The very small units, ranging from 10 to 20 h.p., work best on kerosene, around 44° Baume gravity. These engines are seldom loaded up to full capacity, and so there is little likeli hood of serious preignition, which will occur on full load with kerosene. The kerosene vaporizes very readily, and the bulb can run at a low temperature. This is an advantage as the small engines usually are started and stopped several times each day. It might be well to call attention to the trouble that is experienced in large engines pulling heavy loads while using kerosene. In vestigations covering over one hundred installations where kerosene had been burned revealed that in every instance severe cylinder cutting occurred. It is impossible to definitely deter mine why this condition should exist. In a number of engines 36° Baume distillate had been used with entire satisfaction, but on changing to 44° Baume kerosene cylinder cutting immediately became apparent.
The various make engines show quite different operating char acteristics with oils of practically the same gravity. Table XVI is the result of a series of tests on a 25 h.p. Fairbanks Morse Type Y engine. This engine has a hot combustion cham ber and operates without water injection.
These tests, while not entirely conclusive, indicate that an oil with a higher gravity can be burned without preignition in a water-injection engine than in a "dry" engine. Due to this comparative absence of preignition in the water-injection en gine, there is also less tendency to "hunt." On the heavier oils both engines displayed a smoky exhaust. The exhaust of the water-injection engine always has a dark color accompanied by an oily or tar-like fog, which is, of course, a result of a com bination of the steam and heavy hydrocarbons. So objection able is this oily exhalation that it is never advisable to run a stream of water into the exhaust line, even though this does serve to deaden the noise.
Fuel Storage.—It is seldom profitable to purchase oil in barrel lots when the yearly consumption exceeds 8000 gallons. With an annual demand beyond this figure the difference in the car load and barrel prices will pay for a storage tank. When the yearly requirements are less than two tank cars, the best storage is the vertical corrugated steel tank made of No. 16 gage galva nized sheets. These tanks are not costly and serve their purpose admirably. If leaks develop at the rivets, a little soap will seal the openings until the seam can be soldered when the tank is emptied. There is no danger of an explosion when soldering,
for, unlike a gasolene tank, the small amount of vapor fumes speedily dispel after the tank cover is removed. Red lead is useless as a temporary seal since the oil causes the lead to thin and flow. In the Northern states a cypress storage tank can be employed with success. The cypress tank is lower in price than is the steel one; in the Southern states the oil evaporation from a wood tank is too excessive to justify its utilization. With the larger units, ranging from 50 h.p. upward, the choice of storage lies between the horizontal steel tank, constructed of boiler plate, and the vertical concrete tank. In the majority of installations the steel tank is more advisable since it is difficult to secure the services of a concrete contractor who is able to con struct a leak-proof concrete tank.
Fuel Consumption.—An average of the guarantees of various builders of low-pressure oil engines gives the following values for fuel consumption: Full load Three-quarters load One-half load .651b. .68 lb. 1.00 lb.
These figures are based on the use of fuels from 32° to 44° Ratline. On factory tests or on tests conducted by an engineer who is thoroughly conversant with the peculiarities of the engine tested, these values can usually be more than equaled. However, in actual operation, the engineer cannot hope to better the builder's guarantee; indeed, he is fortunate if he equals it.
Figure 350 is the result of a factory test run on a Bessemer oil engine No. 16668 of 85 h.p. rating. The standard guarantee of this concern is 'Ho pint or from .6 pound to .7 pound', dependent on the weight of the oil per gallon. The test curve shows the full-load fuel consumption to have been .57 pound, while the half load value was .63 pound. This half-load fuel consumption is rather remarkable. Ordinarily, the net mechanical efficiency of a two-cycle low-pressure engine is around 85 per cent. at full load and approximately 70 per cent. at half load. Naturally it would seem that at half load the fuel consumption must be at least of the full-load result. The half-load efficiency would be still less, due to the lower cylinder temperature. That the fuel consumptions, in this case, are practically identical can evidently be attributed to the higher thermal efficiency at half load; the oil vapor and air mixture was leaner than at full load— it has been proved many times that with a lean mixture the thermal efficiency is higher than with a rich mixture.