946

THE COLLIERY GUARDIAN.

May 19, 1916.

RESEARCHES ON FIREDAMP*

By Enrique Hauser.

Firedamp is a mixture of methane with other inert
gases or combustible gases. The inert gases in question
are carbonic acid, water vapour, nitrogen, etc. The
combustible gases are hydrogen, ethane, etc. The study
of the properties of firedamp, therefore, becomes a study
of the properties of methane and the influence on its
properties of the different gases we have mentioned.

Properties of Methane.

(1)	Respirability. — Methane, being a paraffin, is
characterised by its weak chemical affinities. From
the physiological point of view, one could substitute it
for nitrogen of the air at least momentarily, without
experiencing any difficulty in breathing, the mixture
consisting of 21 per cent. O2 + 79 per cent. CH4. From
the chemical point of view, this weak affinity permits
the gas to be burned in certain circumstances with
hydrogen and oxide of carbon in an oxygenated mix-
ture, with the result that the last two gases achieve only
a fractional combustion. This practice has been utilised
with success in the methods of analysis described later
in this paper.

(2)	Retardation of Flashing Point.—A characteristic
property of methane is its retardation of flashing point,
noted by Davy in 1816, by Mallard and Le Chatelier in
1883, and recently confirmed by Taffanel and Le Floch.

When a mixture of combustible gases with air or with
oxygen is submitted to the action of heat, combustion
commences and continues slowly without any external
manifestation indicating the occurrence of a chemical
reaction. But if we raise the temperature little by
little, a point is reached where this combustion takes
place rapidly, manifested externally in the phenomena
of light, and finally in the diminution of volume of the
mixture. It is the occurrence of these latter pheno-
mena which is called the flashing point of the gas.
Thus, according to the experiments of Mallard and
Le Chatelier, in a mixture of carbonic oxide and
oxygen, combustion commences with appreciable
rapidity at a temperature of 400 degs. Cent.; at 447 degs.
Cent, the proportion of the mixture which burns in a
second is 1 in 1,000; at 615 degs. Cent, the proportion
is 1-5 in 1,000; and, at 650degs. Cent., the flashing
point is immediately reached. That is to say, this great
change occurs in a temperature interval of only 35 degs.
At 650 degs. Cent, the whole mixture is burned in some
thousandths of a second.

The same general result is noted in the case of
hydrogen. For example, in an experiment made at
540 degs. Cent., only one-half of the mass was burned;
while at 555 degs. Cent., combustion was completed on
the instant.

In mixtures of methane with air these phenomena of
retardation are very much amplified. Slow combustion
commences at 450 degs. Cent, without arriving at the
flashing point. More rapid combustion occurs at
650 degs. Cent., requiring about 10 seconds to. accom-
plish the burning of the mixture. This duration, or this
retardation, becomes almost negligible at 1,000 degs.
Cent. In mixtures of firedamp and oxygen, the retarda-
tion of the flashing point is a little less.

The scientific reason for the retardation of the flash-
ing point should be sought, at least in part, in the fact
that methane is an exothermic gas which requires for
its dissociation 22-1 calories per molecular weight, and
which does not arrive at the flashing point until the
moment of dissociation. The necessary calories must,
therefore, be supplied by the combustion of a part of the
same gas previous to its breaking into flame. In order
to give an idea of this, we will say that the dissociation
of 100 litres of methane requires a quantity of heat
equivalent to that produced by the combustion of 32
litres of hydrogen. If this is exact, then a mixture of
methane and hydrogen, containing 25 per cent, of
hydrogen, should reach the flashing point without any
retardation. In trials made by the author where heat
was applied at a single point, the quantity of hydrogen
actually necessary was 7-2 per cent, higher.

Be that as it may, it is necessary to consider this
phenomena of retardation of the flashing point as a
protection which hides a danger, and which has often
been the cause of serious accidents, because of too great
confidence reposed in the resistance to the flashing
point of firedamp. The object of the author’s first,
researches on this gas was to reach the flashing point
of methane by. means of incandescent wires. During
this study the phenomenon of retardation became
manifest.

(3)	Flashing of Firedamp by Incandescent Wires.—
The flashing of firedamp by means of metallic wires
heated to a red heat by an electric current has been the
source of many discussions, but, if one studies this
question profoundly, it becomes evident that what
appeared at first to' be divergencies, in reality do not
exist as such. In order to study the flashing by means
of metallic filaments, it is doubtless necessary to employ
a current of low-tension and a circuit without induction,
in order that, at the moment of fusion of the metallic
wire, sparks shall not be produced by the breaking of
the circuit, because these sparks would, of course,
produce the flashing of the mixture.

That is why the author employed in his experiments
a current of 4 volts, obtained from two accumulators
with lead bases of the Dinin model, placed close to the
metallic wire in order to avoid induction effects from
long conductors. Under these conditions it is not pos-
sible to produce appreciable sparks upon rupture. They
would have occurred if he had employed a resistance in
series to absorb the voltage during the dynamic state of
the passage of the current, as was often the case in the
experiments of his predecessors on this subject. On the

* Transactions of the American Institute of Mining
Engineers. -

contrary, in none of the experiments described below
was an explosion obtained by rupture of the wire and
subsequent sparking.

But there is a very important cause , which has con-
tributed without doubt to the apparently contradictory
results of Couriot and Meunier, because these experi-
menters, according to their own statement, worked con-
stantly on the mixture of 9-5 per cent, of firedamp.'
Now, as the higher limit of the flashing point of fire-
damp is 14 per cent, of firedamp, according to the experi-
ments of Burgess and Wheeler, the mixture used by
Couriot and Meunier contained an excess of air of only
14-8 — 9-5 = 5-3 per cent., which is only 1-05 per cent,
of oxygen more than the content required for the higher
limit of flashing point. If this 1 per cent, of oxygen
should be used up by oxidisation of the wire, there would
be left a non-inflammable mixture. This is the reason
why Wullner and Lehmann found the most diluted fire-
damp mixtures (6-66 per cent.) the most easily inflam-
mable. It is on these diluted mixtures that the author
made his experiments, by employing at times methane
prepared in the pure state by means of carbide of
aluminium, or, better, natural firedamp.

Under these conditions, the following results were
obtained :—■

(1)	With wires of ferro-nickel, 0*3 mm. in
diameter, with or without fusion of the wire, there
was no inflammation of the most sensitive mixtures
of pure methane.

(2)	With platinum wires, 0*5 mm. in diameter,
gradually brought to a red heat, inflammation was
obtained in six consecutive trials with mixtures of
from 7 to 7-5 per cent, of natural firedamp, without
seeing the filament melt, although it glowed rapidly
at the moment when the explosion took place. With
filaments of platinum of 0-2 mm. diameter, two explo-
sions occurred in three experiments with natural
firedamp.

(3)	With wires of soft iron, of 0-9 mm. diameter,
the results obtained are very interesting. In employ-
ing a straight filament, either horizontal or inclined,
or a filament curved toward the top or toward the
base, there were six ignitions in 17 experiments with
natural firedamp of from 7-2 to 7*5 per cent. That
is to say, one-third of the experiments resulted in
ignition without fusion to the filament. (In cases
where ignition did not result, the filament was, how-
ever, heated to fusion.) On the contrary, by employ-
ing an iron wire inclined at an angle of nearly 45 degs.,
with a spiral tower toward the middle, there were five
ignitions in six experiments, without fusion of the fila-
ment, and in three of these experiments the same
filament was employed for three consecutive attempts.
In another case, a filament of twisted wire ignited a
mixture which had not been ignited by fusion of the
straight wire.

The explanation offered for these facts is as follows

It will happen in the case of a relatively large filament
of iron that its centre will be at a higher temperature
than the outside, which becomes rapidly covered with a
layer of melted oxide of iron. Now, if a point exists
where the cohesion is somewhat greater than elsewhere,
and where the oxide runs together in the form of a ball
or a pearl, the centre of the wire will be exposed, and
the filament will be partially volatilised, and become an
oxide in the state of vapour, and will produce a flame
which will set off firedamp. Such an event could happen
with a galvanised iron wire at the temperature of the
volatilisation of zinc, which is 675 degs.—or almost
1,000 degs. lower than the temperature of fusion of soft
iron. On the other hand, if we expose a slight obstacle
to the movement of the firedamp, for example, by
placing a cold filament on the incandescent filament, we
would be able to produce the explosion more easily, and
we can produce it with certainty if by rolling the wire
in the form of a helical we heat the gas from tw’O sides,
and require it to pass at least twice across the red hot
filament. In this case the explosion occurs at a tem-
perature lower than in the case of a straight filament.

It is true that Couriot and Meunier have thought
this explanation should be rejected, because they could
not admit that, at a temperature lower than that of its
fusion, “ iron would be volatilised, and, oxidising in a
state of vapour, produce a flame,” and that, even admit-
ting as possible that the volatilisation and the flame took
place, it would be as well produced with a straight wire
as with a curved wire, and better still with a small than
with a large wire. "With all respect due to these gentle-
men, the author submits, on this point, that, being
obsessed with their theory of a gaseous envelope of
oxygen, they did not well understand his experiments,
which often resulted in a contradiction of that theory.
It is not necessary at the moment to discuss extensively
the theory of Couriot and Meunier, but it may be said
that it is not possible to admit that firedamp can be
ignited by radiation alone, because the ignition must
commence at the hottest point, that is, at the contact
with the incandescent wire. Now, as a single contact
does not suffice, because of its short duration, one should
curve the wire in the form of a helical and incline it, to
prolong this contact and transform the slow combustion
of the firedamp into a rapid combustion, that is to say, a
combustion accompanied by flame, which combustion is,
however, only visible at a certain distance from the wire,
where there is an unburned part of the mixture.
Furthermore, as previously observed, the temperature of
the curved wire at the moment of the ignition of the fire-
damp is lower than in the case of the straight wire. It
is between clear orange and white, instead of welding
white or dazzling white. This, therefore, makes the
radiation less important, and makes entirely improbable
the production of a flame by volatilisation of iron of the
centre, which is not exposed as in the case of a straight
wire heated to a dazzling white heat. It may be
remarked here that neither the length of the wire
employed by the author nor the duration of his experi-

ments permitted in any case the absorption by the wire
of oxygen from the explosive mixture, so as to render it
non-inflammable. The amount of firedamp in these
mixtures was never higher than 7-5 per cent., as already
stated.

As to the theory of the possible volatilisation of the
centre of large iron wires, which has been suggested by
the observation of the easy inflammability of firedamp
by a galvanised iron wire brought only to a dark red
heat, the following facts may be cited :•—-First, it is
beyond doubt that a large wire, heated by an electric
current, will have a higher temperature in the interior
than on the surface, because of the loss of heat by radia-
tion and convection. In reality, in the case under dis-
cussion, the difference of temperature between the centre
and the surface of a wire is not very great, because oxide
of iron melts at 1,350 degs. Cent., and its gradual for-
mation should liberate enough heat to compensate in
part for the superficial loss by radiation, and so cause
the hottest zone to be found near the surface. Now,
this layer of oxide, gradually formed, prevents in its turn
the rapid oxidisation of the wire. If this layer should
suddenly disappear by the accumulation of oxide at one
point, the centre of the wire would be exposed, and,
since it is already near its point of fusion, its superficial
oxidisation would take place in this case very rapidly,
and the heating would be sufficient to melt it, with
partial volatilisation. This phenomenon of partial
volatilisation of iron by its heat of oxidisation is compar-
able to' that which takes place toward the end of the
second period of the Bessemer process.

These explanations enable one to admit the possi-
bility of the hypothesis which the author has formulated
to explain the inflammation of firedamp by a large,
straight, incandescent wire.

To confirm these experiments, others were made with
soft steel wires 0*6 mm. in diameter, and with pure
methane or pure carbide of hydrogen. In four attempts
no ignition was obtained with a' straight, horizontal fila-
ment of about 15 mm. length, but ignition occurred once
with a filament about 25 mm. in length, and curved
toward the top. With an inclined filament, having
three helical turns, there were two explosions in two
experiments, using the same mixture which failed
to give ignition in the three consecutive attempts with
the straight, horizontal filament of 15 mm. length. This
last experiment was repeated with the same result,
employing natural firedamp.

The results could not be more conclusive, and even
if the flame or the electric spark were the most appro-
priate means for inflaming firedamp, it is nevertheless
true that incandescent filaments could produce inflam-
mation with equal certitude without the intervention of
the flame, on condition that the wires are not melted
during the time of the retardation of inflammation of the
firedamp, and provided that the temperature of the wire
and its heating surface are such that the firedamp mix-
ture in contact with it can be rapidly brought to the
temperature of quick combustion or of inflammation,
before it burns slowly or its oxygen is absorbed by the
incandescent wire.

The apparatus, employed for the experiments of
inflammation consists of a large cylindrical glass tube
of 40 mm. diameter and 155 mm. of useful length (total
length 180 mm.), placed in a bowl filled with water. It
is closed at the top by a rubber stopper perforated with
three holes. A glass tube passes through the centre
hole, and is joined to a pet cock for introducing the
gaseous mixtures. Glass tubes also pass through the
other two holes, and terminate in metallic points at their
lower ends, between which the spark can be produced.
By means of a drop of mercury the connection is estab-
lished between the metallic points and the wires of the
exterior circuit which enter the tubes. By adjusting the
height of the tubes and points in the cylinder, the spark
can be regulated at the desired height, but, to make the
experiment more sure, the author employed sometimes
two pairs of points, one at the upper part of the test
tube, and the other toward the base. To try ignition
with incandescent filaments, a curved cable with two
conductors, duly isolated so that their points are
separated in the form of a spark, is introduced at the
lower part of the test tube through the bowl of water.
Between the points of these conductors is placed the
filament which is to be brought to a red heat by the
passage of the electric current. The filament is situated
about one-third or one-half of the distance from the
bottom of the test tube. The entire system is bound
together by iron stays on a solid support of wood.

The results of these experiments have also proved that
pure methane prepared in the laboratory conducts itself
in the same manner as natural firedamp as to its igni-
tion by incandescent filaments, from which the author
has been led to believe that it can be used with an
incandescent wire of small diameter, and serve as a
physico-chemical means of characterising the influence
which other combustible gases have on natural firedamp.

(4)	Influence of Combustible Gases on Properties of
Firedamp Mixtures.—Many discussions have taken place
on the existence of hydrogen and hydrocarbons in fire-
damp, as indicated by chemical analysis. The discus-
sions have involved especially the consequences which
the presence of these gases, more inflammable than
methane, would have upon our knowledge of the proper-
ties of so-called “ sharp gas.” The author believes that,
instead of entering into academic discussions based on
samples of gas w’hich it is impossible to procure twice
under the same conditions, it would be much better to
study the properties of synthetic firedamp consisting
only of pure methane, and then to examine the influence
exercised upon the properties of this gas by the addition
of variable quantities of the combustibles mentioned
above, in order to be able to determine at just what
point their mixture with firedamp would augment the
danger. This, however, has not. prevented him from
studying the question of the analyses of firedamp, of
which mention will be made later.