CHAPTER XXXI

CHIMNEY GAS ANALYSIS

We have found in preceding discussions that for practical purposes the gases passing out through a chimney from the central station boiler are usually considered to be composed of carbon dioxide, oxygen, carbon monoxide and nitrogen. Since these constituents are usually determined volumetrically we shall represent them by the symbols V1, V2, V3, and V4, respectively. We shall now proceed to a discussion of the usual methods employed in determining the flue gas analysis during the boiler test.

The Taking of the Flue Gas Samples and Analysis. Certain solutions have been found in the chemist's laboratory that will absorb carbon dioxide and will not absorb oxygen, carbon monoxide or nitrogen. Again another solution has been found that will absorb oxygen but will not absorb carbon monoxide or nitrogen. And still a third solution has been found that will absorb carbon monoxide but will not absorb nitrogen. If then a contrivance can be set up so that a flue gas sample may be successively washed in these solutions, a means is provided for determining an analysis by volume.

Orsat Apparatus.-Let us then see how the flue gas analysis is taken.

The apparatus commonly called the FIG. 169.-A carbon dioxide Orsat Apparatus (see Fig. 171) con

recorder.

pipettes, P', P", and P"". In addition a leveling bottle A and connecting tube T are also provided.

The tube E is connected to the point in the flue at which the sample is to be taken. The instrument is first set in

FIG. 170.—A recorder for combustion operation.

From the discussion in the text it may be inferred that a knowledge of the carbon dioxide component of the flue gas enables us to judge concerning the combustion taking place in the furnace. The principle involved in the type of carbon dioxide recorder as shown is that a change of volume in a gas produces a change of pressure. A continuous sample of the flue gas enters at A and in passing through the absorption chamber the carbon dioxide is absorbed and consequently a reduction in pressure takes place. By the calibration of suitable manometer tubes the instrument may be made to read the carbon dioxide component direct.

operation by closing the stop-cocks f, g, and e, d being open. By lowering the leveling bottle A, a sample of the gas is drawn into the burette B. This preliminary sample is then

expelled to the atmosphere by raising the bottle A and allowing the gas thus put under pressure to pass out through a by-pass at d. This process is continued until it is considered that an average sample has been drawn into the burette B. The leveling bottle A is next lowered so as to cause the water in burette B to come to its zero mark. By raising the bottle A the water is again forced into burette B and the gas sample expelled through stopcock e into the pipette P', in which there is a chemical solution

The Orsat apparatus is a portable instrument contained in a wooden case with removable sliding door front and back, as shown in its simplest form in this illustration, taken from the report of the Power Test Committee of the American Society of Mechanical Engineers. It consists essentially of a measuring tube or burette, three absorbing bottles or pipettes, and a leveling bottle, together with the connecting tubes and apparatus. The bottle and measuring tube contain pure water; the first pipette, sodium or potassium hydrate dissolved in three times its weight of water; the second, pyrogallic acid dissolved in a like sodium hydrate solution in the proportion of 5 grams of the acid to 100 cc. of the hydrate; and the third, cuprous chloride. These chemicals are sold by most of the large dealers. Details of how this apparatus is used to determine the chimney gas analysis were set forth in a previous discussion.

that absorbs carbon dioxide, but will not absorb oxygen, carbon monoxide or nitrogen.

To Ascertain the Carbon Dioxide Content of a Flue Gas.Exactly 100 cc. of gas were originally drawn into the burette B. If now the leveling bottle A is again lowered to draw the gas back through stop-cock e, the volume in the burette will be found to have lessened in quantity so that instead of reading zero it now reads N1 which indicates directly the volume of car

bon dioxide that was present in the gas, for evidently this volume has been absorbed in the pipette P'. Hence, we have

To Ascertain the Oxygen Content of a Flue Gas.—In a similar manner the gas sample in the burette B is now forced through pipette P", in which is a solution that will absorb free oxygen in the sample but will not absorb carbon monoxide or nitrogen. By means of the leveling bottle A, the sample is next drawn back into the burette B and a reading N2 noted. It is now evident that the oxygen content of the flue gas may be computed from the formula

To Ascertain the Carbon Monoxide Content of a Flue Gas. The pipette P'" similarly contains a solution which readily absorbs the carbon monoxide present in the gas, but will not absorb nitrogen. Hence we proceed as in the two former instances and return the gas sample to the burette which now reads N3. Consequently the carbon monoxide which was present in the flue gas is obtained from the formula

To ascertain the Nitrogen Content of a Flue Gas.-We assume that all of the gas which remains in the sample is nitrogen. Consequently the nitrogen content is obtained from the formula

V1 = 100 (V1 + V2 + V3)

1

(4)

An Approximate Check on the Orsat Analysis.-Air is found by weight to have 76.85 per cent. nitrogen and 23.15 per cent. oxygen. By volume this analysis will be found to be 79.09 per cent. nitrogen and 20.91 per cent. oxygen. Since 1 unit by volume of oxygen forms 1 unit by volume of carbon dioxide in the burning of pure carbon the actual percentage of nitrogen in the chimney gases is not altered but should remain 79.09 per cent. if perfect combustion is maintained.

On the other hand, when imperfect combustion is under way, or in other words, when some carbon monoxide is being formed 1 unit by volume of oxygen forms 2 units by volume of carbon monoxide. Hence when pure carbon is the fuel, the sum of the percentages of carbon dioxide, oxygen, and 11⁄2 the carbon monoxide must be in the same ratio to the nitrogen present as the oxygen

in the air is to the nitrogen component, namely as 20.91 : 79.09. This is a convenient check upon a flue gas analysis in the progress of the experiment. Thus if an analysis of chimney gas is found to contain by volume 9.5 per cent. carbon dioxide, 10.2 per cent. carbon monoxide, 5.2 per cent. oxygen, and 75.1 per cent. nitrogen, according to this proportion, we should have

10.2
2

9.55.2+ : 75.1 20.91 : 79.09

Upon investigation this will be found to be approximately true and well within the limit of experimental accuracy.

As California crude oil contains usually about 11 per cent. of hydrogen, the ready checking above indicated proves of no avail since the hydrogen content is not taken account of in the Orsat or flue gas analysis. As the relationship serves, however, to clinch our ideas of volumetric proportions of entering air and outgoing flue gases, it is well to bear it in mind.

In boilers fired by coal containing little hydrogen the CO does not usually exceed 1 or 2 per cent. and the sum of the Orsat readings CO2 + 0 + CO is usually beween 20 and 21 per cent. When burning oil, on the other hand, the sum of these readings may be as low as 16 or 17 per cent. due to the large proportion of hydrogen in the fuel, which means an apparent nitrogen content of 83 or 84 per cent. The reason for this is that the water vapor formed by the burning of hydrogen condenses in the Orsat apparatus and occupies practically no volume, but the oxygen which unites with the hydrogen brings with it the same proportion of nitrogen as does the oxygen that unites with the carbon. Consequently the Orsat indicates a larger proportion of nitrogen than would occur if the fuel were pure carbon.

Chemical Formulas for Preparing the Absorption Solutions. The bottle A and the measuring tube or burette B contain pure