up in proportion. One factor some- 
times overlooked is the fact that the 
difficulty and expense of handling ir- 
radiated samples increases rapidly with 
the flux density available for irradiations. 
Comparatively simple remote-han- 
dling equipment which is satisfactory 
for use in conjunction with a 1-Mvw re- 
actor may well be grossly inadequate 
to handle material from a 20-Mw 
facility. 
Safety Considerations 
Discussions of reactor safety gener- 
ally fall under three headings, (1) in- 
herent self-control in the reactor, (2) 
reliability and effectiveness of the 
safety and control system, and (3) ex- 
clusion area and containment. 
Self-regulation. There is little rea- 
son for building a research reactor that 
does not have a negative temperature 
coefficient. Thus all research reactors 
should be inherently self-limiting to 
some degree. The negative coefficient 
of reactivity is due largely to expansion 
of the moderator although expansion of 
the fuel element and increase in the 
mean thermal neutron temperature 
(determined by the temperature of the 
moderator) contribute to the effect. 
The self-limiting is of greatest impor- 
tance in protecting the reactor against 
unforeseen actions that will cause sud- 
den positive excursions of reactivity, 
and hence the time constant of the 
moderator is as important as its magni- 
tude. 
The homogeneous reactor has an 
ideal response in that the moderator 
temperature follows the fuel temper- 
ature practically instantaneously and 
there is no time lag in the reduction in 
reactivity. The danger in the homo- 
geneous reactor appears to:be the pos- 
sibility that the containing vessel may 
rupture after sudden applications of 
positive reactivity. 
On the other hand, the formation of 
steam between the plates of the hetero- 
geneous reactor is delayed until the heat 
released in the plates is conducted into 
the water. After a large step increase 
in reactivity, the high temperature 
gradients necessary for a high conduc- 
tion rate may allow the plates to reach 
the melting point before sufficient 
steam is formed to shut down the reac- 
tor. In the case of MTR elements 
immersed in light-water moderator 
(4) the step change required to cause 
melting appears to be of the order of 3- 
4% in 6k/k. 
Scramming controls. The safety 
circuits operate to ‘‘scram” the reac- 
tor if the neutron flux rises above a pre- 
determined level or if the period be- 
comes shorter than is considered safe. 
The period protection usually extends 
four or five decades below full power. 
Reliability is obtained by careful choice 
of components and circuits and by 
duplication of equipment. In addition 
interlocks can be provided to scram 
the reactor if a potentially dangerous 
condition should develop in the experi- 
mental equipment. 
Exclusion and containment. It is 
conceivable that both the self-limiting 
features and the safety circuits might 
fail to protect the reactor against a 
positive reactivity excursion and that 
subsequent melting of the fuel elements 
would release radioactive fission prod- 
ucts into the reactor building. To 
protect the general public against this 
eventuality, the reactor is surrounded 
by an exclusion area or the reactor 
building is made sufficiently gas tight 
to contain the radioactive products. 
C. R. McCullough has summarized 
the problem of guarding against reactor 
accidents and the general philosophy 
with which the U. S. Advisory Com- 
mittee on Reactor Safeguards investi- 
gates a proposed reactor installation 
(10). Several years ago this committee 
suggested the empirical relation. 
R = 0.01 VP 
where R is the radius of the exclusion 
area in miles and P is the reactor oper- 
ating power level in kilowatts. This 
relation postulates little containment 
and is really applicable only to high- 
power reactors. The exclusion dis- 
tance determined supposedly provided 
sufficient time to warn and evacuate 
personnel downwind when the wind 
was light and, if it was strong, allowed 
time for the cloud to diffuse so that the 
total exposure was not too great as the 
radioactive cloud was blown over a 
person at the edge of the exclusion 
area. Today more detailed calcula- 
tions are desired and the rule-of-thumb 
given above is useful only as a first 
approximation to be employed during 
the initial examination of possible re- 
actor sites. It must be recognized 
that under some extreme meteoro- 
logical conditions a person at the edge 
of an exclusion area thus determined 
can receive a severe overdose of gamma 
radiation. 
More accurate parameters in terms 
of various atmospheric conditions have 
been published (1/1) and permis- 
sible ingestion and inhalation rates 
set up. The three most toxic radio- 
active isotopes are U2, Pu239, and Sr9° 
for which the tolerance doses* are 
1,690, 32 and 1.3 micromicrograms/ 
cubic meter respectively. The fatal 
dose* is some 10,000 times this in each 
case. If the installation is safe from 
the viewpoint of these three sub- 
stances, the concentrations of all other 
radioactive isotopes will be well with- 
in permissible limits. An exception 
would be the hypothetical case where 
these three isotopes were continuously 
removed and all others allowed to 
remain in the reactor. 
Frequently a combination of. exclu- 
sion area and building containment is 
the least expensive solution to the 
problem. The building must be suff- 
ciently strong to withstand the effects 
of the largest reactor power excursion 
that can be reasonably postulated and 
sufficiently gas tight so that personnel 
outside the exclusion area will not re- 
ceive an excessive dose before they can 
be evacuated. The provision for some 
exclusion area relaxes the leakage re- 
quirements for the building and may 
result in appreciable cost saving. 
BIBLIOGRAPHY 
1. L. D. P. King in ‘International Conference 
on Peaceful Uses of Atomic Energy,” vol. 2, 
p. 372 (United Nations, New York, 1955) 
2. R.H. Graham, D. G. Boyer. Reactor safety 
experiments, Nucueonics 14, No. 3, 45 (1956) 
3. W. M. Breazeale. The swimming pool—a 
low cost research reactor, NuciLeonics 10, 
No. 11, 56 (1952) 
4. J. R. Dietrich in ‘‘ International Conference 
on Peaceful Uses of Atomic Energy,"’ vol. 
13, p. 88 (United Nations, New York, 1955); 
W. E. Nyer, et al. Transient experiments 
with the SPERT-1 reactor, NucLeonics 14, 
No. 6, 44 (1956) 
5. A. M. Weinberg et al. in ‘International Con- 
ference on Peaceful Uses of Atomic Energy,” 
vol. 2, p. 402 (United Nations, New York, 
1955) 
6. W. H. Zinn. ANL reactors, in ‘‘Inter- 
national Conference on Peaceful Uses of 
Atomic Energy"’ vol. 2, p. 456 (United 
Nations, New York, 1955) 
7. ORNL memo CF-55-8-201 
8. D. J. Hughes, ‘‘Pile Neutron Research,” 
(Addison-Wesley Publishing Co., Boston, 
1953) 
9. C. D. Bopp, O. Sisman. How radiation 
changes polymer -mechanical properties, 
Nucveonics 13, No. 10 (1955) and references 
at the end of this paper 
10. C. R. McCullough, Mark Mills, Edward 
Teller in ‘International Conference on 
Peaceful Uses of Atomic Energy," vol. 13, 
p. 79 (United Nations, New York, 1955) 
11. J. E. Holland. Meteorology and atomic 
energy (Weather Bureau, U. S. Department 
of Commerce, Washington, D. C., 1955) 
* Tolerance dose is the maximum level 
that can be tolerated every day for 8 hr, 
equivalent to 0.043 rem/day. Fatal dose 
is that which gives only 50% survival if 
the dose is acquired within an 8-hr period. 
73 
