em is
necessarily hidden from view, and, further, it is not easily accessible.
The leakage taking place in any system, if there be any, must, however,
be detected and stopped.

Fig. 63 is given to illustrate a danger peculiar to the self-contained
oil system, in which the oil and oil-cooling chambers are situated
adjacently in the turbine bedplate. One end of the bedplate only is
shown; B is a cast-iron partition dividing the oil chamber C from the
oil-cooling chamber D. Castings of this kind have sometimes a tendency
to sponginess and the trouble consequent upon this weakness would take
the form of leakage between the two chambers. Of course this is only a
special case, and the conditions named are hardly likely to exist in
every similarly designed plant. The capacity of oil, and especially of
hot oil, to percolate through the most minute pores is well known.
Consequently, in advocating extreme caution when dealing with oil
leakage, no apology is needed.

[Illustration: FIG. 63]

It may be stated without fear of contradiction that the oil in a
self-contained system, namely, a system in which the oil, stored in a
reservoir near or underneath the turbine, passes only through that one
turbine's bearings, and immediately back to the storage compartment,
deteriorates more rapidly than when circulating around an "entire"
system, such as the gravity or other analogous system. In the latter,
the oil tanks are usually placed a considerable distance from the
turbine or turbines, with the oil-cooling arrangements in fairly close
proximity. The total length of the oil circuit is thus considerably
increased, incidentally increasing the relative cooling capacity of the
whole plant, and thereby reducing the loss of oil by vaporization.

The amount of oil passing through the bearings can be ascertained
accurately by measurement. With a system such as the gravity it is only
necessary to run the turbine up to speed, turn on the oil, and then,
over a period, calculate the volume of oil used by measuring the fall of
level in the storage tank and multiplying by its known cross-sectional
area. In those cases where the return oil, after passing through the
bearings, is delivered back into the same tank from which it is
extracted, it is of course necessary, during the period of test, to
divert this return into a separate temporary receptacle. Where the
system possesses two tanks, one delivery and one return (a superior
arrangement), this