CHAPTER 25
ADDITIONAL LIQUID METAL REACTORS

In this chapter three other types of Liquid Metal Fuel Reactors will be
discussed. The first of these is the Liquid Metal Fuel Gas-Cooled Reactor.
In principle this reactor is similar to the LMFR previously discussed, but
it has many features that are ditferent; for example, it has a noncirculating
fuel, and the heat is removed by cooling with helium under pressure.
Advantages and disadvantages of this design over the circulating fuel
LMFR will be discussed in the following pages.

The second reactor discussed in this chapter is the LAMPRE. Thisis a
molten plutonium fueled reactor which is under development at the Los
Alamos Scientific Laboratory. Although only in its beginning stages of de-
velopment, it is conceived as a high temperature (650°C) fast breeder re-
actor utilizing plutonium as the fuel.

The third type of reactor is based on a liquid metal-TO; slurry fuel.

25-1. Liquip MErar FreL Gas-CooLep REacTor*

25-1.1 Introduction and objectives of concept. The Liquid Metal Ifuel
Gas-Cooled Reactor (LMF-GCR) design 1s unique in that it combines
inert gas cooling with the advantageous liquid fuel approach. The LMF-
GCR concept has a high degree of design flexibility. It is a high-tempera-
ture, high-efficiency system that may be designed as a thermal converter,
uranium thermal breeder, or plutonium fast breeder; that may produce
heat, electric energy, or propulsive power; and that may power either a
steam or a gas turbine.

The fundamental principle of the LMEF-GCR is the utilization of an
internally cooled fixed moderator-heat exchanger element with fluid fuel
center.  The fuel is circulated slowly through the core to assure proper
mixing and to facilitate fuel addition. The core is cooled by gas that is
pumped through it in passages that are separated by a suitable high-
temperature material from the fuel channels. The many well-known
advantages of fluid fuels are thereby gained without the penalties of
circulating great quantities of corrosive, highly radioactive fuel-coolant
solution and of tying up large amounts of expensive fuel outside the core.

*American Nuclear Power Associates: Rayvtheon Manufacturing Co., Waltham,
Mass.; Burns and Roe, Inc., New York City; The Griscom-Russell Co., Massillon,
Ohio; Clark Bros. Co., Olean, New York; Orange and Rockland Utilities, Inec.,
Nyack, New York. Reference design by Raytheon Manufacturing Co. This sec-
tion 18 based largely on contributions from W. A. Robba, Raytheon Manufac-
turing Co.

930
25-1] LIQUID METAL FUEL GAS-COOLED REACTOR 931

25-1.2 Reference design characteristics of an LMF-GCR. Maferials.
Internal gas cooling avoids the corrosion and material problems encoun-
tered in reactor concepts that require the circulation of liquid fuels or
coolants as a heat-transport medium. Helium has been selected as the
gas coolant because it is inert and has better heat-transfer properties than
other inert gases. Graphite has been chosen for the moderator and core
element structural material in a thermal reactor, because of its excellent
moderating and high-temperature properties. Its resistance to corrosion
by bismuth has been fairly well established, and the operating temperature
ix high enough so that energy storage in the graphite should not be a
problem.

Referenee design. A reference design of an LME-GCR nuclear power
station has been produced. A summary of the design parameters is given
in Tuble 25-1. Tt is o graphite-moderated thermal reactor employing highly
enriched uranium-bismuth fuel and helium coolant. The coolant leaves the
core at 1300°17 and is circulated through a superheater and steam generator,
where it produces steam at 850 psig, 900°T".  Since it is inherently self-
regulating, has little excess reactivity, and i cooled by inert helium, it 1s
extremely safe.

In order that the capital cost of the first plant be low, the reference
design ix for a small plant producing approximately 16,000 kw net electrical
sutput. However, it is large enough to demonstrate the practicability of
i LAF-GCR and provide operational experience applicable to com-
mercinl-size plants. By assuming the feasibility of constructing a 13-ft
dizneter pressure vessel for a design pressure of 1000 psi, it appears possible
1o desien 2 gas-cooled reactor plant having an electrical capacity of 240
Mo,

A U235 fyeled thermal reactor was chosen for the design because it will
demonstrate the practicability of the LMEF-GCR concept in a relatively
Smple reactor. A breeder is more complicated because 1t requires two
sitnilor systems for fuel and blanket solutions.

The reactor building and the general arrangement of components as
conceived in the reference desien are shown m g, 25-1. The reactor,
primary coolant system, fuel system, and steam generator are enclosed in
1 gus=tieht steel containment shell.

The reactor core, reflector, internal fuel and gas piping, and pressure
vesscl are shown in ig. 25-2. The core, consisting of an array of graphite
Clenents, has an active length of 56 in. and a eross section approximating
+eirele of 36-in. ditmeter. Fig 25-3 is a picture of a sample section of the
core element. The larger rectangular holes are vertical fuel channels that
would be 56 in. long in the reactor. The small erosswise slots are for helium
coolunt flow.  This graphite element, which separates the two fluids, is
<milar to o heat exchanger that conducts heat from the fuel to the gas
932 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

|

Steam Plant
Building

Containment Sheli

 

 

 

F1c. 25-1. Artist’s concept of LMF-GCR nuclear power station.

channel surface, where it is removed by convection into the coolant stream.
The principal problem associated with the LMF-GCR is the development
of an impervious graphite core material that will prevent significant leakage
of bismuth or fission-product gases into the coolant stream, or of helium
into the fuel.

The machining operations required to produce the core section of the
element have been demonstrated to be feasible. The reflector is made up of
vartous machined graphite shapes. The fuel piping completes the core and
the reactor assembly.

By volume, the core region is approximately 659, graphite, 259, fuel,
and 109 void space for coolant. The fuel solution contains fully enriched
uranium dissolved in bismuth. With these proportions of fuel and moder-
ator, the minimum eritical dimensions as calculated for a cylindrical re-
actor are height and diameter of approximately 42 in. For this application,
a larger core size is required in order to have sufficient heat-transfer area.
Since the graphite core elements are a permanent part of the reactor and
are not changed in routine refueling procedure, it is not required that they
be interchangeable. A considerable amount of design flexibility is thereby
achieved, and variations of the fuel channel, moderator, and gas channel
geometry provide control over the nuclear and heat processes.

For the reactor described above, it is calculated that 900 atomic parts
of U235 per million parts of bismuth are necessary for criticality, if there is
no poisoning of the reactor. However, if the effect of xenon and samarium
25-1] LIQUID METAL FUEL GAS-COOLED REACTOR 933

®QQ

117
d1d

O O 0O

C

    
  

 

 

 

 

04/ D

/,

Section BB

  
   

 

 

 

 

 

 

 
 
 
 
  

View DD

 

Slots For Safety
Controls Fuel Outlet Pipe

 

— (G as

Broken-Out

Section AA Section CC
A

Fig. 25-2. Reactor and pressure vessel assembly.

equilibrium poisoning is included, 1010 ppm of U?3% will be required for
criticality.

The buildup of fission produets and uranium isotopes as a function of
time was calculated to determine the fuel concentration necessary for
criticality after various time periods of operation. Since the solubility of
uranium in bismuth is limited to 6560 ppm at 965°F, the lowest fuel tem-
perature in the LMF-GCR, the reactor fuel must be replaced or processed
after the poisons build up to such a level that this solubility limit is exceeded
by criticality requirements. With the total fuel inventory in the system
equal to 1.2 times the fuel in the core, the fuel lifetime will be 220 megawatt-
vears. This corresponds to an operating period of 4.8 years with a plant
utilization factor of 80%,.

At the end of the fuel lifetime, the fuel solution will contain 3370 ppm
934 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

   
 
 
  
 
 
 
  
     

F1a. 25-3. Model section of nuclear core element for LMF-GCR liquid metal
fuel gas-cooled reactor.

 

Bismuth
Liquid Bismuth Condenser
Charcoal

Trap

 

Fuel Feeder

 

 

 

 

 
     

 

 

 

 

 

 

 

 

 

 

& Sampler
"_"; \ T | 850 psig
AN Q00°F
Bismuth | Bismuth-Uranium Helium 3 4
Charge Tank : Solution
I FiqUid Degasser Super-
| i Nitrogen [ Eleetric Heater Heater
| : Trap
| [
Spent Fuel : |
Shipping Tank | : 3
| | »
! Ditf. Press. 1300°¢
i Controller |
! I Helium
: Cover Gas
Storage
| Reactor Steam
: Generator
|
|
!
I

 

 

 

b -
Spent

Fuel
System

 

 

 

 

R

[m e e —

 

 

 

 

Electromagnetic
Pump

 

 

 

 

 

T
Speni

Storage
Fuel Tank Coolant Pressure Controi

Fia. 25-4. Over-all plant flow diagram.
25-1] LIQUID METAL FUEL GAS-COOLED REACTOR 035

of 17233 1960 ppm of U2 and 1230 ppm of U2 which make up the 6560
ppm of uranium allowed by the solubility limit.

In producing the 220 Mw-yr of heat, 98.7 kg of 17235 will bhe either fis-
stoned or transmuted into U23%, Since 23.8 ke of U?3% remain in the reactor
at the end of fuel lifetime, approximately 8097 of the total amount of U235
added to the reactor during its operation will have been “"burned.”

The systems required in the plant are shown by the flowsheet of I'ig. 25-4.
The heat 1 removed from the reactor by helium at 500 psia, which leaves
the reactor at 1300°F and returns at 900°I. This heat is removed from the
helium in a steam generator that produces superheated steam ut 830 psig,
900°I°. The steam is utilized by a standard turbine generator plant.

A steam-cyele generating plant was incorporated, since it is highly de-
veloped. A closed-cyele gas turbine, the most probable alternative, has
not vet heen developed sufficiently for general utility application, but may
be advantageously combined with the LME-GCR at some later time. In
<uch w0 =ystem, the reactor coolant would serve also as the eycle working
fluid. eliminating the imtermediate heat exchanger.

Although the reference LMEF-GCR 1s envisioned as a high-enrichment
reactor, 1t 15 possible, by changing the parameters, to use fuel of only 209
enrichment.  This low-enrichment reactor would have the advantage of
producing a sizeable fraction of its own fuel by ercating Pu?3* through
neutron absorption in U238,

Purametrie calculations of low-enrichment reactors have been made
u=ing i two-group, two-region spherical geometry computer code developed
for the IBM 650 digital computer. The results show that to have a fuel
litetime long enough (about 1 yr) to be of practical value, the dimensions of
the redctor core should be equivalent to a sphere at least 6 ft in diameter.

25-1.3 Fuel and fuel system. Fuel system. The fuel system is completely
sepurate from the heat-removal system. The main fuel loop flow rate is
approximately 2 to 4 gpm, which is sutticient to provide for uranium makeup
atd Tor gas separation in the degasser.

I'uel flows upward through the reactor core and into the degasser. From
there, the flow gods down into the sump tank and back into the reactor
inlet. The fuel is pumped electromagnetically and flow is measured by an
orltiee or an electromagnetie flow meter.,

The =ump tank acts as a receiver for all the fuel in the loop when the
core 15 to be drained. To keep the sump tank nearly empty during operation,
ihe pressure differential between the helium cover gas in the sump tank
and the degasser must be kept equal to the bismuth static head. The
fuel 1+ antomatieally drained into the sump tank when the pump is de-
crneraized and the two cover gas lines are connected together. Thus there
are no valves i the primary fuel loop which must be operated in order to
drain the reactor.
 

936 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25
TABLE 25-1
SUMMARY OF DESIGN PARAMETERS

Over-all plant performance
Reactor core thermal power 57,000 kw
Helium blower power 5,530 kw
Net electric power generated 16,470 kw
Plant efficiency 28.99%

Thermal data on reactor at full power
Helium pressure 500 psia
Coolant inlet temperature 900°F
Coolant exit temperature 1300°F
Coolant flow rate 389,000 Ib/hr
Coolant velocity in core ~= 560 fps
Number of flow passes 1

Average thermal power density
Peak thermal power density
Peak to average heat flux ratio (average over life)
Design heat output
Maximum graphite temperature
Maximum fuel temperature
AP/P through reactor
Steam plant data
Pressure
Temperature
Flow rate
Number of extractions
Turbine heat rate
Condenser pressure
Turbine speed
Gross turbine output
Pressure vessel
Material
Outside diameter
Thickness
Over-all length
Weight
Type of closure
Insulation

Core

Neutron energy
Fuel, clean

Fuel lifetime

 

 

0.714 Mw/ft3
=(. 922 Mw/{t3
=1.29
1.94 x 108 Btu/hr
1650°F
1755°F

4.3%

850 psig
900°F
188,300 lb/hr
4
9,645 Btu/kwh
1.5in. Hg
3600 rpm
22,000 kw

Stainless steel
94 in,
21n.
123 in.
30,000 1b.
Boited
4 in. of diatoma-
ceous earth

Thermal
900 ppm of U23?
93.59%, en-
riched U in Bi
220 Mw-yr

 

 
25-1] LIQUID METAL FUEL GAS-~COOLED REACTOR 937

TaBLE 25-1 (continued)

 

 

Reprocessing interval (0.8 plant factor) 4. 8 yr
Fuel burnup 809, of U235
Moderator 1.9 g/cc graphite
Bismuth in core 11,400 Ib
Bismuth volume fraction 259,
Graphite volume fraction 659,
Void (helium) fraction 109,
Average core radius 28 in,
Core height 56 in.
Core volume 79.8 ft3
Power 57 Mw
Specific power, average over fuel lifetime ~3700 kw/kg
Power density (based on core volume in liters) 25.2 kw/liter
Average thermal flux (clean) 5.9 x 1014
Average thermal flux (average over life) ~3 X 1014

: Average fast flux (clean) ~6 X 1014

i Average moderator temperature 800°C

‘ Tempeizture coeflicient, average over fuel lifetime ~0.5% 10746k/°C
(ritical mass (clean, enriched U) 5.6 kg

| ('ritical mass (xenon at equilibrium, enriched T) 6.3 kg
Inventory (xenon at equilibrium, enriched U) 7.6 kg
Inventory volume = 1.2 core volume of bismuth 24 ft3
U235 in system at end of fuel lifetime 23 .8 kg
Reflector 1.9 g/cc graphite
Reflector thickness 1.5 ft
Reflector void fraction 59,

 

 

 

Fuel. Uranium makeup is added to the fuel solution on a day-to-day
basix, thus keeping excess reactivity to a minimum. The operating lifetime
of the fuel is nearly 5 yr at full power and 809 plant utilization factor. Fuel
burnup may be as much as 80%, and total U235 inventory varies from
about 7 kg at the beginning of fuel life to about 24 kg at the end of fuel life.

The LMF-GCR tends to be self-regulating. Under the influence of its
negative temperature coeffictent, the reactor will tend to operate at the
same average moderator temperature at all power levels. This temperature
will be maintained by controlling the uranium fuel solution concentration.

Spent fuel. After 4 to 5 yr, nonvolatile fission-product poisons and non-
fissionable 1sotopes of uranium accumulate to such an extent that a new
fuel charge is required. The used fuel is drained into the spent fuel tank and
the reactor fuel loop is then ready to receive a new fuel charge. The spent
fuel ix transferred into a number of small, shielded shipping tanks for ship-
ment to a chemical processing plant.
038 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

25-1.4 Reactor materials. The critical problem associated with the
LMF-GCR is the development of a core element. As a basic core element
material graphite is extremely attractive because it is a very good moder-
ator, possesses excellent high-temperature strength, has unexcelled re-
sistance to thermal shock, 1s not attacked by bismuth, has a low neutron
absorption cross section, possesses a satisfactorily high thermal condue-
tivity, and shows evidence that radiation damage is rapidly annealed at
high temperature. Presently available graphite is not impermeable to bis-
muth or gases, as the core element material of the LMF-GCR must be in
order to separate the fuel and coolant satisfactorily. However, recent de-
velopments indicate a chance for suceess in this area.

The other aspect of core element development is to find a suitable means
for joining the graphite to the upper and lower fuel system headers. The
graphite-to-metal bond must have adequate mechanical strength and be
resistant to corrosion, thermal cyeling, and radiation damage. Bonds of
this type have been prepared by means of high-temperature brazing tech-
niques, and the work has shown that numerous additional bonding agents
are available. Preliminary work is encouraging and indicates that with
improvements in bond design, bond techniques, and test methods, solutions
to the bonding problem may be achieved.

Alternate materials as the basic core element structural material are
under investigation as a backup to the graphite development. These
include KT silicon carbide, molybdenum, molybdenum carbide, niobium,
niobium carbide, zirconium carbide, tantalum, and tantalum carbide, all
of which have properties indicating promise for LMF-GCR application.

25-1.5 Plant operation and maintenance. The LMF-GCR is primarily
self-regulating, having a temperature coefficient of approximately —0.5 X
10~ 4/°C. Large changes in power output are controlled by varying coolant
flow rate while keeping the gas temperatures approximately constant.
Coolant flow rate will be varied by controlling the helium blower speed, and
by changing the coolant gas density (pressure) with the compressor and
accumulator system.

The main plant and reactor control room will be outside the reactor
containment shell in the steam plant geunerator building. A full thickness
of shiclding wall separates the boiler and blower compartments from the
reactor, and operating personnel will be able to conduct maintenance and
inspection of these items while the reactor is in operation. This wall 1s
penetrated by the concentric piping which carries the primary gas into and
out of the reactor. To attenuate radiation streaming through the pipe, a
turn is made within the shield.

The core and pressure vessel assembly have been designed so that the
core, and also the reflector if necessary, may be replaced in the event of a
25-2] MOLTEN PLUTONIUM FUEL REACTOR 939

failure. During operation, the core and reflector are supported at the bot-
tom of the pressure vessel. However, the core assembly is attached to the
pressure vessel head so that the two will be lifted together when the head
1s removed. The reflector is also constructed with a metal support structure
s0 that it can be lifted out of the pressure vessel as a unit.

The fuel loop components and piping are arranged so that maintenance
can be carried out in a safe and reliable manner. Since the parts are rela-
tively inexpensive, it will probably be cheaper to replace than repair them.

25-1.6 Plant capital and power cost. For a 16,000-kw (electrical) LMF-
GCR plant, the cost of power, at an 80% plant utilization factor, is es-
timated at 14.6 mills/kwh, made up of 8.6 mills/kwh for fixed charges,
2.7 mills/kwh for operation and maintenance, and 3.3 mills/kwh for fuel.
The total power cost using a 609 plant utilization factor is 18.4 mills/kwh.
A fixed charge rate of 1569 was used.

The capital cost for a 16,000-kw LMF-GCR nuclear plant has been
estimated at $409/kw of installed capacity. These cost figures are based
on estimates for the important equipment in the plant, and on recent AEC
tuel prices.

25-2. MoLtEN PrLuroNtum FurL Rreacror*

25-2.1 Introduction. The long-range utility of nuclear power based on
uranium fission depends upon the development of a plutonium-fueled
reactor capable of being refueled by an integral, or associated, breeding
evele. If full utilization of the energy content in the world’s supply of
uranium is to be accomplished, the more abundant U238 must be converted
imto the easily fissionable isotopes of plutonium. The need for this full
utilization 1s apparent when it is realized that the economically recoverable
U235 content of uranium ores [1,2] is sufficient to supply projected world
power requirements for only a few decades. Breeding on the plutonium
cyvele extends fission power capabilities by a factor of 140, yielding thou-
s1nds, instead of tens, of years of world energy reserves.

The high values of the capture-to-fission ratio at thermal and epithermal
neutron energies for the plutonium isotopes preclude these types of reactors
from an integral plutonium breeding cycle system. To obtain an appreci-
able breeding gain, a plutonium-fueled reactor must be either a fast or a
fust-intermediate neutron spectrum device where breeding ratios of the
order of 1.7 may be expected from suitably designed systems. One of the
power-producing reactors of the future must logically be a fast plutonium
breeder.

*This section is based largely on material from Los Alamos Scientific Laboratory,
[LA2112, R. M. Kiehn.
940 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

In order to maintain a fast-neutron spectrum, fuel densities in a plu-
tonium breeder will be high, and coolants must be either molten metals or
salts. The latter characteristic will permit large amounts of power to be
extracted from relatively small volumes, thus obtaining & large specific
power. Hydrogenous and organic coolants are eliminated because of their
attendant neutron moderation properties, high vapor pressures at high
temperatures, and relatively poor resistance to radiation damage. For
efficiency reasons the system temperature should be as high as is compatible
with a long operating life. Therefore, to be in step with modern electrical
generation techniques, this would imply coolant outlet temperatures of the
order of 650°C.

25-2.2 Basic components. Before discussing the Los Alamos Molten
Plutonium Reactor (LAMPRE) proposal in detail, the following resume
will treat some of the possibilities for the three basic components of a
power reactor: the fuel, the container, and the coolant.

Molten plutonium fuels. Plutonium metal melts at 640°C, a temperature
that is somewhat high, but not beyond the bounds of utility. Iortunately,
some alloys of plutonium have significantly lower melting temperatures.
Specifically, eutectic alloys of plutonium with iron, nickel, and cobalt
all have melting temperatures in the vicinity of 400 to 450°C. Ternary
and quaternary alloying agents will further lower these melting tempera-
tures by a few percent. One characteristic of these transition metal alloys
is that they do not dilute the fuel volumetrically to a great extent in their
eutectic compositions.

Other alloys of plutonium which are more dilute in fuel and have not
too unreasonable melting temperatures are the magnesium-plutonium and
bismuth-plutonium alloys. The spatial dilution of fuel atoms alleviates the
high power density problem but, unfortunately, these alloys have melting
temperatures significantly higher than the transition metal alloys.

A compilation of the interesting fuel alloys, their melting points, and
eutectic compositions appears in Table 25-2.

Container materials. A material capable of being fabricated into various
shapes and resistant to high-temperature corrosion by the fuel alloy is a
necessity if practical use is to be made of the low melting temperature
plutonium alloys. Since the transition metals readily form low melting
point alloys with plutonium, the normal constructional materials, steels
and nickel alloys, are eliminated.

The next alternatives, the refractory metals, have been used with meas-
urable success to contain the various alloys of plutonium. Tungsten and
tantalum have been somewhat better containers than molybdenum and
niobium and much better than chromium, vanadium, and titanium. The
requirement of fabricability eliminates several of the refractory metals, such
25-2] MOLTEN PLUTONIUM FUEL REACTOR 041

as tungsten and molybdenum, because of the poor state of their peculiar
welding art.

The limitations of metallurgical knowledge at present lead to the con-
clusion that tantalum will be one of the best contuiner materials for these
plutonium alloys. The high-temperature strength properties and the heat-
transfer properties of tantalum are excellent; moreover, it is weldable. The
parasitic capture cross seetion of tantalum would be intolerable in an
epithermal or thermal power breeder reactor and, although relatively
large in a fast spectrum, its effect on neutron economy in a fast reactor can
be made small, if not minor, by careful design.

Dynamic corrosion tests indicate that tantalum’s resistance to corrosion
by molten sodium, a possible coolant, will be adequate. Long-term static
corrosion tests (9000 hr at 650°C) indicate that the fuel is compatible with
tantalum at proposed operating temperatures.

Coolant. The desire to obtain a high power density at high temperatures
and low pressures in a high radiation field dictates the use of molten metal
or =ult coolant. The list of possibilities is topped by sodium and bismuth.
A few words about the properties of these coolants are probably appro-
priate at this point.

Sodium 1s advantageous because of its low melting point, good heat-
trauster properties, low pumping power requirement, and because there
haz been considerable engineering experience with it. Its poor long-term
corrosion properties when in contact with the better container materials
such as tantalum and its explosive burning property when exposed to water
or molst air are distinet disadvantages.

Bizmuth, on the other hand, does not react explosively with water, nor
doex 1t burn in air.  Pumping power requirements some five times larger
than for sodium, its higher melting temperature, and the polonium buildup
problems are disadvantageous factors of a bismuth coolant. However,

TaABLE 25-2

Furr AvLoys

 

 

 

Futectic Melting Approximate

Alloy composition, point, density,

a/o °C g/cc
Pu-Fe 9.5 Fe 410 168
Pu-Co 10 Co 405 16
Pu-XNi 12.5 Ni 465 16
Pu-Mg 85 Mg 552 3.4
Pu-Bi Noneutectic 271-900

 

 

 

 

 

 
942 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

the corrosion resistance of tantalum in dynamie, high-temperature bis-
muth is excellent, according to the Ames experiments [3].

25-2.3 LAMPRE. A first step in solving the plutonium power reactor
problem is to prove the feasibility of operating and maintaining a molten
plutonium power reactor core. To this end, the reactor assembly known as
LAMPRE T has been devised. The LAMPRE system has the following
essential features:

Fuel alloy: Molten plutonium-iron
(eutectic composition, 9.5 a,/0 IFe)

Container: Tantalum

Reflector: Steel

Shield: Graphite, iron, concrete
Coolant: Sodium

Power: 1 Mw heat

Heat transfer: Internally cooled core

Tube-shell
Heat exchanger
Heat rejected to air

Breeding No breeding blanket

Core. The LAMPRE core consists of three parts: fuel alloy, container,
and coolant. A proposed design, deseribed in detail below, yields a strue-
ture which 1s approximately 5097 by volume fuel alloy, 1577 strueture, and
30Y% coolant. The minimum tube separation 1s shightly under 1/16 .
At reasonable heat-transfer rates, this configuration is capable of develop-
ing a specific power of better than 250 watts/g. More efficient systems can
utilize a similar structure but must dilute the fuel volumetrically to ob-
tain a larger heat-transter surface per unit of contained fuel. The larger
area-to-volume ratio can be obtained by going to smaller diameter tubes
and/or closer gpacing of the tube array. In the tube-shell arrangement,
the fuel is located on the outside of the tubes and the coolant flows through
the tubes. Such a scheme preserves the volumetrie integrity of the fuel.
Other radiator-type schemes, which also preserve fuel integrity, are
concelvable.

The over-all assembly will be designed so that the core will be completely
filled during operating conditions. The estimated core height is 6.5 in.
25-2] MOLTEN PLUTONIUM FUEL REACTOR 943

Tantalum expansion, filling, and draining tubes will be attached to the
core structure. A reference core assembly would be:

Container Tantalum
Tubes (547) 3/16-in. OD, 0.015-in. wall,
hexagonal array
Cage shape Right eylinder,
6.25-in. OD, 6.5-in. height
Headers and shell 0.040 to 0.080 in.
Critical mass 26 kg plutonium alloy

R flector. No attempt to breed will be carried out in the first LAMPRE
concept.  Although the over-all coolant container will be made of stainless
<teel, the fast-neutron reflector will be made of steel and will be cooled
by the main sodium stream. The thickness of the radial steel reflector will
be adjusted to be thin enough, neutronwise, to obtain adequate external
reflector control, but will be too thick to allow the thermalized neutrons
returning from the graphite shield to build up a power spike at the core
-urface. The core, although slightly coupled to the reflector and shield, will
have a mean fission energy greater than 500 kev, ensuring a high possible
breeding gain. The top and bottom stainless-steel reflector slugs will also
be <odium-cooled and will be essentially “mnfinitely” thick to fast neutrons.
The coolant channels will be drilled or machined into solid slug or disk
castings,

Control.  The control of LAMPRE will be effected by reflector-type
mechanisms.  An annular shim control digplacing the innermost 4 in. of
<hicld with aluminum will be used as a coarse criticality adjustment
mierhanism,  Several replacement eylinders, replacing the inner portions
of aluminum with void, will be used ax fine controls. A rotating control
cvhmder will be built into the system in anticipation of safety and neutron
Kineties experiments,

The radial thickness of the steel fast-neutron reflector 1s adjusted so that
the 1=t and intermediate neutrons returning to the core from the alummum
reflector and graphite are worth approximately 109, to the core critical
i<, Displacement of the aluminum reflector effectively reduces the
newutron refleetion back to the core, yielding an external, large-effect control
mechani=m adequately cooled by aluminum conduction and air convention.

The LAMPRE critical experiments have proved that aluminum-void
replicement mechauisms are effective and operable.  The annular shim
hias been shown to be almost ineffective at distances greater than 2 in.
944 ADDITIONAL LIQUID METAL REACTORS [cHAP. 25

  
   
  
    
  
  

SODIUM IN—

=S OR OUT

TOF REFLECTOR
POONTAINMENT

GAS VENT

IHIM _
‘ e ——HEAT EXCHANGER
FUEL CONTAINER H ; BAYONET °

BOTTOM REFLECTOR

 

 

 

CONTROL RBOD THIMBLE

 

FUEL ADDITION LINE

RADIAL REFLECTOR—r———gr |

 

 

 

FUEL RESERVOIR
CATCH PAN

 

LOCATING PIN

Fia. 25-5. The Los Alamos Molten Plutonium Reactor Experiment.

above or below the core height for the geometry. These results have been
incorporated into the LAMPRE design as presented in Fig. 25-5.

25-3. Liquip MeTAL-UranNtuM OXIDE SLURRY REACTORS

There has been some work done at other locations on uranium oxide
slurry reactors. At Knolls Atomic Power Laboratory, a uranium oxide-
bismuth slurry reactor has been explored [4]. In this reactor, the fuel,
consisting of uranium oxide suspension and liquid bismuth, is pumped
through a moderator matrix and then through an external heat exchanger.
The reader will recognize that this is the same as the single-region LMFR
described in the preceding chapter.
REFERENCES 945

The studies at KAPL were encouraging. A small amount of experimental
work indicated that dispersions of uranium oxide and bismuth can be made.
These workers found that at 500 to 600°C titanium is the best additive for
promoting the wetting of UOs by bismuth. An 8 w/0 UQOg-bismuth slurry
was actually pumped with an electromagnetic pump at 450°C.

At Argonne National Laboratory, uranium oxide-NaK slurries have been
studied as possible reactor fuels [5]. This fuel would be suitable for a
fast-breeder reactor. Investigations have been carried out at a maximum
concentration of 4.3 vol. 9% UO; in eutectic NaK. Two loops have been
operated at temperatures ranging from 450 to 600°C. A slurry with 4.3
vol. % actually has a very high weight percent, 36.0 w/o.

The tests in the two loops indicated uniform suspension at flow rates of
2 fps.

The U032 dropped out of suspension at temperatures above 500°C but
would resuspend at lower temperatures. When a very small amount of
uranium metal was added to the slurry, better wetting of the particles
was obtained and no further settling above 500°C was observed.

Work on the uranium oxide slurries is continuing, and the incorporation
of these results into liquid metal fuel reactors can be expected.

REFERENCES

1. 2. GLassTONE, Principles of Nuclear Reactor Engineering. Princeton, N. J.:
D. Van Nostrand Co., Inc., 1955. (pp. 1-2)

2. P. C. Purnawm, Energy in the Future. Princeton, N. J.: D. Van Nostrand
Co.. Ine., 1953, (p. 214)

5. R, W. Frsuer and G. R. Winpers, High Temperature Loop for Circulating
Liquid Metals, in Chemical Engineering Progress Symposium Sertes, Vol. 53,
Nao. 200 New York: American Institute of Chemical Engineers, 1957. (pp. 1-6)

4. D. H. Anmany et al., A UOo-Bismuth System s a Reactor Fuel, USAEC
Report KAPL-1877, Knolls Atomic Power Laboratory, July 1, 1957.

5. B. M. ABranawm et al,, UO2-NaK Slurry Studies in Loops to 600°C, Nuclear
Net. and Eng. 2, 501-512 (1957).