is coke
or not. In the mine, to discriminate the coke scale
from the coal dust, there is only one test—to illuminate
the deposit with an electric lamp; the coke scale
possesses a metallic lustre, while the coal dust shows a
dull colour. In the most important case, the two
samples have been collected from both sides of the
props, and these have been examined under the micro-
scope, or subjected to chemical analysis. The reverse
current of air does not necessarily return along the
former direction of the explosion wave, but in order to
cause a vacuum, as the air current comes from the
nearest point; if there is a working face in its vicinity,
it frequently occurs that the current carries such articles
as tin cans, baskets, paper, etc., to the vacuum place
from the working faces. At this stage coke particles
may be driven into facing crevices in the ribs or cracks
in timbers. As the velocity of the explosion wave
increases, it no longer deposits coked dust on the facing
side of props or posts, except occasionally as scattered
particles, but forms thin scales, often bright, on the
reverse side of timbers or projections.

On the point far distant from the origin, though flame
diminishes, if the wave still posseses a full velocity,
corners of brick walls, surfaces of timber and planks
become shaved like, as by a sand blast. This shaved
surface always faces-the direction of the explosion wave.

The third kind of coke belongs to this stage :—The
coked particles are generally bright, and always close
enough to form a thin scale. Such a coke indicates a
violent sweep of the explosion wave, too strong to allow
deposition on the windward side, but depositing coke
scale on the lee-side.

o
°o
o
o°
%°
o

Fig. 9.—Section of Gallery.

A = Coke globules hanging from roof; b = Coke
globules deposited on windward side of prop.

When the explosion flame encounters a newly-formed
dense dust cloud, here it causes the secondary explo-
sion ; then the coked scale adheres on the surface facing
this secondary explosion, but at ■ this stage, as the
velocity of the explosion wave is great, a portion of
cohering coked scale is shaved away by the force of the
blast, and, according to.the relative position of scale to
the direction of the current, it exhibits prism, prismoid,
and other similar forms. At this stage the position of
cohering coke scale is not necessarily on the lee-side of
the: current, but mainly faces the direction of the
current, or parallel to the wind. For the above
reason, in an investigation of the direction of the explo-
sion wave, it requires circumspection to find out the
true direction of the wave, especially at this stage. The
above relation is explained graphically in figs. 5 and 6.

As regards points furthest distant from the point of
origin, when the velocity of the explosion wave decreases
gradually, and begins to die out, the coked dust does
not cohere to the lee-side of the current, but adheres
to the facing side of wind (side faces to the wind). Also,
if the explosion wave rushes out to open spaces, such
as working faces on the longwall system, after passing
through narrow passage ways, then the velocity of the
current decreases rapidly, and the coked scale adheres
on the surfaces facing the current. This relation is
explained in figs. 7 and 8.

The fourth kind of coke occurs at this stage :—Minute
globules of -coke as large as A ^n- diameter are
produced when the coal has strongly coking properties.
These cokes are perfectly round, and the interior of the
coke is nearly hollow, consisting of one or more air cells.
Such a globule attaches on the surface facing to the
current. (See fig. 9.) These globules are like tiny
balloons, and do not adhere together on account of their
being cooled in transit through the atmosphere. When
these globules become attached to the roof by the melted
bituminous matter, they hang down from the roof, as
in fig. 9, A, and may be found on the roof between the
timbers on the lee-side corner of the gallery. If these
globules fly to the gas at the end of the gallery, a gas
explosion may follow,