182 
It is important to remember too that 
in the design of the fuel package, and 
in fact in the design of all parts ofa 
pile, reliability must be achieved toa 
degree which transcends most of our 
experience. This is the result of 
possible radiation hazards from fail- 
ures, and the extreme difficulty of 
inspection and repair of a “‘hot’’ (in 
the radioactive sense) pile. 
In addition to the problem of fuel 
design, there are many other features 
of a nuclear reactor for power pur- 
poses which pose some nice questions 
for engineers and scientists. ‘These 
are not necessarily less important or 
difficult than the fuel problems, and 
involve many of the general con- 
siderations and details mentioned in 
the foregoing. Some of these are 
reviewed briefly in the following 
paragraphs. 
Heat transfer fluid.—The selection of a 
fluid to carry the heat from the nuclear 
reactor to the heat exchanger in which 
it may be converted to steam or a hot 
gas is a vital matter. Gases, liquids, 
and liquid metals all must be con- 
sidered. The medium must have 
low neutron absorption. It must be 
reasonably stable under intense radia- 
tion. It must have, of course, good 
heat-transfer characteristics, must be 
noncorrosive, preferably nontoxic and 
nonexplosive. Availability and cost 
also are important. 
If the medium is to be allowed to 
boil in the reactor, then the neutron- 
absorption characteristics of the vapor 
must be approximately the same as 
those of the liquid. Otherwise a sud- 
den flashing to vapor of a part of the 
liquid might result in such a sudden 
change of nuclear characteristics in 
the pile that the control mechanism 
could not compensate for it. 
Radiation shield ——The shield for pro- 
tection of personnel from the lethal 
neutron and other radiations is an 
absolutely necessary part of any power 
pile. If research and development 
can reduce considerably the volume 
and weight of shield required, it will 
be very helpful in facilitating the ap- 
ANNUAL REPORT SMITHSONIAN INSTITUTION, 1948 
plication of atomic power plants to 
ships, and elsewhere where weight and 
size are of great importance. 
Control—The reliability of the con- 
trol system must be absolute, so that 
any failure must result in safety rather 
than permit a runaway. It must be 
able to handle all possible variations 
in reactivity of the pile, either normal 
or accidental, gradual or fast, such as 
result from the slow consumption of 
the fuel, or the sudden loss of part or 
all of the heat-transfer fluid. It must 
be effective at all times, under any 
possible set of conditions. It must 
make it impossible for the “nuclear 
boiler” ever to explode. These re- 
quirements are rigid, and must be met. 
AN ATOMIC POWER SYSTEM 
Of what factors might the land- 
based atomic power system of the 
future be comprised? In all the dis- 
cussion on the subject the fact is 
implicit that at present no possibility 
is foreseen for obtaining usable electric 
power directly from an atomic pile. 
The atomic energy will appear as heat 
which, when converted into steam or 
hot gas, will feed conventional turbo- 
generators. In other words, from the 
steam-supply pipe on to the consumer 
the system will be similar to present 
systems using coal or oil. The nuclear 
reactor and heat exchanger, however, 
will replace the boiler. There will be 
involved components comparable in 
function to present fuel and ash trans- 
portation, storage, preparation, and 
disposal installations, but vastly more 
important from economic and security 
points of view. 
The system may inculde both “‘pri- 
mary” and “‘secondary”’ reactors. By 
a primary reactor is meant one which 
produces not only power, but fission- 
able material as well. The reactors 
currently operating at Hanford, in- 
asmuch as they produce heat (at 
present dissipated in the Columbia 
River) and plutonium, are by this 
definition primary reactors. Under 
the international control plan proposed 
