Popular Science Monthly/Volume 24/February 1884/Dangerous Kerosene and the Methods for its Detection




KEROSENE, in virtue of its cheapness and the brilliant light it gives, has found its way into almost every house. And yet frequent and often horrible accidents prove that much of the oil now sold is of a most dangerous character. It is the recognized duty of the State to render the sale of such oil impossible by proper inspection. Almost daily reports of loss of property and life, as the result of the use of unsafe kerosene, show, however, that this official control fails to effect its object. This may be due, in a measure, to the undoubted negligence of cities and towns to appoint competent inspectors—if, indeed, any appointment is made—or to the carelessness of the inspectors; but of greater importance even than this are the low standards adopted, and the unreliability of the tests which are used to determine the character of the oil.

It is the object of this paper to consider the conditions of safety in an oil used for illuminating and heating purposes, and to give a brief sketch of the principal methods which have been proposed for determining this important point.

Petroleum, from which kerosene is prepared, is, as is generally known, a mixture of a large number of intimately related compounds of widely differing volatility. Some are gaseous, and escape in this form as the petroleum issues from the ground, while others form the solid paraffine. The middle portions of the crude oil are separated from the more and less volatile compounds by distillation, and after a further process of purification go into the market as kerosene. The entire removal of the lighter and more volatile portions, which are known as naphtha and benzine, is of the utmost importance, for it is in their presence that the danger lies. Alone, they are easily ignited, and alone or mixed even in small proportion with kerosene, they readily emit vapors which are inflammable and which with air form an explosive mixture.

An oil is safe only when it will not yield these dangerous vapors at any temperature which it is liable to assume. This temperature depends obviously (1) upon that of the place where the oil is kept or used, and (2) upon the influence of the heat of the burning wick in warming the oil in the reservoir of the lamp. As the result of carefully conducted experiments with lamps of different patterns, it has been found[1] that the maximum increase of temperature of the oil in a burning lamp is some 16° Fahr. (9° C). Before the lamp is lighted the oil in it will in most cases have the temperature of the air about it. Our rooms in summer often have a temperature of 90° Fahr., and reach 100° Fahr. in a few exceptional days, while in winter the oil assumes even a higher temperature than this when the lamp is placed—as it often is—near a stove or an open fire.

Hence, it is plain that the lowest temperature at which an oil may evolve inflammable vapors and be considered safe must be put at 116° Fahr., or better still at 120° Fahr.

What, now, are the means for determining the temperature at which these vapors appear, and thus for deciding upon the safety or danger of an oil? It seems at first thought a simple and certain matter. Put a little oil in a cup and suspend a thermometer in it; warm it slowly, and, as the temperature rises from degree to degree, pass a lighted match just above its surface. Presently the match will cause a tiny explosion. This indicates that the dangerous vapors are appearing, and the thermometer now gives the so-called flashing-point of this oil. Go on heating and testing as before, and at last the oil will take fire and continue burning by itself. The mercury is now at the burning-point. But repeat the experiment with fresh samples of the same oil, and you will find that a trifling variation in the conditions will alter the flashing-point to a wonderful extent. The quantity of oil used for the test, the rate of heating, and the range of temperature through which the oil is heated, the distance above the surface at which the match passes—each and all have a marked influence on the determination.

The burning-point—or fire-test, as it is often misleadingly called—is of little value; for not only does it always lie above the flashing-point—which is the real danger-point—but it bears no simple relation to the latter, so that its determination gives really no clew to the temperature at which the oil becomes unsafe.

The unreliability of this simple method of testing and the importance of the problem have called forth numerous suggestions for improvement. Within the last fifteen years no fewer than twenty-five different instruments have been proposed, presenting as many more or less widely modified forms of the simple cup-tester indicated above. The most essential variations are (1) in the size and form of the oil-holder or cup, which in some apparatus is open, in others partly or wholly closed; (2) in the dimensions of the water-bath—which is invariably employed in all as the best means for communicating a slow and uniform increase of temperature to the oil; (3) in the means used for igniting the vapor—a burning match, waxed thread, small gas-jet, electric spark, or little oil-lamp standing on the cover of the oil-cup being the chief devices for this purpose.

But, notwithstanding all the ingenuity displayed, and the elaborate and costly apparatus to which it has in some instances given birth, we find Engler and Haass,[2] at the close of a careful investigation into the reliability of petroleum-testers, in which all the more promising methods were laboriously examined and compared, laying down these general principles, which are to be observed in the construction and use of this class of testers:

1. The quantity of oil must be the same in all experiments.—In the Saybolt tester, for instance, which was adopted in 1879 by the New York Produce Exchange (chiefly, however, for the purpose of determining the burning-point), variations of one millimetre, or about one twenty-fifth of an inch, in the height of the oil, cause differences of some degrees in the flashing-point.

2. The oil must be heated slowly and uniformly.

3. The temperature of the oil at the beginning of the test must be at least 18° Fahr. (10° C.) below its flashing-point (which is approximately determined by a preliminary test). Hence, a low-grade oil, which flashes not far from the air temperature, must be cooled down before an accurate determination can be made.

4. The size and intensity of the flame or spark used to produce the flash must remain unchanged in all tests. Increase in size or intensity lowers the flashing-point.

5. The distance of the flash-flame or spark from the surface of the oil must be the same in all tests. The flashing-point is lowered by decreasing this distance. Care must be taken that this distance is not so small that a local evolution of vapor from the surface occurs.

6. The time during which the flame or spark acts must be reduced to a minimum, increase in the time causing a sensible lowering of the flashing-point.

7. On account of the practical purpose for which the tests are made, the conditions under which the vapor is formed in the tester should correspond as closely as possible to those which determine its formation and explosion in lamps, etc.

Comment upon methods which depend for trustworthy results upon such a formidable array of conditions is hardly necessary; the best apparatus must be electrical and costly, and even then unreliable except in the hands of an expert. We are not surprised to find Mr. A. H. Elliott, in his report of a similar investigation ordered by the New York State Board of Health, giving as his general conclusion: "Of all the apparatus examined, not one can be called perfectly satisfactory. . . . Of the electric testers it may be stated, that any advantage obtained from the use of electricity is more than overcome by the trouble necessary to maintain the galvanic battery and induction-coil." But, even if the performance of some of these instruments is such as to yield concordant results, when all the precautions are carefully heeded, these results can have only a relative significance, and agreement of different testers can only be secured by selecting one with its manipulation as an arbitrary standard, and adopting conditions in the others which shall give corresponding results. Nor can it be affirmed that all the conditions under which explosions in lamps are liable to occur are provided for in any single instrument

Fig. 1.—The Saybolt Tester.

of this class. The oil-reservoirs of our lamps differ much in size and shape, and hence have different capacities. Moreover, the quantity of oil, its surface, and the amount of air in the reservoir with which the vapor mingles, are constantly changing while the lamp is in use and the danger greatest. Again, it is not alone in quietly burning lamps that accidents occur. Probably half are due to upsetting or breaking, and the oil, which would have been safe otherwise, gives rise to explosion or flames under these more dangerous circumstances.

If it is important to test the oil, it certainly is wise to employ, if possible, a test which shall indicate the lowest temperature at which, under any conditions, inflammable vapors can be evolved, and not to trust to a method which merely proves an oil safe under certain arbitrary conditions.

Besides these instruments which aim at a direct determination of the temperature at which an oil becomes dangerous, others have been proposed in which the character of the oil is tested in an indirect manner, by finding the elastic force or tension of its vapor at a given temperature. The tension is measured by the height of the column of water which it sustains. By comparing the tension which any oil gives in this apparatus with that of some kerosene which has been selected as a standard, the quality of the former is ascertained—a higher tension indicating a more dangerous oil. It is plain that the reliability of this method depends upon the assumption that a definite relation exists between vapor-tension and flashing - point in all kerosenes. It has, however, been shown in the most conclusive manner, that this is not the case.[3] Four different oils, which all had a flashing-point of 28·5° to 29·5° C, as determined by one of the most trustworthy of the testers before described, were found to give, at 28° C, vapor-tensions of 75, 104, 118, and 168 millimetres (of water); and, at 40° C, tensions of 126, 149, 165, and 201 millimetres. Further, seven different kerosenes gave, when tested by the two methods, the following results:

OIL. 1. 2. 3. 4. 5. 6. 7.
Flashing-point 25° C. 26° C. 26° C. 28° C. 30° C. 44° C. 48° C.
Tension at 35° C. 95mm. 160mm. 201mm. 73mm. 45mm. 13mm. 5mm.

It thus appears that the results obtained by the measurement of the vapor-tension are quite worthless as indications of the dangerous character of kerosene, and the method must be regarded as far less reliable than even the imperfect ways of testing which have been already discussed.

The uncertainties of the foregoing methods are entirely avoided by a distillation test, which also enables one to decide the quality of the oil as an illuminating material, and thus gives the fullest information in regard to its nature.[4] The oil is separated by the distillation into three fractions: a light oil distilling below 150° C.; illuminating oil coming over between 150° and 270° C.; and a heavy oil which boils above 270° C. The first fractional distillate represents the dangerous constituents, and should not exceed, according to Bielstein, five per cent of the whole. The heavy oil affects the freedom with which the kerosene burns in a lamp, and, in American kerosene, should not form more than fifteen per cent of the oil. The operation must be conducted with care, in a flask provided with a dephlegmator, and the fractions, as well as the original sample, must be weighed. These circumstances are likely to prevent the general adoption of a method which is otherwise so simple and satisfactory, and kerosene will probably be tested in the future, as now, by the determination of its flashing-point.

In 1879 Victor Meyer[5] suggested a principle by which the minimum, or, as he called it, "true or absolute" flashing-point, could be determined. It is to saturate air with oil-vapor at the test-temperature. His method is simply this: A glass cylinder of about 200 c. c. capacity is partly filled with oil, stoppered with a cork through which a thermometer passes, and heated by plunging into warm water; when the temperature is reached at which the test is to be made, the cylinder is briskly shaken, the stopper removed, and a small flame introduced. Flashing-points obtained by this plan are considerably lower than those given by the methods which have been discussed, and are found, moreover, to be largely independent of the conditions so essential to success in the latter.

Haass[6] has described an elaborate and clever apparatus based on the same principle, and differing essentially from Meyer's only in the substitution of an electric spark, at a fixed distance from the surface of the oil, for the flame which the latter employed. In both of these methods the flashing-point depends upon the time allowed between the shaking and testing, Haass recommending an interval of one minute after the bubbles have disappeared from the surface of the oil, in order to permit the suspended oil-particles to settle. The shaking, which must be repeated from degree to degree, is a troublesome feature of these methods, and, though Meyer's apparatus is certainly simple and inexpensive enough, that of Haass is difficult of construction, electrical, and costly. The general principle of these methods is, however, without question the correct one for obtaining a minimum (and approximately "absolute") flashing-point, and it is to L. Liebermann[7] that we owe the suggestion of an ingenious and successful plan for

Fig. 2.—Liebermann's Tester.

avoiding the difficulties mentioned above. In Liebermann's method the saturation of air with vapor is accomplished by forcing an air-current through the oil as it is warmed from degree to degree; and the test made by bringing a small flame to the mouth of the oil-holder at the same instant.

It has, however, been shown that the intermittent current of air which is recommended gives somewhat irregular results, and that more concordant flashing-points are obtained by letting a continuous current run through the oil for at least one minute before the flash occurs. It may perhaps seem, at first thought, that a continuous current of air would dilute the vapor to such an extent that the flashing-point must be materially raised, and that this effect must be more marked as the velocity of the current is increased. This is, however, not the case. On the contrary, while a slow, continuous current raises the flashing-point appreciably, a sufficiently rapid one gives nearly the same results as the intermittent method; nor does any further increase in the velocity alter the flashing-point to a sensible extent. It has indeed been found that a large dilution of kerosene-vapor with air is necessary to furnish the conditions for the most violent explosion; and these conditions are also those for the readiest flash by this method of testing. The most explosive mixture, according to Chandler, is formed by nine parts of air to one of vapor. The passage of a large quantity of air through the oil tends, of course, to make the flashing-point higher, by carrying away with it the more volatile portions which determine the flash, and this effect is greater when the quantity of oil is small and the air-current long continued. It is, consequently, necessary in the employment of this method to know the minimum quantity of oil and the maximum duration of air-current which will permit concordant results. These limits have been ascertained in a recent investigation,[8] the results or which are given a little further on.

A tester of still simpler construction than that of Liebermann has also Fig. 3.been proposed.[9] It consists, as shown in the cut, of a glass cylinder, closed at one end by a cork, through which a small bent tube, d, c, b, passes. Just within the cork the end of this tube contracts to a small orifice. The other end of the tube connects with a small bellows, or other source of slightly compressed air, the flow of which can be regulated by the pinch-cock e.

Experiments made with cylinders of different dimensions have shown that the best results are obtained when the diameter is between 2·5 and 4 c.m. The length (if only great enough to allow at least the minimum quantity of oil to be used) makes no difference. Cylinders of the same diameter but of different lengths, when filled with oil to within the same distance from the top, all give the same flashing-point. Change in length in such cases is simply equivalent to change in the quantity of oil employed in the test, and it has been proved that the quantity of oil does not affect the determination when it is above a certain minimum.

The distance of the oil, or rather of the foam into which the surface is broken by the air-current, from the top of the cylinder, however, makes a considerable difference in the results—the flashing-point falling as this distance is decreased, until at about 5 to 6 c. m. it reaches a minimum.

These considerations lead to the following statements and directions for the use of this method:

1. The oil-cylinder should have a diameter of 2·5 to 4 c. m. It may be of any convenient length, provided it holds, when filled, for the test not less than 50 c. c. of oil. With a diameter of 2·5 c. m., the length should be at least 16 c. m.; with a diameter of 3 cm., the least length should be 13 c. m. A good tester may be made from the chimney of a student-lamp, by cutting off the lower part, a little above the contraction. (Glass is easily cut by filing a deep notch at one point, and letting a little gas-flame play slowly back and forth across it in the line of the proposed section, until a crack springs quite through the glass; this crack can then be led in any desired direction by keeping the little flame just ahead of it on the glass.) The whole chimney may also serve as an oil-cylinder by corking the large end. The irregularity of shape at the bottom does not affect the results; but the length makes it rather inconvenient by requiring a correspondingly deep water-bath.

2. The cylinder is filled with oil to a point such that, when the air-current is running, the top of the foam is 4 or 6 c. m. below the mouth.

3. The oil is heated by means of a water-bath, into which the cylinder is plunged to the level of the oil. The temperature of the oil should not rise faster than two degrees a minute.

4. Air is forced through the oil with such velocity that about (and not less than) 1 c. m. foam is maintained on the surface, and a flash-jet brought to the mouth of the cylinder every half degree, or oftener in the vicinity of the flashing-point. The approach of the flashing-point is announced by the appearance of a faint blue halo of burning vapor around the flash-jet; this finally detaches itself and runs down to the surface of the oil, and the reading of the thermometer at this instant gives a trial flashing-point, which may be a little too high if the air-current has been running too long, or not long enough.

The test is now repeated with a fresh sample of the oil, and the air-current started in full strength not less than one nor more than three or four minutes before the flash occurs. It is a good plan, however, to let a very slow current of air bubble through the oil from the time that the tester is put in the water-bath, so as to secure regularity in the heating of the oil.

A very good flash-jet is a little gas-flame from the tip of a blowpipe, or glass tube drawn out to a point.

The advantages of this method are:

1. Simplicity of apparatus. It can be made in a few moments by any one who can bend a glass tube.
2. Simplicity of manipulation. A manufacturer, asked to try it, obtained concordant and accurate results at the first trial.
3. Trustworthiness of the results, which are independent of arbitrary conditions, and have a significance wholly wanting in methods based upon other principles. The flashing-point determined is the lowest.

The lowest flashing-point for illuminating oils in New York is fixed by law at 100° Fahr., and this is as determined by a modification of the Wisconsin tester, an instrument which demands all the precautions so emphatically given by Engler and Haass. In Massachusetts, method and flashing-point are apparently both left to the wisdom and discretion of the inspector.

And yet we have seen that 116° Fahr. is the very lowest flashing-point consistent with safety, and this should mean the minimum flashing-point determined by some fully reliable method. We must not be misled in this matter by the statement that our best kerosenes are "150° or 160° test" oils; for this has reference, not to the flashing-point, but to the fire-test, or burning-point, which, as has already been shown, gives little indication of the character of an oil. The best oils sold flash at about 109° Fahr., while the cheaper grades have much lower flashing-points—at least as low as 85° Fahr.

We need not be surprised, then, at the numerous accidents; they will not diminish until a much more efficient and intelligent system of inspection is enforced than now. We are far too much inclined to take our risk, even in the midst of constant warnings; we leave our kerosene to the ignorant and careless handling of our servants; we buy, perhaps, a cheaper grade from motives of economy, only to find that the oil in which we thoughtlessly trusted has occasioned loss of property, horrible suffering, or even death.

As long as unsafe kerosene is offered for sale, we may be sure purchasers will be found. The only safe way is to banish the dangerous stuff from the market.

  1. "Zeitschrift für anal. Chem.," xxi, 332.
  2. "Zeitschrift für anal. Chem.," xx, 1.
  3. Engler and Haass, loc. cit.
  4. "Zeitschrift für anal. Chem.," xxii, 313.
  5. Wagner's "Jahresbericht," 1879, 1175.
  6. "Chem. Industrie," 1880, 123, and "Zeitschrift für anal. Chem.," xx, 29.
  7. "Zeitschrift für anal. Chem.," xxi, 321.
  8. "American Chemical Journal," vi, No 1.
  9. Ibid, iv, No. 4, 285, and "Ber. d. Deutschen chem. Gesell.," xv, 2555.