September 1, 1916. THE COLLIERY GUARDIAN. 405 Table I.—Analyses of Gas Samples taken Subsequent to the Explosion. Source of samples. Date of sampling. 1913. Results of analysis. co2. o2. CO. ch4. N->. Behind No. 40 Perc. Perc. Per c. . Per c. P.c. stopping* ... Mar. 2If ... 4'8 ... 11'8 . .. 0'2 . ... 0'8 .., . 82'4 Do. Do. X ... 4'8 ... 11’8 . .. 0'2 . ... 0'7 .., . 82'5 Do. Mar. 23 ... 5'4 ... 9'3 . .. 0'2 . ... 1'3 ... . 83'8 Do. Do ... 5'2 ... 9'6 . .. 0'2 . ... 1'3 ... . 83'7 Do. April 5§ ... 3'5 ... 12'6 . .. 0'2 . ... ro ... . 82'7 Do. — ... 3'5 ... 12'5 . .. 0'2 . .. 1'2 ... . 82'6 Behind No. 40 stopping ... April 20 ... 2'2 ... 15'4 . .. 0'2 . .. 0'7 ... 81'5 Borehole May 13 ... 6'7 ... 5'3 .. .. 3'3 ... 84'7 Do. May 16 ... 6'4 ... 5'4 . .. 3'6 ... 84 6 Do. May 19 ... 6'6 ... 5’2 . .. 3'5 ... 84'7 Do. June 20 ... 7 5 ... 6'2 . .. 2'2 ... 84'J Do. 8ept. 6 ... 5'3 ... 6'2 .. go' ' .. 4'3 ... • 84 2 Pitmouth Do. ... 3'7 ... 11'9 . pj • .. 2'4 ... 82'0 Behind No. 61 stopping ... Sept. 11 ... 3'0 12'1 . rZ .. 2'8 ... 82'1 Behind No. 40 CQ stopping ... Do. ... 3’7 ... 11'3 . .. 2'3 ... 82’7 Behind No 61 stopping ... Sept. 13 ... 3'4 ... 11'2 . .. 2'9 . 82'5 Borehole Do. ... 5'2 ... 7'1 . .. 3'8 ... 83'9 Do. Sept. 19 ... 5'3 ... 6'9 . .. 4'0 ... 83'8 * At main haulageway. t 2.30 p.m. X 3 p.m. § 9.20 p. m. On June 7, 1913, an attempt was made to reopen the mine, or at least to determine definitely whether the introduction of fresh air would make conditions worse. Air was forced into the workings. At the same time, samples of air were collected at the borehole and at one of the drift mouths. Samples were taken regularly at both places to observe the change in the mine atmosphere with the introduction of fresh air. Table II.—Analyses of Gas Samples taken Subsequent to the Explosion and Immediately Following Introduction of Fresh Air. Results of analysis. Sample Date of Time of No. sampling, sampling. CO2. O«. CO. CH4. N2 1913. p.m. It .. . May 24 . .. 3.20 .. . 6'0 ... 4'8 ... 1 ... 3'6 ... 2f .. . Do. . .. 3.20 ... . 6'1 ... 4'8 ... | ... 3'4 ... 1a+ .. . June 7 . .. 1.00 ... . 6'4 ... 5'2 ... I ... 3'6 ... M .. . Do. .. 1.00 ... . 6'3 ... 5'0 ... g ... 3'7 ... 3§ •• Do. . .. 8.30 ... 6'1 ... 5'3 ... ~ ... 3'6 ... 4 .. . Do. .. 8.30 ... . 5'7 ... 7'7 ... § ... 3'0 ... 5 .. Do. . .. 8.45 ... 5'2 ... 7'2 ... £ ... 3'0 ... 6 .. . Do. . .. 8.55 ... 5'8 ... 5.8 ... ... 3 7 7* .. . Do. .. 9.00 ... 5'3 ... 7'2 ... ® ... 3'0 ... 8 .. Do. .. 9.30 ... 5'9 ... 5'5 ... ... 3 6 ... 9* Do. .. 9.45 ... 5'5 ... 6 9 ... 1 ... 3'0 10 .. . Do. . .. 9.55 ... . 5'7 ... 5'7 ... | ... 3'6 ... 11* .. . Do. . .. W.15 ... 5'6 ... 7 2 ... 0'2 ... 3'2 ... 12 .. . Do. . .. 10.20 ... 6'0 .. 5'7 ... 0'3 ... 3'8 ... 13* .. . Do. . .. 10.45 ... 5'3 ... 7'6 3'0 ... § 14* .. . Do. . . 11.15 ... 5'2 ... 7'6 3 0 ... rt 15 .. . Do. .. 11.20 ... 5'8 ... 6'4 3'6 2 16 .. Do. . .. 11.45 ... a.m. 5'6 ... 6'4 3'8 ... g S 17 .. . June 8 . .. 12.15 ... 5'6 ... 6'4 3'6 ... 18* .. Do. . .. 12.30 ... 5'6 ... 7'2 ... ci ... 3'0 ... 19* .. . Do . .. 1.30 ... 5'2 ... 7'8 ... o ... 2'8 ... 20 .. Do. .. 1.15 ... 5'2 ... 7'8 ... g ... 3 2 ... 21* . Do. . .. 2.30 ... 4'8 ... 8'4 ... ... 2'8 ... 22 ” . Do. . .. 2.15 ... 4'8 ...10'0 ... * ... 2'8 ... 23 .. Do. . .. 3.15 ... 4'8 ...10'7 ... | ... 2'6 ... 24* .. . Do. . .. 3.30 ... 4'4 ... 9'3 ... ... 2'8 ... 25* .. . Do. .. 4.30 ... 4 6 ... 9'4 2'6 ... 26 .. . Do. . .. 4'15 ... 4'2 ...11'8 2'0 ... 27 .. Do. . .. 5.20 ... 4'4 ...12 7 2'0 ... 28* ... Do. .. . 5.30 ... 4'0 ...10'2 2'4 ... 29* ... Do. .. .. 6.30 ... 4'0 ...10'9 2 2 ... 3> ... Do .. 6.20 ... 4'3 ...13'6 2'0 ... * following number of sample indicates that sample was taknn at drift mouth ; other samples taken from borehole. t Sample taken to determine conditions in the mine if possible ; air had not been forced in previous to taking the sample. The mine had been sealed as far as possible for several weeks. I Sample taken about two weeks after samples 1 and 2 had been taken and prior to forcing air into the area. It was believed that the fire had been smothered. § Sample taken after air had been forced into the mine for about 15 minutes. A sample was also collected (at 6.20 a.m., June 8, 1913) in the main haulage way, behind a brick brattice, which had been previously built to seal oft the mine, The analysis of the sample showed : Carbon dioxide, 3-2 per cent.; oxygen, 16’8; carbon monoxide, nil; methane, 1-4; nitrogen, 78’6 per cent. About the time samples 29 and 30 were taken, an exploration was made into the mine to determine, if possible, by actual observation, the effect the incoming air was having on the fire area. The oxygen content by this time had gradually increased until, as shown by the analyses, it had reached about 11 or 13 per cent. It was thought that with this percentage of oxygen pre- sent, fanning of the fire, if the coal, embers, etc., were sufficiently hot, would take place. Smoke was dis- covered, and evidence obtained that the fire was increasing in intensity. This indicating premature re-opening, the mine fire area was sealed again. Gases Responsible for the Explosion. A study of the analyses compels the conclusion that the combustible gases produced by the fire were largely responsible for the explosion. Of the many samples collected in this mine by the Bureau’s representatives, following the fire, not one contained an explosive pro- portion of methane. The examination of the gases pro- duced in the mine extended over a period of several months. The highest proportion of methane ever dis- covered was 4’34 per cent. The sample containing that percentage was collected several months after the fire started. At no time was it possible to get a gas sample directly over the seat of the fire. The samples were collected at other accessible places, and, as far as the combustible gas was concerned, represented that given off normally by the mine. Most of the time, of course, the fan was shut down; hence, the methane content represents that which accumulates normally in the mine when air is not introduced to remove it. Only traces of carbon monoxide were found in the sample examined, it is true, but it should be noted that carbon monoxide with any hydrogen or methane formed by the fire would have been produced at the fire—a small place compared with the entire area sealed. Hence the carbon monoxide would have been largely diluted with air from this area, especially at remote points and subsequent to the discovery of the fire. Regarding the cause of the explosion, the fire had, of course, been increasing in intensity from the time it had been discovered up to the time the explosion occurred, about 24 hours, subsequently. The fan had been operating all this time, and air was being forced through the fire area—not much, probably because the fire occurred near a working face, where the air current was still. There was enough, however, so that fresh air was being driven into the area from one side and reacting with coal and probably some wood, while the products of combustion were being removed from the other side of the fire area. With the reversal of the fan there was, of course, no sudden and complete reversal of the air current, but a gradual change of the movement of the air. For a short time after the reversal of the fan most of the air would still continue in its original direction. The movement in this direc- tion would diminish, however, until finally all the air would be driven in the opposite direction. Just how long a time would be required for the complete reversal in any given mine would be problematical. There would be countless eddies formed as one current met another. The effect on the gases produced by the fire would be a dilution of them with the fresh air and a sweeping back of the mixture into the fire area. The amount swept back would at first be small, but would keep growing larger until a mixture would be obtained that contained an explosive proportion of fire gas. If the temperature at any place in the area were hot enough to ignite this mixture, an explosion would result. This is what pro- bably happened in the case of the fire explosion mentioned. Air Movements as Factors in Explosions. Although the sudden reversal of the fan probably precipitated the explosion, yet many explosions following mine fires happen in mines where the ventilation is entirely shut off as far as the fan is concerned. Anything that would cause a movement of the air and thus force explosive gaseous mixtures over a fire area would pro- duce the same result. A fall of rock could sweep gases to a point of ignition. Also, simple diffusion, eddies, and convection currents in a mine where the air was otherwise still could mix explosive gas with air and bring the mixture in contact with flame or witty embers hot enough to ignite them. Explosibility of Gases from Mine Fires. It is interesting to consider the explosibility of those gases that may enter into mine fire explosions. Those that play a significant role are undoubtedly methane, hydrogen, and carbon monoxide. Unsaturated hydro- carbons, such as ethylene, are also produced, but, it is believed, in such small proportions as to be negligible in a consideration of the important gases. The same can be said of ethane. Methane can come from pores and pockets in the coal, at ordinary temperatures, and can be produced by heat reactions. Hydrogen and carbon monoxide are not found normally in coal mines; hence their presence in the gases of mines that are on fire is due to the destructive distillation of the coal or wood, the action of carbon dioxide and hot coal or wood, and the action of steam on hot carbon. Characteristics of Methane, Hydrogen, and Carbon Monoxide. In explosive characteristics, these three gases, methane, hydrogen, and carbon monoxide, differ con- siderably. Both hydrogen and carbon monoxide when mixed with air ignite at lower temperatures than does methane, and have much wider explosive ranges, although their lower limits of explosibility are higher than the low limit for methane, as is shown in the following table :— Table III.—High and Low Explosive Limits of Three Gases. Gas. Low High limit.* limit.* Per cent. Per cent. Methane ...... .......... 5'5 ... 14'00 Carbon monoxide (moist)... 15'00 ... 73’00 Hydrogen ................ 10'00 ... 66'00 Ignition tempera- ture. f Degs. C. 650 to 750 651 585 * Determined by the authors. t Dixon, H. B., and Coward, F. H., Ignition temperature of gases : Chem. News, vol. 99, 1909, p. 139. These results mean that when, in a mixture of air and methane, there is present 5-5 to 14 per cent, of methane flame will travel entirely through the mixture when it is ignited at one point. Similarly, flame will travel through mixtures of hydrogen or carbon monoxide and air, but the limits are higher. With appreciably less and greater proportions than the percentages given above, pronounced flashes would occur upon ignition, with the development of considerably pressure, but such mixtures could not propagate flame indefinitely. As far as the low limit is concerned, methane is the most dangerous, because a smaller proportion of gas is needed to form an explosive mixture than in mixtures of hydrogen or carbon monoxide and air. Hydrogen and carbon monoxide are more dangerous because their explo- sive ranges are so much wider. As regards the combustible gases present in a coal mine, mixtures of all of them in varying proportions are found. In a gaseous mine that normally produces a large amount of methane, the methane would predomi- nate in any part sealed because of a mine fire, for the methane would be accumulating in all parts of the mine, whereas the carbon monoxide and hydrogen would be chiefly concentrated in close proximity to the fire area. In addition, some methane could be produced by the fire reactions. Experiments to Determine Explosibility of Carbon Monoxide, Methane, and Air. As regards a fire in a non-gaseous mine, as in the mine fire described above, only small quantities of methane would accumulate from the coal at ordinary tempera- tures; the quantities of carbon monoxide and hydrogen would be large or small according to the fire reactions. At any rate, in mine fire investigations one is largely concerned with mixtures of the three gases and air, hence some experiments were made by the author having to do with the explosibility of mixtures of carbon monoxide, methane, and air. The experiments were con- ducted in a Hempel explosion pipette, a small spark being used as the source of ignition, and the following results were obtained :— Table IV.— Explosive Limits of Mixtures of Carbon Monoxide, Methane and Air. Mixture. Low High r CO. CH? limit. limit. Per cent. Per cent. Per cent. Per cent. 20 80 6'91 15'89 40 60 7'31 18'26 50 50 8'15 18'66 60 40 9'59 ... 21'18 80 20 10'69 30'17 . 90 10 .. 12'20 41'45 In further tests having to do with the explosibility of mixtures of methane, carbon monoxide, and air, enough carbon monoxide was added to prepared non- explosive mixtures of methane and air to make the resultant mixture explosive. The results are given in Table V. Table V — Inflammability of Special Mixtures of Carbon Monoxide , Methane, and Air. Methane. Carbon monoxi e. Effect of spark. Per cent. Per cent. 4'25 2'92 No inflammation 3 92 4'22 Do. 3'94 4'52 Complete inflammation 3'87 4'60 Do. 3'65 4'98 Do. 3'11 6'06 No inflammation 3'18 6'23 Do. 2'95 6'47 Complete inflammation 2'99 6'18 Do. 3'12 692 Do. It will be seen that as the amount of methane is decreased the carbon monoxide must be increased by an amount that brings the total combustible gases to a pro- portion greater than 5-50, the low limit for methane. This conclusion necessarily follows from the low limits of complete inflammation of the two gases. When about 4 per cent, of methane is present there is needed about 4-50 per cent, of carbon monoxide to produce an explo- sion. When about 3 per cent, of methane is present there is needed about 6-50 per cent, of carbon monoxide to produce an explosion. The explosive properties of mixtures of hydrogen, methane, and air would be some- what similar. Production of Combustible Gases During Mine Fires. In a fire above ground, combustion may be complete, because sufficient air can be supplied to the burning material to permit the formation of carbon dioxide and water as the principal products of combustion. However, a mine fire is usually quickly covered with fallen coal, slate, timber, etc., and the oxygen content of the mine atmosphere is more or less rapidly diminished, and does not have ready access to the burning fuel. As a conse- quence*, combustion soon becomes incomplete, and inter- mediate products of combustion, such as carbon monoxide, are formed. The character and quantity of the gases produced differ according to the stage of the fire, from the time it is first noticed, through the fighting of it, and up to and through the time it is sealed if sealing becomes necessary. The changes that take place may be studied by com- paring mine fire combustion processes with changes that coal undergoes and the gases that are given off when the coal is subjected to different combustion processes above ground. Such studies have to do with the destructive distillation of coal in retorts, as in making illuminating gas, the manufacture of coke in coke ovens, the produc- tion of producer gas, and the burning of coal in boiler furnaces. Chemical reactions may occur in mine fires that are related to the reactions that take place during any of the above well-known processes, from that of complete combustion, as in properly fired boiler plants, to destructive distillation, as in gas-house retorts, where practically no oxygen enters into reaction except that from the coal. Gases from the Destructive Distillation of Coal. The results of the destructive distillation experiments of Burgess and Wheeler* have shown that, with rise in temperature, the production of carbon monoxide increases from about 4-60 to 8-75, the methane decreases from 64-50 to 12-85, the hydrogen increases from 7-0 to 72-90, the carbon dioxide decreases from 10-95 to 0-20, and the so-called illuminants decrease from about 14-15 to 1-00. At the lower temperatures carbon monoxide and hydrogen may be produced in proportions nearly . alike and in large and explosive proportions: but the tendency is for the carbon monoxide to be subordinate to the hydrogen. At high temperatures this tendency is still more noticeable. In mine fires when conditions have reached the point where au explosion might occur, that is, after the fire has burned for some time, tempera- tures will probably have reached their maximum, and have started to fall, owing principally to lowering of the oxygen content, and to incombustible material, such as slate from the roof, falling on the fire. Hence the pro- ducts of combustion that distil at low temperatures may also be of importance. However, when the fire has once reached a high temperature it may cool slowly, for * Burgess. M. J., and Wheeler. R. V.. “The Volatile Constituents in Coal.’’ Jour. Chem. Soc., 1910. vol 97, pt. 2. pp. 1917-1935.