lattice but a smaller critical mass, the 
actual ratio depending on the effective- 
ness of the reflectors in use. However, 
at a given power level the light-water 
type provides 5-10 greater fast-neu- 
tron flux. This is a very important 
consideration in the planning of radi- 
ation-damage experiments. The neu- 
tron lifetime in a heavy-water reactor 
is perhaps four times that in the light- 
water type. Thus the period resulting 
from a given increase in prompt reac- 
tivity is proportionally longer and the 
transient correspondingly less severe. 
This is regarded as a safety factor. It 
must be added, however, that this fac- 
tor is offset to a considerable degree by 
the fact that a given physical disturb- 
ance in a heavy-water reactor (flooding 
a beam hole, etc.) has a much greater 
effect on reactivity than in a light-water 
assembly. 
Neutron Flux 
The relation between the average 
flux and reactor power in a thermal re- 
actor is given approximately by 
P=5 X 10 darG 
Where P is reactor power in watts, da» 
is average thermal flux in neutrons/ 
em?/sec, and G is actual reactor loading 
in grams of U-235. 
The ratio of the peak to average flux 
in the core depends on core geometry, 
reflector material, etc. but is generally 
1.5-2 to 1. It must be borne in mind 
that in lightly loaded reactors, such as 
the water boiler or CP-5, insertion of an 
Average 
Power thermal flux 
Designation (kw) (n/cm?/sec) 
~ 10 x10" 
=~ 1.25 x 10” 
Water boiler 
Open-pool, 
10-plate 
assemblyt 
Open-pool, ~14 x10" 
10-plate 
assemblyt 
Tank, 
10-plate 
assemblyt = 
Tank, 
19-plate 
assemblyt + 
1.3 10" 
1.25 x 10" 
TABLE | - Characteristic Reacto 
Critical 
mass 
Moderator Reflector | (kg) (kg) 
Light water 
Light water 
Light water 
Heavy water 
Light water 
absorbing sample of appreciable size in 
the region of maximum thermal flux 
will substantially reduce this flux. 
Fuel Requirements 
Limitation of enrichment of export 
material to 20% increases the U2** re- 
quirement relative to that needed if 
material enriched to 90% is available. 
As far as the homogeneous reactor is 
concerned, the increase is not great; the 
resonance escape probability is about 
0.95 and the thermal capture cross sec- 
tion of U?*8 is only 2.5 barns. How- 
ever, the increased concentration of the 
solution makes it even more strongly 
corrosive. 
In the case of the heterogeneous 
(MTR-type) fuel elements the U?**'re- 
quirements depend strongly on the fuel- 
element structure. The uranium- 
oxide fuel elements used in the Geneva 
reactor contained about 50 wt % of U, 
but this design has never been put into 
further production in this country. At 
present only Al-U-alloy fuel elements 
are being made and the maximum 
amount of U in fuel elements manufac- 
tured to date is about 25 wt %. Asa 
result of the weight limitation and the 
enrichment constraint, the present ex- 
port elements can contain only 5 wt % 
of U?%®, The added Al to carry the 
U in the elements increases the Fermi 
age of the core as well as increasing 
parasitic capture. The extra U235 
needed to raise the thermal utilization 
sufficiently to offset the greater fast 
leakage appreciably increases the criti- 
cal mass. 
(20% - Enriched Fuel) 
Total 
ence 
Graphite 
Light water 
Graphite 
Graphite 
Beryllium 
*Total requirement is the amount of fuel necessary to fuel the reactor and to compensate for xenon 
and samarium poisoning, temperature variations, the presence of experimental equipment, etc., in 
addition to 15% burnup. 
tAlloy cores of fuel plates have 40 wt % U. 
tThese are similar to CP-5 and MTR respectively except that the CP-5 and MTR use fully enriched 
fuel and thus require smaller critical masses. 
However, there is good reason to 
believe that in the near future Al-U- 
alloy fuel elements containing approxi- 
mately 40 wt % U (10 wt % U5) will 
be available forexport. A lattice made 
up of such elements will have a critical 
mass only moderately greater than one 
employing material enriched to 90%. 
The number of plates per element is 
determined by heat-transfer require- 
ments; elements intended for higher- 
power reactors require more plates, 
more Al and consequently a larger 
critical mass. Between }4 and 24 of 
the Alin a plate is found in the cladding 
that protects the Al-U alloy core. The 
following example illustrates the differ- 
ence in critical masses: a light-water- 
reflected-and-moderated active lattice 
made up of 19-plate MTR-type ele- 
ments has a critical mass of about 3.5 
kg, but a lattice of 10-plate elements 
under similar conditions requires only 
2.8 kg to reach critical. Both types of 
elements employed in making these 
measurements contained the same 
amount of fuel, ~160 gm of U**> en- 
riched to 90%. 
The optimum number of plates is a 
compromise between critical mass and 
primary-coolant pumping power. The 
film drop (the temperature differential 
across the thin layer of nearly motion- 
less water next to the surface of the 
plate) must not be so great that local 
boiling takes place. Fewer plates re- 
quire a higher water velocity and more 
pumping power. A typical design for 
a 5-Mw open-pool reactor shows a film 
drop of 125° F at the hottest spot, a 
flow rate of 3,000 gpm and an average 
rise in coolant temperature of approxi- 
mately 12° F. While the optimum 
number of plates is different for each 
power level, it has been suggested that 
in the interest of manufacturing econ- 
omy, industry standardize on a 10- 
plate element for research reactors 
which will be operated -at powers less 
than 5 Mw and on a 19-plate element 
for reactors intended for 5 Mw or more. 
The fuel requirement also is deter- 
mined by the materials used for the 
moderator and reflector. Typical 
values for several reactors employing 
fuel enriched to 20% are shown in 
Table 1. Because fuel plates with 
alloys containing as much as 40 wt % U 
will probably be available for export in 
the visible future, values for the hetero- 
geneous reactors are calculated on that 
basis. Actually there are several thou- 
sand possible permutations and combi- 
