THE COLLIERY GUARDIAN AND JOURNAL OF THE COAL AND IRON TRADES. Vol. (TV III. FRIDAY, AUGUST. 21, 1914. No. 2799. The Limiting Conditions for the Safe Use of Electricity in Coal Mining.* By W. M. THORNTON, D.Sc., D.Eng.t At the present time there are 3| million workers in coal mines throughout the world, and there are few mines in which electricity is not used for signalling, lighting, or power. On a conservative estimate the value of the electrical machinery now being installed every year in coal mines approaches d£10,000,000, and with the development of the Chinese coalfields this may be expected to increase. But the possible danger from electrical ignition of firedamp or coal dust has only been at all fully considered in the last few years. Within five years there have been three great colliery explosions which have had a possible electrical origin. At West Stanley, in Co. Durham, direct electrical ignition of coal dust by a flash from a faulty fuse box was sug- gested; at Hulton, in Lancashire, ignition of gas at a faulty switch; at Senghenydd, in South Wales, igni- tion of gas by the sparks on a signalling bell circuit. These disasters have directed public attention to the subject with increasing force, and the present notes, compiled at the suggestion of H.M. Electrical Inspector of Mines, give the more important conclusions which have been reached by recent research. It must be said, however, that there is no class of work in which electricity can be used with more advantage than in coal mining; but to ensure safety it is necessary to take precautions which are not necessary above ground. For some years competition lowered the quality of elec- trical mining gear below that on the surface. This is now less usual, and the quality of mining work is in many cases better than surface work, and, I am told, is specially good in the latest installations in Man- churia. With regard to danger from electric shock there is nothing peculiar to coal mining, and the risk is not here discussed. Lower Limit of Ignition. The initial ignition in most colliery explosions is of firedamp alone. For this to occur sufficient gas must be present to form a lower limit mixture. If the com- bustible gas is pure methane, CH4, the lower limit of inflammability is at 5-6 per cent, of gas in air by volume, at atmospheric temperature and pressure.J If, however, there is ethane or other higher hydrocarbon present, the limit is lowered in inverse ratio to the heat of combustion of the mixture. The calorific value of methane is 189-1, and of ethane 336-6 kilogramme calories per gramme molecule. Thus a mixture in which ethane formed 30 per cent, and methane 70 per cent, of the combustible gas would have a lower limit of inflammability at 4-5 per cent, of gas in air. Such large fractions of ethane are unusual in most large coalfields. The most recent work on the influence of temperature on limits of inflammability is that of M. Taffanel, of the French Coal Dust Research Station at Lidvin, and M. Le Floch. They have shown § that as the tempera- ture is raised the lower limit of inflammability falls. In the case of methane they give the following values. Table I. Initial temperature (degs. Cent.) ... 20 ...175 ...237 ...312 ...555 ... 690 Lower limit ....... 5* * * § *8 ... 5*25 ... 4’75 ... 4’3 ... 3’4 ... 3 These are of interest in the case of heat from gob fires, but electric arcs from breaks are transient and local, and we may in the absence of coal dust take 5-6 as the *A paper read at the meeting of the British Association, Sydney, N.S.W. t Professor of Electrical Engineering in Armstrong College, Newcastle-upon-Tyne. J “ The Lower Limit of Inflammation of Mixtures of the Paraffin Hydrocarbons with Air,” by R. V. Wheeler and M. J. Burgess. Trans. Chem. Soc., 1911, vol. 99, p. 2013. § “ Sur la Combustion des Melanges G-azeux,” par MM. Taffanel et Le Floch. Comptes Rendus, tom. 157, No. 15. October 13, 1913, 1, 595. lower limit in coal electrical mining practice. It will be shown later that even | per cent, of gas with suffi- cient dust present helps to forward an explosion. The usual measure of inflammability of a mixture between the limits has been by the velocity of an explo- sion wave in it, but the measurements are difficult to make. For the purpose of safeguarding the use of elec- tricity in mines, a more convenient test is to find the electric current which when broken in the mixture causes ignition by its breakspark. The results, using non-inductive circuits, that is, with power factors not less than 0-95 when the current is alternating, are as follow for methane and ethane.* Table II.—Methane. Least igniting current. Percentage of gas in 4 f air by volume. Continuous at Alternating at 100 volts. 40 N. 200 volts. 5*6 .. Lower limit . 5'0 .. 1'9 .. 12'5 7'0 .. 1'20 ,. 7'5 8'0 .. 1'0 .. 4'9 9'0 .. 1'07 .. 3'2 10'0 .. 1'20 .. 2'7 12'0 .. 1'50 .. 4'9 13'0 .. 1'65 .. 7'5 14'0 .. 2-10 .. 12'0 14-8 .. Upper limit .. Most inflammable mixtures .. 8 per cent. . 10'2 per cent. Table III.—Ethane. Least igniting current. Percentage of f gas in air. Continuous at Alternating at . 100 volts. 40 N. 200 volts. 3*1 . Lower limit .. 3'5 .1'6 . 12'0 4-5 . 1'0 . 7'0 5'0 . 0'98 . 6'0 6'0 .. 1'08 . 4'5 7'0 . 1'25 . 4'0 8'0 . 1'45 . 4'5 10'0 .. 1'82 . 10'2 10'7 . Upper limit .. Most inflammable mixtures . 5'0 per cent. .. . 7’0 per cent. The curves drawn from the continuous and alternating current values are quite different in type; the former is chisel-pointed, the latter U-shaped and parabolic at the lowest values. In neither case is the most sensitive mixture that for complete combustion. For continuous currents it is below it, for alternating current above it. The least igniting continuous current is very nearly the same for all the paraffin gases.f The influence of ethane is, therefore, not to increase the sensitiveness of the mixture to electrical ignition other than by lowering the limit of inflammability. For alternating circuits ethane actually requires greater current than methane,J and the very great increase in the magni- tude of alternating as compared with continuous current is the first and strongest argument in favour of the use of alternating current from the point of view of risk from breaksparks. Influence of Inert Gases and Circuit Voltage. A small excess percentage of nitrogen appears to have a marked effect on the magnitude of the igniting currents. Pit gas containing from 6 to 12 per cent, of nitrogen, diluted with an air to form mixture having measured percentages of combustible gas, required the igniting currents to be from one and a-half to twice as great as with methane when direct currents were used, and to as much as four times greater when * “ The Ignition of Coal Gas and Methane by Momen- tary Electric Arcs,” by W. M. Thornton. Trans. Inst. Min. Eng., 1912, vol. xliv. f “ The Electrical Ignition of Gaseous Mixtures.” W. M. Thornton. Roy. Soc. Proc., vol. A, fig. 2. X Log. cit., fig. 12. alternating currents at a frequency of 40 were used for ignition.* This point requires fuller examination. The change of igniting current with voltage is not the same in continuous and alternating circuits. In the former the product of voltage and current is approxi- mately constant over part of the working range, in the latter the current does not always diminish as the voltage is raised, but remains constant over a considerable range of it. An alternating pressure of about 500 is the safest, in the sense that at this pressure a greater power can be broken without ignition than at any other. Table IV.—Igniting Currents for a 9’5 per cent. Mixture of Methane and Air. Voltage. Continuous current at 100 volts. Alternating at different frequencies. I 40. 60. 80. 100. 50 2'5 16 — — — 100 1'0 7 16'0 ... 20 .. .. 29 200 0'4 3'8 .. .. 12'2 ... 14'2 .. . 19 300 0'25 ... 3'5 .. 7'0 ... . 12'5 .. .. 17'0 500 0'2 3'5 . .. 6'0 ... . 11'0 .. .. 13'0 700 — 3'0 . .. 4'5 ... 10*0 . .. 10'7 1000 — 0'75 .. .. 1'5 ... 5'5 .. . 8'0 Thus, at 1,000 volts, alternating current at 40 frequency the igniting current is a little less than at 100 con- tinuous volts. It can be shown that at a frequency of about 100 a breakspark ignites gas with more difficulty than at any other frequency. For all circuits other than large power circuits this frequency is to be recommended from the point of view of safety from ignition by break of cable. Influence of Self-Induction in a Circuit. Inductance in a circuit, either continuous or altern- ating, increases the igniting action of a breakspark. The energy of the circuit, which just causes ignition, is the same at all inductances from 0’02 to 0-6 henry,t that is, the igniting current is inversely proportional to the square root of the inductance in a circuit. This is most important in signalling circuits in which the inductance of a bell may reach 0-5 henry. The igniting current at this for a single break is 0*5 ampere. Bell circuits in which open sparking may occur may there- fore be regarded as dangerous, unless otherwise protected. The voltage for this purpose in Great Britain is limited to 25, but there are no conditions specified as to its source. It is most working usually six to 15, except on long haulage roads. The current taken by an ordinary mining signalling bell may be from 0-15 to 0-4 ampere. The inductance of a bell changes with the setting of the vibrating armature, and may be from 0-1 to 0-5 henry. These values are large from an ignition point of view, but the ringing force varies as the square of ampere turns on the windings (and the breakspark, either at the trembler contact or the ring- ing point on the wires, also as the square of the ampere turns), so that some such figures cannot be avoided, for if the current is less the turns, and therefore the inductance, are greater. It is, however, possible by shunting the inductance to remove altogether the risk of ignition. Dr. Wheeler’s tests on the Senghenydd bells gave the following figures.J Mixtures of methane and air were fired by the spark at the bell trembler, when there was a current of 0*45 amperes passing and 4*5 volts from secondary batteries, having an internal resistance of 0*1 ohm per cell. The bell resistance was 10 ohms and its mean inductance 0*34 henry. A bell having 0*13 henry inductance required 7’5 volts at 0*7 ampere, with 10 ohms resistance. No ignition was obtained at the sparks on the signalling wires when the voltage was raised to 26 and the current to 1*05 amperes. Shunting the bell coils * ‘‘ The Comparative Inflammabilitv of Mixtures of Pit Gas and Air Ignited bv Momentary Electric Arcs.” Trans. Inst. Min. Eng., 1912. vol. xlvi., part 1. pp. 112-24, 1913. t “ The Ignition of Coal Gas and Methane.” Loc. cit., fig. 5. X Report of the Chief Inspector of Mines on the Seng- henydd Explosion, p. 36.