has shown that Pettersson's method is slow, and that at least 20 minutes must be allowed for the complete removal of the moisture from the air. When the apparatus shown in Figure 5 was used, however, two or three fairly rapid passages of the sample over the phosphorus pentoxide sufficed for the com plete removal of the moisture. The apparatus shown in Figure 5 consists essentially of the burette a, the phosphorus pentoxide tube b, the water jacket g, and the mercury reservoir h. To make a determination draw the gas sample into the burette through the three-way stopcock e, and measure it against the air in the compensating tube c by bringing the mercury in the manometer tube d to the mark f. Then pass the sample slowly through the phosphorus pentoxide tube b into the mercury reservoir h. Three or four passages of the gas sample back and forth between the burette and the mercury reservoir are enough. Measure the gas sample again as at first beginning, and record the loss in volume as moisture. с α h FIGURE 5.-Apparatus for determination of moisture in gas mixture. For explanation, see text The following series of results compare the percentages of water vapor in the laboratory air as determined by means of the sling psychrometer and by absorption with phosphorus pentoxide. The thermometers were graduated to 0.1°. ANALYSIS OF MINE ATMOSPHERES REASONS FOR USING PRECISE METHODS Some of the constituents in mine air are present in extremely small proportions, less than 0.1 per cent of methane, for example, and an equally low content of carbon dioxide. If any methane is present, the management of a mine will want to know it, because any derangement of the ventilating system may allow large quantities of the gas to accumulate. Such accumulations have caused disastrous explosions. Small percentages of methane in the return air current may represent a large volume, measured in terms of cubic feet, expelled in 24 hours. If a current of 100,000 cubic feet of air per minute con-tains only 0.1 per cent of methane, it discharges 144,000 cubic feet of this gas every 24 hours. A current of the same volume carrying 0.5 per cent of methane, a proportion that is not by any means unknown, discharges 720,000 cubic feet of methane in 24 hours. Carbon monoxide, which is produced in mines by fires and explosions, is harmful even in very small proportions,11 and any information as to the amount produced is important. In addition, certain questions, some a matter of dispute among mining men, can be answered only by a close study of mine air. One of these is whether traces of carbon monoxide, and olefin hydrocarbons, hydrogen, or paraffin hydrocarbons other than methane may be found in normal mine air,12 and whether they are given off from the pores of the coal and the cracks in a coal bed, or whether under some conditions they are formed by the action of the oxygen of the air on the coal at normal temperatures. A study of the effect of barometric pressure on the escape of methane in coal mines demands that analyses be performed with especial precision, in order that small changes in the composition of the mine air may be detected. After explosives have been fired in mines, the air may contain certain explosive and noxious gases in extremely small proportions, and because some of these gases are very harmful, even in such proportions, apparatus with which exact work can be done must be used. OLDER TYPES OF APPARATUS Winkler's 18 titration apparatus for the examination of mine air for methane is greatly favored by some mining companies. In connection with Winkler's apparatus, Hesse's 1 apparatus for de 11 Sayers, R. R., and Yant, W. P., Dangers of and treatment for carbon monoxide: Repts. of Investigations, Serial No. 2476, Bureau of Mines, May, 1923. 12 By "normal mine air" the authors mean the air in a mine that is operating under normal conditions. 13 Winkler, Clemens, Handbook of technical gas analysis. 2d English ed., trans. from 3d German ed. by George Lunge, 1902, p. 156. 14 Winkler, Clemens. Work cited, p. 103. termination of carbon dioxide is used. Both apparatus provide for the absorption of carbon dioxide by a standard solution of barium hydrate and the subsequent titration of the excess of barium hydrate with a standard solution of oxalic acid. The method is accurate, the necessary apparatus inexpensive, and their parts easily replaced when broken. A sample measuring at least 200 or 300 c. c. should be taken for analysis. With still larger samples more accurate work can be done. Carbon dioxide and methane only can be determined with this equipment. When the methane is present in proportions at or above the lower explosive limit of mixtures of methane and air (about 5 per cent methane), some other method must be adopted. By Winkler's method the carbon dioxide produced by the combustion is assumed to represent the methane originally present in the mine air. This assumption is undoubtedly true for most mine atmospheres, but does not hold for some samples representing conditions other than normal. The types of apparatus designed by Hempel, Elliot, Orsat, and others for the analysis of producer gas, illuminating gas, or flue gas are scarcely accurate enough for use in examining mine gases. In the hands of a skilled gas analyst Haldane's 15 apparatus gives excellent results. Carbon dioxide, oxygen, carbon monoxide, hydrogen, and methane can be determined with an accuracy of about 0.01 to 0.02 per cent. The determination of methane alone necessitates only the transfer of the sample to the burette, its measurement therein, its combustion, and the final measurements of the contraction in volume produced by the combustion. The determination can be accomplished in about 10 minutes. The instrument must be given the care accorded any correspondingly accurate eudiometric apparatus. It should be protected from drafts, the mercury and the burette should be kept clean, and the solutions should be brought exactly to their respective marks before the burette is read. MODIFIED HALDANE APPARATUS APPLICATION AND USE The Haldane apparatus (see Pls. II, p. 9, and III, p. 18) used by the Bureau of Mines for the analysis of mine air and similar gases is patterned after the original device used by Haldane, and differs only in minor features which have been introduced to simplify its construction and make it easier to operate. The apparatus is especially adapted for the analysis of gases containing small quantities of combustibles, carbon dioxide, and varying amounts of oxygen. The percentage of combustibles present must be below the explosive limit, and enough oxygen must be pres 15 Foster, C. LeN., and Haldane, J. S., The investigation of mine air. 1905, p. 101. ent in the gas itself to burn the combustibles completely. From the contraction in volume, the carbon dioxide produced, and the oxygen consumed when the combustibles are burned, the kind and quantity of the combustible gases present can be determined. However, with this apparatus only three combustibles can be determined, and under certain conditions only two. The Haldane apparatus is useful for the analysis of gases in which the constituents are present in amounts too small for accurate determination by the usual methods-that is, less than 0.1 or 0.2 per cent. It is accurate enough for determining constituents in amounts of 0.02 per cent or more but can not be used for accurate determinations below 0.02 per cent; for proportions as small as this special methods using large quantities of the gas must be employed. The construction of the burette requires that at least 75 per cent of the sample must consist of an inert gas, such as nitrogen, or a constituent which is not analyzed. These restrictions limit the application of the apparatus to certain types of analysis. The Haldane apparatus is especially suitable for the analysis of mine gases containing small quantities of combustibles and atmospheres to which workmen are continually exposed. The Bureau of Mines uses this apparatus almost exclusively for studying the many ventilation problems pertaining to coal and metal mines. It can also be used for the accurate analysis of flue gases, synthetic mixtures of carbon dioxide, natural gas, gasoline vapors, benzene, carbon monoxide, hydrogen, methane, and air below the explosive limit; and to determine small amounts of hydrogen or oxygen in mixtures of the two gases. FACTORS AFFECTING THE ACCURACY OF ANALYSIS The following factors must be considered and eliminated, or corrections must be made for them, to insure accurate analysis of a gas. 1. Change of temperature and pressure during the analysis. 2. Change of water-vapor content. 3. Solubility of the gases in the confining liquid. 4. Solubility of the gases by the absorbent. 5. Graduation of the burette. CHANGE OF TEMPERATURE AND PRESSURE DURING THE ANALYSIS When the temperature of a volume of gas is increased 1° F., the gas expands approximately one four hundred and ninety-second of its initial volume at 32° F., if the pressure does not change. In other words, a change of 5° F. during an analysis makes a difference of about 1 per cent in the volume. Since gas analysis is in reality a series of observations of changes of volume when the gas is treated by different absorbents or subjected to various processes, a temperature change may indicate certain constituents when none are present. For example, assume that during the analysis for carbon dioxide the temperature falls 5° F.; here an apparatus not compensated for changes of temperature would indicate about 1 per cent too much. carbon dioxide in the gas. Likewise, a change of 1 mm. pressure of mercury will change the volume of a gas by one seven hundred and sixtieth of its initial volume at sea level and constant temperature. It is therefore very important to apply corrections for change of temperature and pressure when each volume reading is recorded, or to use a compensating device which corrects for these changes. The Haldane apparatus compensates for change of temperature and pressure (barometric changes) by an auxiliary tube inclosed in the water jacket with the burette, so that any such changes during an analysis affect the gas in the burette and the compensating tube to the same extent. This device, which eliminates the effect of change of temperature and pressure during the analysis, is more fully discussed in the text explaining the operation of the apparatus. CHANGE OF WATER-VAPOR CONTENT The variations of gas volume due to change of water-vapor content, temperature, and pressure are given by the following formula: V2 (P2¬W1⁄2) (273+T1) (P-W1) (273+T2) where V2 observed volume of gas at the observed temperature T2. V1 calculated volume of gas at the temperature T1. P2 and P1 respective barometer pressure in millimeters of mercury. W2 and W1 respective vapor tension of water vapor in millimeters of A rise of temperature increases the tension of water vapor, therefore, unless certain precautions are taken, the compensating device described will not compensate for change in the vapor tension of water vapor in a gas. To accomplish this, always keep a film of water in the burette over the mercury, so that any gas drawn into the burette becomes saturated with water vapor. Likewise, keep 1 or 2 c. c. of water in the compensator tube so that the air in this tube is saturated. When the temperature of the gas in the burette increases, and the vapor tension becomes proportionately greater, the vapor tension of the gas in the compensator tube increases to the same extent and corrects for this error. |