Vol. CV.

THE COLLIERY GUARDIAN

AND JOURNAL OF THE COAL AND IRON TRADES.

ERID AY, JUNE 20, 1913.

No. 2738.

DEVELOPMENT OF COALDUST EXPLOSIONS.

Fourth Report of the Explosions in Mines Committee.

The following is the fourth Report to the Secretary
of State for the Home Department upon the enquiry
now being conducted at the Experimental Station at
Eskmeals. The Report contains certain preliminary
observations on the progressive development of a
coaldust explosion as it travels along a gallery, and on
the effects due to obstructions in its path. The Report
is signed by the following members of the Explosions
in Mines Committee:—Sir Henry Cunynghame, Mr.
R. A. S. Redmayne, Oapt. A. P. H. Desborough, Prof.
H. B. Dixon, and Mr. W. C. Blackett.

The following is the text of the Report:—

On Coaldust (Explosions.

As it is our main intention in this Report to describe
some of the phenomena that occur during the explosive
combustion of mixtures of coaldust and air, such as
take place in what are commonly called “ coaldust
explosions,” it is necessary to consider briefly the
conditions under which such explosions can be
generated.

At the outset it seems desirable, to avoid misunder-
standings, to state what we mean by “ inflammation ”
and “ explosion ” of coaldust. Although in the
combustion of gaseous mixtures there is no well-defined
border line between inflammation and explosion, the
different terms being employed merely to indicate
differences in the rapidity of combustion,* yet in the
combustion of mixtures of coaldust and air it appears
possible, and indeed necessary, to draw a line of
distinction between an inflammation and explosive
combustion. Put briefly, our view is that during an
inflammation of coaldust (at all events from bituminous
coals) the flame is propagated largely by the burning of
gases distilled from the dust; whereas during explosive
combustion the portion of each dust particle that aids in
propagating the flame is burned as a whole, both the
volatile matter and the fixed carbon being consumed.
Fundamentally there appears to be no reason why the
inflammation of mixtures of combustible dust and air
should not produce results of the same nature as those
observed during the inflammation of gaseous mixtures,
for in both cases the phenomena are the results of rapid
combustion. The more finely divided the combustible
dust, the more nearly does it approach to the condition
of a gas or vapour, and, if of sufficient lightness to
remain in suspension in the air and thus form an
intimate mixture with the air, there seems to be no
reason why a flame once started in such a mixture
should not be propagated throughout its extent—it
being presumed that the dust can be inflamed readily.

In making comparison, however, between the combus-
tion of a gas, such as methane, and the combustion of
particles of coaldust, the great difference in size
of gaseous molecules compared with that of dust
particles must be borne in mind. If it be assumed,
for example, that a coaldust particle has a diameter
of to’oo mra-> this is over 3,000 times the computed
diameter' of a molecule of oxygen. Moreover, gas
molecules are always in very rapid motion, the mean
speed of oxygen molecules at ordinary temperature and
pressure being about 425 metres per second. Compared
with this speed, that of the floating dust particle is
inappreciable. A dust particle floating among oxygen
molecules might, perhaps, be compared to a balloon
lashed by a storm of hailstones. Despite the difference
in size of the reacting bodies, however, there are several
analogies that may be drawn between explosions of
gaseous mixtures and of mixtures of a combustible dust
and air. As in the case of gaseous mixtures, there will
be a definite ignition-temperature for each kind of dust
and air mixture, and inflammation will not be propa-
gated unless and until each successive “ layer ” of the
mixture has been raised to that ignition-temperature.
Similarly, there will be a “lower limit ” as regards the
quantity of any given dust that must be in suspension
in the air to enable self-propagation of flame to take
place.

The Propagation of Inflammation.

When considering the question of the propagation of
the initial inflammation of coaldust the nature of the
means of ignition must not be overlooked. If a
previously-formed dust cloud in the air be presumed it
can be taken for granted, on a priori grounds, that, once
ignition of a sufficient volume of the cloud has been

* The “ explosion wave ” in gases—the most rapid form
of combustion—-is, however, of a different character from
ordinary combustion. The rate of travel of the explosion
wave is a definite physical constant for any particular
mixture of gases in which it can be set up. Fortunately,
it does not appear possible to set up the explosion wave in
mixtures of firedamp and air—or of coaldust and air.

effected—no matter by what means—propagation of
inflammation will take place throughout the cloud if the
fineness of the dust, the quantity of dust in suspension,
and the ignition temperature of the dust are suitable.
In a mine explosion, however, a dust cloud does not
necessarily exist ready formed; but the cause of inflam-
mation has, in most cases, to produce the dust cloud I
simultaneously by disturbance of the air. In such cases, |
therefore, where the dust is not already in suspension in
the air, both flame and concussion are necessary for
the initiation of more than a purely local inflammation.

For the propagation of inflammation in a mine,
therefore, it appears to be essential either that (1) a
ready-formed dense cloud should be present for a
considerable distance along the roadway ; or that (2) the
source of heat that causes ignition should also create a I
disturbance of the air and thus raise in suspension !
deposited dust. When inflammation occursand the flame |
proceeds along a gallery, the expansion of the gaseous '
products of combustion causes a movement of a column j
of air which is pushed forward in front of the flame. !
The opposition to this movement may be approximately j
calculated for a smooth gallery, since it depends on the
inertia of the air to be moved, or the velocity of the air I
pushed forward, and on the size of the tube.

But in the roadway of a mine this opposition may be
enormously increased by irregularities and obstructions
which cause waves and eddies in the air. A cloud of
dust in air so compressed and disturbed burns with
greater rapidity than in still air. The effect on the
intensity of the advancing flame is thus cumulative. It
may therefore be the case that after- the flame has pro-
ceeded a certain distance in the ready-formed dust cloud,
the column of air driven in front of the flame is capable
of raising the dust lying on the floor and other surfaces,
and so creating an inflammable cloud, independently of
any source of disturbance. This conception of the manner
in which an inflammation of coaldust and air is capable
of increasing in intensity and of propagating itself
throughout any length of roadway containing at the
outset dust merely deposited is due, we believe, to Mr.
W. C. Blackett, who from a study of the conditions
existing in the Seaham, Tudboe, Trim don Grange and
Usworth Collieries, after the disastrous explosions that
occurred, formulated his theory of a “ pioneering cloud.”*

by the induction coil sparks at the nozzle. A large jet
of flame was thus produced, from 10 to 14 feet long, and
of about 1 ft. diameter at its widest part, resembling an
ignited “ blower ” of firedamp.* No unburntj gas
escaped into the gallery.

A large flame of this nature (unaccompanied by
concussion) was found capable of igniting a ready-
formed cloud of coaldust, but was unable to effect the
propagation of flame for- any considerable distance when
the coaldust was simply placed in a heap in its path. It
was therefore necessary to create beforehand a dust
cloud over a certain distance, into which the flame from
the igniter could be projected. This was effected by
spreading the dust on thin horizontal boards, 6 in. wide
and 12 ft. long, suspended from the roof of the gallery
and extending end to end over a distance of 150 ft. from
the igniter' towards the open end of the gallery. These
boards, which hung about 8 in. below the roof of the
gallery, were caused to fall simultaneously, at the same
instant as the gas-flame was produced, by the release of
electrically-controlled catches; and, after falling 6 in.,
still in a horizontal position, were suddenly jerked over
sideways, so that the dust resting on them was projected

Gas-Jet igniter

Explosion lube

Right-angle bend

From Fan

Fig. 1.

OPEN

END

This theory was founded, as aforesaid, on practical
experience obtained during the course of investigations
made after colliery “ explosions.” It remained to be
seen how far it was borne out by what takes place
during experimental inflammations on a large scale. A
number of experiments for the purpose of obtaining
information on this question were made by Dr. Wheeler
for the Mining Association in the Altofts gallery. The
results, which are given below, have not hitherto been
published. Arrangements were made by which dust
clouds of different extent could be ignited by a large
gas-flame, unaccompanied by any concussion. The
general arrangement for these experiments is shown in
fig. 1. The gallery (the Altofts gallery) consisted of a
wrought iron tube 74 ft. in diameter and 600 ft. long.
One end was open to the air and the other was joined,
by means of a series of right angle bends to a fan. For
a more detailed description of this gallery, the Record
of the Mining Association should be consulted.

The igniting arrangement is shown in fig. 2. It
consisted of a vertical cylinder A of about 2 cubic feet
capacity, fitted with a weighted piston B. This cylinder
was filled with coal gas under a pressure of 20 lb. per
square inch, the pressure being obtained by weighting
the piston with lead discs C placed on a platform carried
by the piston rod. An outlet tube from the bottom
of the cylinder was bent upward at right angles and
closed at the top by an electrically-controlled valve D.
Beyond the valve the outlet tube was continued through
the side of the gallery (as indicated in fig. 1), and ended
in a 2 in. nozzle fixed centrally in the gallery and
directed towards its open end. Just before an experi-
ment, sparks from an induction coil were passed across
the nozzle of the tube inside the gallery. When the
valve on the cylinder was released, the weighted piston
fell and forced out the coal gas which was then ignited

* Vide “The Combustion of Oxygen and Coaldust in
Mines,” Trans. Fed.Inst.Min.Eng., vol. 7, p. 54, 1894. See
also Proceedings of the National Association of Colliery
Managers. March 1891.

Fig. 2.

D

into the air. In this manner a ready-formed dust cloud
was produced extending 150 ft. Beyond this 150 ft. of
previously formed cloud, the coaldust was simply
deposited on the floor of the gallery over the distance
remaining to its open end—namely, 350 ft. The coal-
dust, which was obtained by pulverising Altofts Silkstone
nut coal, was present throughout in quantity corre-
sponding to 0'4 oz. per cubic foot of air-space.' A
current of air, of insufficient velocity to disturb the dust
deposited onjthe boards, was forced towards the open end
of the gallery during each experiment.

Three experiments were made in this manner. In one
of them the flame travelled only 200 ft. from the point
of ignition — that is to say, only 50 ft. beyond the
distance over which the ready - formed dust cloud
extended. In the other two experiments the flame
travelled the whole length of the gallery, the average
velocity over successive distances of 100 ft. slightly
increasing, and issued 115 ft. out of the open end into
the air.

It is thus apparent that if an inflammation initiated
without concussion takes place in a gallery along which
for a certain distance a ready-formed dust cloud extends,
sufficient force is developed at the end of this cloud to

* After numerous trials it was found that with a 2 in.
nozzle, and with the gas under a pressure in the cylinder
of 20 lb. per square inch, immediate ignition by the induc-
tion coil sparks was obtained when the valve was released.
A smaller nozzle gave a longer flame, but ignition of the gas
was uncertain.