is confirmed by the pyrotannic acid blood method, which will positively detect 0.01 per cent.20

ANALYSIS OF HYDROCARBON GASES BY FRACTIONAL DISTILLATION AT LOW TEMPERATURES AND PRESSURES

For the precise determination of the composition of mixtures of hydrocarbon gases-methane, ethane, propane, butane, etc.—as well as certain other gases, it is necessary to resort to the fractional distillation method that depends upon the difference in the boiling points of various liquefied gases.

The procedure is to liquefy the mixture by a surrounding bath of liquid air to the temperature of the latter, then by adjusting the temperatures successively to the different boiling points to draw off the desired gas. The method is slow and suited only for special laboratory research work.21

EFFECT OF INERT GASES ON INFLAMMABILITY LIMITS

2

The effect of inert gases on the inflammability limits of methaneair mixtures, a matter of importance in fighting mine fires, has been studied by several investigators. Clement,22 by adding CO2 to a methane-air mixture, found that the range of inflammability narrowed, the lower limit rising from 5.8 per cent with virtually normal air to 6.5 with 14 per cent of oxygen and the higher limit falling from 14.2 to 6. 9 with 14 per cent of oxygen. Burgess and Wheeler 23 obtained a slightly different range with these methods and apparatus, but, essentially, inflammability ceased when the oxygen content of the CO2-methane-air mixture was 15.8 per cent. These investigators found that, if the inert gas added is nitrogen, inflammability ceased with 13 per cent of oxygen in the air. This limit is lower than that when carbon dioxide is present, mainly because nitrogen has a lower specific heat than carbon dioxide.

Different proportions of carbon dioxide and (excess) nitrogen in the methane-air mixtures would evidently cause corresponding changes in the limiting extinctive atmosphere when the contents of

20 Sayers, R. R., and Yant, W. P., The Pyrotannic Acid Method for the Quantitative Determination of CO in Blood and in Air: Tech. Paper 373, Bureau of Mines, 1925, 18 pp. 21 Burrell, G. A., Seibert, F. M., and Robertson, I. W., Analysis of Natural Gas and Illuminating Gas by Fractional Distillation at Low Temperatures and Pressures: Tech. Paper 104, Bureau of Mines, 1915, 41 pp. Shepherd, Martin, and Porter, Frank, "An improved method for the separation of gas mixtures by fractional distillation at low temperatures and pressures": Ind. and Eng. Chem., vol. 15, November, 1923, p. 1143.

22 Clement, J. K., The Influence of Inert Gases on Inflammable Gaseous Mixtures: Tech. Paper 43, Bureau of Mines, 1913, 24 pp. Coward, H. F., and Hartwell, F. J., The Limits of Inflammability of Fire Damp in Atmospheres Which Contain Black Damp: British Mines Dept., Safety in Mines Research Board Paper 19, 1926, pp. 5, 9, 10.

23 Burgess, M. J., and Wheeler, R. V., The Limits of Inflammability of Fire Damp and Air: British Mines Dept., Safety in Mines Research Board Paper 15, 1925, 7 pp.

oxygen is between 13 and 15.8 per cent. Jones and Perrott 24 found that in hydrogen-air mixtures at about the upper limit of inflammability-72 per cent—as little as 3.9 of oxygen would permit propagation of flame. At the lower limit of inflammability-40 per cent-it required 20.1 per cent of oxygen to permit propagation of flame.

On mixing carbon dioxide with the air and hydrogen when there was 6 per cent of hydrogen and 56.5 per cent of CO, in the mixture, only 7.8 per cent of oxygen was required to propagate flame.

In mixtures of carbon monoxide, air, and CO, at the upper limit of inflammability-72 per cent only 5.9 per cent of oxygen was needed. At the lower limit, with 13.3 per cent of CO, 18.1 per cent of oxygen was needed; and with 16.5 per cent of CO, and 34 per cent of CO2 only 10.3 per cent of oxygen was required to propagate flame.

The various mixtures of inflammable gases and extinctive gases cited above are not identical in composition with coal-mine fire gases behind seals. Mixtures of such gases, however, must differ widely in composition, especially if the sealed area gives off methane. The limits of explosibility of a range of synthetic mine-fire gas mixtures should be determined. It is proposed that this be done in the Bureau of Mines gas laboratory when opportunity arises.

The experiments so far made to test the effect of different inert gases in mixtures with inflammable gases and oxygen on the propagation of flame have shown that although there are differences in effect proportionate to the differences in specific heat of the respective inert gases, such differences are not great; and for practical purposes, as in mine-fire control, the ratio of the oxygen to the inflammable gases is the vital factor. The problem is to determine the ratios for a wide range of inflammable gas mixtures, as has been done for many individual gases.

AMOUNT OF AIR REQUIRED TO VENTILATE A COAL MINE

While working hard a man requires about 35 cubic feet of fresh air a minute. That amount, however, does not nearly meet a man's needs underground, as the gases given off in his working place must be swept away, and the oxygen absorbed by the coal and by oxidation of the timber must be replaced.

Every State coal-mining law but one calls for a minimum of 100 cubic feet a minute for each man and in gassy mines 150 to 200 cubic feet a minute (the latter figure is required by the anthracite regulations of Pennsylvania). Most State regulations, probably on the theory of limiting the number of men exposed to fumes

24 Jones, G. W., and Perrott, G. St. J., " Oxygen required for the propagation of hydrogen, carbon monoxide, and methane flames": Am. Chem. Soc., April, 1927.

from a fire, specify that not more than 70 or 75 men remain in one split of air current. A few States require less, but 75 seems to be a reasonable limit.25

Only a few laws specify where the measurement of the air currents shall be made. Generally they permit the volume of air entering the mine to be measured at the foot of the shaft or in the entrance of a slope or drift. Tests have shown that in 16 Illinois coal mines 26 only 7.6 to 33 per cent (or an average of 18.6 per cent) of the air leaving the fans reached the last crosscuts nearest the respective faces, because of leakage at doors and stoppings and through crevices in pillars and roof. This leakage represents a great waste of air and power. Good practice requires that at least 50 per cent of the air entering a mine shall reach the faces. A better requirement than that calling for a minimum average volume of ventilation per underground employee, which may be measured at or near the entrance of the main intake, is to require a minimum volume for each and every man in his. working place. If that is done, so far as the health of workers is concerned 75 to 100 cubic feet a minute for each man would be ample in the average nongassy or slightly gassy mine. In order that methane may be carried away, gassy mines may require a current of much larger volume than is needed to provide good air for breathing. Air in working places and used entries should contain not less than 20 per cent of oxygen, not over 0.25 per cent of carbon dioxide, and as little inflammable gas as possible.

Even in a mine that is rated gassy men should be withdrawn from a working place when 2 per cent of inflammable gas is diffused through the atmosphere there, although this percentage is below the lower limit of explosibility of methane in a mixture with air5.2 per cent. Where 1 per cent of methane is found in moving air in any split, it is almost certain that somewhere along that split, at the face or in a roof cavity, a dangerously high percentage of inflammable gas will be found.

The best practice in gassy mines is to keep the gas content of every split below 0.5 per cent of methane by increasing the current of the split or reducing the area served by the split, which sometimes may require the subdivision of a split.

Bodies of gas should never be fanned out by men waving coats or brattice cloth, for that practice has caused many disasters. Gas should be removed by suitable coursing of the air, aided by line

25 For detailed figures on minimum air requirements and maximum number of men on a split, see p. 138.

26 Williams, R. Y., Mine Ventilation Stoppings with Especial Reference to Coal Mines in Illinois: Bull. 99, Bureau of Mines, 1915, p. 14.

59495°-28-3

brattices, when no man or source of ignition is on the return from that split. If the body of gas is large-say, 25 cubic feet of pure methane or 500 cubic feet of explosive mixture-no one except those necessary to manage the ventilation should be permitted in the mine until the proportion of gas is lowered to 1 per cent.

In some mines in Belgium, France, and certain districts of Canada the volume of gas given off is so great that violent outbursts occur; these necessitate special provisions, such as keeping boreholes ahead of all advance workings, testing the quantity and character of the gas given off, advancing the headings slowly to allow gas to drain off, not turning rooms until the area has been well drained, and blasting electrically when all persons are out of the mine. Such outbursts have been inconsequential and rare in mines of the United States, but they may occur at depth in regions where the coal beds are much faulted or folded.

GASSY OR NONGASSY" MINES

Precise definition of a gassy mine is of vital importance in coal mining. None of the State regulations specifically define a gassy mine, except to say that it is determined as such by official decision of the mining department of the State concerned. The Operating Regulations to Govern Coal-Mining Methods and the Safety and Welfare of Miners on Leased Lands on the Public Domain, formulated under the auspices of the Bureau of Mines in 1921, do give methods of determining a gassy mine. Recent developments in mining methods at the face and in electric installations and additional experience from explosion accidents make it seem necessary to lower the limits given in those regulations.

The bureau is frequently asked to specify what constitutes a gassy mine, and thus enable mine operators and State mining departments to formulate requirements for mine ventilation and regulations governing the installation and use of electric machinery and power lines. The bureau believes that all coal mines are potentially gassy but admits that there are wide differences in the proportion of inflammable gas found and hence in the degree of danger. Scott Turner, director, approved the following classification by the mine safety board of the Bureau of Mines on May 6, 1926, for policy and for teaching by the bureau's mining staff.

27 Although the word " gaseous has been widely used in this country, it is literally incorrect; hence the word " gassy has been adopted by the Bureau of Mines for this bulletin and for future publications.

CLASSIFICATION OF COAL MINES BY GAS CONTENT

Class 1 coal mine.-A practically nongassy mine in which inflammable gas in excess of 0.05 per cent can not be found by systematic search.

Class 2 coal mine.-A slightly gassy mine in which (a) inflammable gas can be found,28 but in amount less than 2 per cent, in still air in any active or unsealed abandoned workings; or (b) inflammable gas can be found,28 but in amount less than 4 per cent, in some place from which the ventilating current has been shut off for a period of one hour; or (c) inflammable gas can be found,29 but in amount less than 0.25 per cent, in a split 30 of the ventilating current; or (d) inflammable gas enters a split 30 of the ventilating current at a rate 31 not exceeding 25 cubic feet per minute.

Class 3 coal mine.-A gassy mine in which inflammable gas is found in amount greater than specified for a class 2 coal mine.

NOTES ON CLASSIFICATION OF MINES BY GAS CONTENT

With rare exceptions, the inflammable gas found in coal mines is methane. In coal-mining fields where deep wells that pass through the mines or through near-by ground reach natural gas in beds below the coal there have been rare instances of leakage of gas from a well. Natural gas almost always contains more than 85 per cent of methane, and usually ethane, propane, and traces of butane, and higher hydrocarbons as well; therefore the presence of these gases in mine air indicates leakage from a gas well. The lower limit of explosibility of methane-air mixtures in a state of turbulence is 5 per cent, and of natural gas-air mixtures with about 10 per cent of ethane and associated hydrocarbon gases is 4.6 per cent. The limit, therefore, varies with the character of mixture. Also the limit is lowest when ignition is at the bottom of the mixture and the mixture is in a state of turbulence, such as that caused by a blast.

To classify a coal mine properly it is advisable that systematic testing and sampling be done in the return of each split and in areas under suspicion at least three times in not less than 72 hours. All but one of the tests and samples of mine air must show less

28 By employing an approved flame safety lamp, with flame drawn low, or by employing an approved gas detector, or by sampling and analysis with an approved gas analytical apparatus.

20 By sampling and analysis with an approved gas analytical apparatus or by employing an approved gas detector.

30 If but one continuous ventilating current is employed in a mine, this shall be considered a "split for the purpose of this definition.

31 Determined by sampling and analysis and measurement of the ventilating current.