Measurement of the volatility of motor fuels by batch distillation is regarded by the author as unsatisfactory, because the carbureted fuel is vaporized in an internal-combustion engine by continuous distillation, hence there is great difficulty in correlating the temperatures of test with those of actual use. Whereas formerly gasoline was produced by batch distillation in the refinery, it is produced now by removing the gasoline from crude oil by continuous distillation or is produced by cracking and continuous fractionation. Therefore the temperatures of production also bear no rational relation to those of test by batch distillation. Similarly, in an engine, fresh gasoline is supplied continuously by the carbureter and is vaporized continuously in the manifold and cylinder, all of the constituents being present at any time in any cross-section of the manifold.
To determine whether or not volatility can be measured by a continuous-distillation method, and temperatures of such test translated readily into those of production and use and the practicability of such tests for routine use, simple laboratory test-apparatus was constructed and test runs were made with various fuels. The apparatus is illustrated and described.
Results of the tests show two marked differences from results of batch distillation. The temperature range of distillation is much smaller and the percentage of bottoms, or unvaporized fuel, shows approximately a linear relation with temperature. With the gasolines tested, practically nothing was vaporized below 160 deg. fahr. and all of the gasolines were vaporized completely below 320 deg. fahr. It is held that if this laboratory method is truly representative of the process of distillation in an engine, it should be possible to translate the temperatures of test into those of use if an estimate can be made of (a) the reduction of hydrocarbon-vapor pressure due to the presence of air and (b) the effect of pressure on the temperatures of vaporization. A simple method of estimating actual partial pressures of hydrocarbon vapor in an air-gasoline mixture from the molecular weight of the gasoline, the ratio of weights of air and gasoline in the mixture and the total pressure of the mixture in the manifold is given.
Tests made with the apparatus when fitted with a bubble tower apparently confirm the surmise that the overhead temperature of a fractionated vapor-stream will correspond closely with the temperature of zero bottoms when the tower stream is free from steam and liquid entrainment and that the effect of steam can be corrected with reasonable accuracy. A test made with a blend of equal volumes of benzol, zylol and toluol indicates that fractionation in the apparatus is practically negligible, which may account for the linear relations found with gasolines. Estimation of the volatility of blends is comparatively easy by mathematical computation from continuous-distillation curves of the blended materials.
The author believes that simple laboratory continuous-distillation methods should receive serious consideration and that the proposed test method should give rise to the development of a definition of gasoline volatility that will be more acceptable to producer and user than methods now in use.