66

THE COLLIERY GUARDIAN.

January 14, 1916.

(3)	The electric current is turned on, thereby
burning out the combustible gas, and the valve is
closed.

(4)	The instrument is shaken to facilitate cooling,
whereupon the water in C falls to a point, on the scale,.
indicating the percentage of combustible gas.

A determination requires less than two minutes.

If the detector were not shaken at the close of a test,
thus forcing the water into the combustion space and
cooling the gases quickly, cooling would be slow, and it
would be a matter of five minutes or so before the gas
in the combustion space would return to the original
temperature of the gas before burning.

This equalisation of temperature is necessary, because
a difference in temperature causes the gas to expand or
contract, and thus alters the level of the water in the
glass gauge C. Hence the gas must be of the same
temperature at the beginning and end of a test.

The aluminium scale of the instrument is easily
removable so that the proper graduated side, corre-
sponding to the combustible gas ;hat is being detected,
can be placed against the glass.

Detecting Combustible Gases in Air.

An instrument for detecting combustible gas in air
is illustrated in fig. 3. It consists essentially of a U
tube, one limb of which is formed by the part T S and
P, and other limb by the part A B C Y and Y1.

To make the instrument ready for use, the head piece
X is unscrewed, and water is poured into the reservoir
B, until the level of the water stands at 0 in one limb
of the U tube, and at S in the other limb.

To take a sample of gas for test, one blows into the
tube T, thereby depressing the level of the water in that
side of the U-tube to some point in P, and raising the
level of the water in the other side of the U-tube to A.,
Next, the water is allowed to resume its former position,
i.e., with level at S and C, thus drawing a sample of
the air to be tested into the combustion chamber B.

The valve A is then closed, and the little platinum
wire (shown attached to the screw plug P) is heated by
means of electrical energy derived from the dry cells
D and E. The hot platinum wire burns the combustible
gas in the sample, whereupon a contraction in volume of
the sample takes place. The water consequently rises
a certain distance in the combustion chamber B, and
falls a corresponding distance in the glass tube K, i.e.,
falls a distance corresponding to the amount of com-
bustible gas originally in the sample. A graduated scale
is placed at M, alongside the glass tube K, and is
marked in percentages of combustible gas, so that a
direct reading is obtained of the percentage, by observ-
ing the level to which the water falls opposite the scale.
The scale is accurately calibrated beforehand, once and
for all, by means of experiments made with known per-
centages of gas. Four graduated percentage columns
are placed on this scale, for the detection of different
combustible gases.

The two small dry cells that furnish the electric
current are made by the Ever-Ready Company, of the
National Carbon Company, and are stock articles. They
provide energy enough for a minimum of 20 determina-
tions. Each determination requires less than two
minutes. The instrument is made of brass and
aluminium, except for the stout glass gauge K. The
weight is under 2 lb.

When the dry cells are exhausted, they are taken out
of the instrument by unscrewing the brass piece N, and
new ones are inserted. A small spring is shown at 0
for holding the two dry cells tightly together and against
the contact .F, which leads to the screw plug P, and
thence to the platinum wire G.

The two dry cells are placed in an insert in the brass
tube Z. Water surrounds this brass tube, and is in com-
munication with P through the place marked by an
arrow. A small contact screw (not shown) is placed at
the base of the instrument, for turning on the electric
current through the platinum wire G.

The instrument can be used for detecting different
combustible gases, i.e., methane in mine air, gasoline
vapour in air, hydrogen in air, coal gas in air, natural
gas in air, acetylene gas in air, carbon monoxide in air
or in flue gas, water gas or producer gas in air, and
other combustible gases in air.

Instead of water, one can use an aqueous solution of
caustic soda, or caustic potash can be used in the device,
thereby absorbing carbon dioxide as soon as formed, and
accelerating the contraction in volume, with the net
result that the instrument is more sensitive than if water
alone were used in the device.

After the combustible gas has been consumed, and the
electric current has been turned off, the lowering of
the water level in the glass gauge is not instantaneous
unless the instrument be agitated, thereby shaking the
water into the combustion space and cooling the burned
gases. The point is that the gases in the combustion
space must be of approximately the same temperature
at the beginning and end of an experiment, or a true
reading will not be obtained.

In the United States, such patents as may be granted
for the device belong to the Bureau of Mines, who am
at present working out a method for exploiting the
devices, so that the public interests will best be served.
Abroad the patents are the property of the 'author.

Imports of Pit Props in 1914.—The imports of pit props
into the United Kingdom during the rear just closed aggre-
gated 2,490,739 loads, the value being £4,786,301. The
respective figures for the two preceding vears were
2,476,854 loads, valued at .£3.259,346, and 3,451,328 loads,
valued at £4,445,066. During the month of December
140,192 loads were imported, valued at £461,632, as com-
pared with 133,320 loads, valued at £224,592 in December
1914, and 228,528 loads, valued at £373,059, in December
1913,

Notes on the Ignition of Explosive Gas Mixtures by
Electric Sparks.*

By J. D. MORGAN, A.M.I.E.E

The ignition of an explosive gas mixture by a spark is
commonly considered to depend upon the communica-
tion of heat.from the spark to the gas. When approach-
ing the subject, it is natural to suppose that the ability
of the spark to ignite the gas can be expressed in terms
of the heat energy of the spark. . On examining the
subject experimentally, however, a suspicion is soon
created that ignition depends partly, if not entirely,
upon some cause other than heat. It is a common-
place of ordinary experience that to produce ignition
the temperature of the igniting means must not be
below a certain definite value. This fact appears to
have led to the belief that temperature is the deter-
mining factor in ignition. It is wmll known, however,
that if the temperature is not accompanied by a suffi-
cient quantity of heat, ignition will not occur. Certain
diminutive electric sparks will not ignite a highly
inflammable gas, yet their temperature may be well
above the ignition temperature. Glowing cordite
emitting a spray of sparks “ cannot ignite a coal gas
jet in spite of the obviously high temperature of the
sparks.’’j- Assuming that heat alone, when accom-
panied by sufficient temperature, is capable of causing
ignition, it would apparently be right to suppose that
the mode of producing electric sparks containing suffi-
cient heat could have no effect upon the igniting pro-
perty of such sparks. This, however, is not found to be
the case.

Prof. Thornton^ obtained curves for the least single
sparks in continuous current and. alternating current
circuits which will ignite a coal gas and air mixture,
which curves showed that a greater amount of energy
is required to produce an igniting spark by an alter-
nating current than by a continuous current: and the
relationship between the number of volts and amperes
in the circuits immediately prior to the production of
the sparks differs in character in the two cases. This
fact is in itself sufficient to prompt the question : Does
ignition depend upon some factor other than heat? A
variety of experiments suggest a reply. If an iron wire
heated by an electric current (continuous) be held over
the disc of a charged electroscope, it will be found that
when the wire first becomes visibly hot there is no
effect upon the electroscope, and gas cannot be ignited.
On gradually increasing the current, a condition of
temperature is attained at which the electroscope
steadily discharges. It is at this temperature that
ignition occurs. In a paper by Prof. Thornton,§ a
similar experiment with platinum wire is mentioned,
and in the same paper there is recorded the most
interesting and important result found by Mr. J. R.
Thompson, that “ it is possible to ignite a cold explo-
sive mixture by the incidence of X-rays on a platinum
surface in it.”

Another experiment consists in so adjusting the spark
gap between a pair of pointed poles in the high-tension
circuit of an induction coil that in neither air nor coal
gas alone can a spark pass, but the poles emit a faint
blue glow or brush discharge, visible in darkness. If the
poles are contained in a small chamber, into which an
explosive coal gas and air mixture is introduced, it is
found that after an interval, which varies with the size
of the gap, the gas explodes. The time can.be made
to vary from a fraction of a second to as much as two
minutes. If the gap is too large, no explosion can be
produced. When the explosion flame appears, a spark
at once passes, due to' the greater electric conductivity
of the ionised gases, and often persists for a second or
more after the flame has vanished. This experiment is
somewhat at variance with one performed by Prof.
Thornton. In another paper|| he states, in connection
with a brush discharge from leaky steam pipes, that
“ gaseous mixtures cannot.be fired by such a discharge,
or by the more active discharge from needle points at
extra high pressure, unless a definite spark passes.” As
a matter of fact, the spark appears to be the conse-
quence, and not the cause of combustion.

Experiments such as those above described all suggest
the ionic origin of ignition. It has been shown that
where a hot wire or spark is the source, ignition only
occurs when ionisation is produced, and ionisation alone
without heat has been found to be capable of causing
ignition.

Dr. H. F. CowardU has expressed the opinion that
“ the ignition of an inflammable gas mixture is largely
governed by the two factors, namely (a) its thermal
conductivity, and (b) the energy degraded when the dis-
charge is passed.” He also stated, however, “ that the
experiments (described in his paper) do not prove
whether the ignition is ultimately a thermal or an elec-
tronic effect.” Investigators of ignition phenomena
are inevitably driven to suspect, if not to accept, the

* From a paper read before the Institution of Eicetrieal
Engineers, Birmingham.

f H. F. Coward/C. Cooper, and C. H. Warburton. “ The
Ignition of Electrclvtic Gas bv the Electric Discharge.”
Transac'ions of the Chemical Society, vol. 101, n. 2278. 1912.

J W. M. Thornton. “ The Ignition of Coal Gas and
Methane bv Momentary Electric Arcs.” Transactions of
the Institution of Mining Engineers, vol. 44. p. 145, 1912.

§ W, M. Thornton. “ The Electric Ignition of Gaseous
Mixtures.” Proceedings of the Royal Scciotv. A, vol. 90,
p. 272, 1914. (See Colliery Guardian, Julv 3, 1914, p. 19).

|| W. M. Thornton. ‘“The Limiting Conditions for the
Safe Use of Electricity, in Coal Alining.” Report of the
British Association, p. 513, 1914.

UH. F. Coward, C. Cooper, and J. Jabobs. “ The Igni-
tion of Some Gaseous Mixtures by Electric Discharge.”
Transactions of the Chemical Society, vol. 105, p. 1069, 1914.

ionic origin of ignition; and there is sufficient evidence
as above indicated for stating that ionisation alone is
capable of causing ignition, and ionisation accompanies
the common electrical methods of ignition.

In a general way it is well known that gas mixtures
are only combustible when the proportions lie within
certain limits. These limits, for methane and air, have
been carefully worked out by Dr. Wheeler, and the least
single igniting spark for mixtures between these limits
have been investigated by both Wheeler and Thornton.
Mixtures of methane and air containing less than 5-6
per cent., and more than 14-8 per cent., of methane
are incapable of ignition.* * * § *

Wheeler f has shown how the least continuous current
t required to produce a single igniting spark varies with
variations in the gas mixture. Single sparks were pro-
duced between platinum points by a contact breaker
arranged in a bell circuit, in which the voltage and induc-
tance were kept constant. The most sensitive mix-
tures were found to lie between 7-5 per cent, and 9 per
cent, of methane.

Using continuous current, Prof. Thornton found the
curve to assume a different form from that already
mentioned. The ordinates of the two curves are not
comparable, since the least igniting current diminishes
as the voltage or inductance increases. From the fact
that a portion of the second curve followed closely a
straight line which passes very nearly through the
origin, Prof. Thornton argued that the igniting current
is proportional to the number of molecules of com-
bustible gas in unit volume of the mixture.

Using alternating current, the character of the curve
alters and assumes a symmetrical form, the current
varying as the “ square of the excess of either of the
combining gases on each side of the point of maximum
inflammability.

A . common method of defining the least spark which
will ignite a given gas mixture is by specifying the
number of volts and amperes, or the number of amperes
and the inductance in the circuit prior to the formation
of the spark. On the assumption that this gives a
measure of the ability of a spark to ignite a gas (or the
“ incendivity ” of the spark), the validity of the method
has been rightly questioned. For both inductive and
non-inductive circuits there seems to be no sufficient
reason, as will be explained later, for the assumption

ior-

5

o

Current o-4 ampere approximately throughout

io

Seconds

Fig. 1.

that the energy associated with a circuit prior to spark-
ing can be regarded as a measure of the incendivity of
the spark. Nevertheless, there is a practical value in
curves showing the relationship between the number of
volts and amperes or amperes and inductance in circuits,
which, when broken, give rise to sparks capable of
igniting a given gas mixture, for they indicate conveni-
ently the practical conditions under which dangerous
sparking becomes possible.

The foregoing researches showed how, at a fixed
voltage (90 volts), the least current capable of producing
an igniting spark in an 8 per cent, mixture of methane
and air varies with the inductance of the circuit, and
that, in an inductive circuit at relatively low voltages,
the current varies but little over a wide range of voltage
variation (see figs. 1 and 3). The amounts of voltage
and current, or current and inductance, capable of pro-
ducing dangerous sparking are comparatively small;
hence the necessity, already well known, of adequately
safeguarding electrical apparatus to which explosive
gases are accessible.

In experiments with single sparks, when a given spark
was of the intensity just below that required for ignition,
it was reproduced a number of times, and if ignition was
never obtained, the spark was regarded as ineffective in
the particular gas employed.. It must be borne in mind,
however, that the repeated sparkings which were found
by the experimenters to be incapable of igniting the gas,
wore produced at a comparatively slow rfite by rotating
one of the poles by hand or a small motor; and unless
this fact be noted, the above assumption is liable to be
misleading. The author finds that a single spark, which
when repeated slowly will not ignite a gas, will after a
more or less definite interval produce ignition when
repeated rapidly. The element of time seems to him
to be a factor of importance .in ignition phenomena. If
instead of.a single break device a vibratory make-and-

* M. J. Burgess and R. V. Wheeler. “The Limits of
Inflammability of Mixtures of Methane and Air.”
Transactions of the?Chemical Society, vol. 105, p; 2591, 1914.

1 R. V. Wheeler: Home Office Report on Battery-Bell
Signalling Systems. 1915. (See Colliery Guardian,
April 23, 1915, p. 855).