URI GAT and J. R., ENGEL Oak Ridge National Laboratory
P O, Box 2009, Building 9108. Oak Ridge, Tennessee 37831-8088
H. L. DODDS University of Tennessee
Nuclear Engineering Department, Knoxville, Tennessee 37996
Received December 30, 1991
Accepted for Publication May 28, 1992
KEYWORDS: molten salt reactor,
dismantled weapons fuels,
The molten salt reactor (MSR) option for burning fissile fuel from dismantled weapons is examined. It is concluded that MSRS are potentially suitable for beneficial utilization of the dismantled fuel. The MSRs have the flexibilify to utilize any fissile fuel in continuous operation with no special modifications, as demonstrated in the Molten Salt Reactor Experiment, while maintaining their economy, The MSRS further require a minimum of special fuel preparation and can tolerate denaturing and dilution of the fuel. Fuel Shipments can be arbitrarily small, which may reduce the risk of diversion. The MSRS have inherent safety features that make them acceptable and attractive. They can burn a fuel type completely and convert it to other fuels. The MSRs also have the potential for burning the actinides and delivering the waste in an optimal form, thus contributing to the solution of one of the major remaining problems for deployment of nuclear power.
There are expectations that fissile material from nuclear arms
reduction will become available and require
disposition.1 Proposals for this disposition vary
widely from disposal as waste, with all the associated issues of
monitoring, safeguard, verification, and long time periods of needed
control, to utilization as a nuclear fuel for beneficial energy
production. The concerns are to retain control of the fissile
material over its lifetime, to avoid any recycle into weapons, and to
maximize the economic benefits and minimize any risks. The use of the
dismantled fissile material in light water reactors (LWRs) is
The emphasis here is on the use of the fissile material as a fuel
for the beneficial generation of power in molten salt nuclear
reactors (MSRs). Specifically, the potential advantages of MSRs as
versatile, flexible fuel utilizers are discussed. Some safety
features of MSRs proliferation-inhibiting properties and
possibilities for advantageous handling of waste are pointed out. The
MSRs utilize the fuel in the form of fluid fissile material. Fluid
fuel reactors have some unique possibilities associated with the
ability to circulate the fuel.3
The molten salt programs of the U.S. Department of Energy (DOE)
and its predecessor agencies had their manifestation in two very
successful reactor experiments: the Aircraft Reactor
Experiment4 (ARE) and the Molten Salt Reactor
Experiments (MSRE). Both of these reactors were designed, built, and
operated by Oak Ridge National Laboratory (ORNL). Researchers at ORNL
also conducted many extensive studies of various possible MSR
concepts The ARE was a product of the Aircraft Nuclear Propulsion
Program and operated successfully in 1954 (Ref. 6).
That program was subsequently discontinued, but a civilian-oriented
Molten Salt Reactor Program5 (MSRP) that began in
1956 continued development of this general technology. The primary
goal of the early MSRP and for most of this program was the
development of molten salt breeder reactors' (MSBRs) using the
Th-233U fuel cycle, which could compete with other concepts using the
238U-Pu fuel cycle. Consequently the effort was focused on a system
with integral, on-line chemical processing. The MSBR effort was
discontinued in 1972, resumed as a technology development program in
1974, and finally closed out in 1976, A small design study was
undertaken in 1978 as part of DOE's Nonproliferating Alternative
Systems Assessment Program.8 This study examined
additional MSR concepts that might offer greater resistance to
nuclear proliferation than the LWRs operating on a once-through fuel
cycle. The study led, ultimately, to two similar conceptual MSRs a
breakeven breeder 8,9 using a
complex, on-line fuel processing plant and a simplified
converter10 with a oncethrough 30-yr fuel
Molten salt reactor studies have been undertaken in many places.
One of the larger programs was conducted in the Federal Republic of
Germany with the Molten Salt Epithermal (MOSEL)
Reactor.11,12 The MOSEL reactor
forgoes the graphite in the core, which is used as a moderator in
other MSR concepts, to achieve an epithermal spectrum for enhanced
breeding in the thorium cycle, More recently, some concepts in
Japan13 and at ORNL (Ref. 14)
addressed simplicity of design and enhanced safety as the primary
This technical note is based on the earlier studies and previous work. No specific calculations have been performed to confirm the potential capabilities of the MSRs suggested here, Many of the ideas proposed are conceptual. Several of the past concepts have been combined into new concepts. Not all of the possible rcsulting interactions have been explored. Thus, further studies are necessary to fully understand all the implications of the ideas suggested here.
FUELS FOR MSRs
Molten salt reactors are usually geared toward the Th - 233U fuel
cycle. They were developed initially when there was high emphasis on
breeding. The MSRs were conceived as near thermal reactors with a
graphite moderator. The preferred salts are fluorides, including
beryllium and lithium fluorides, because of their desired nuclear and
thermodynamic properties. Both the beryllium and the fluorine cause
significant neutron moderation. To achieve breeding with the soft
neutron spectrum, it is necessary to select the thorium
cycle.15 To enhance breeding, the MOSEL concept
removed the graphite moderator of the thermal concept to harden the
spectrum and reach into the peak region of the 233U neutron yield in
the epithermal spectrum. 11
The MSRE was operated initially with 235U as the fissile fuel at
35% enrichment. That operation spanned 34 months beginning in 1965
and included a sustained run of 188 days (partly at low power to
accommodate the experimental program). All aspects of operation,
including the addition of fissile fuel with the reactor operating at
power, were demonstrated. Subsequently, the mixture of 235U and 238U
was removed from the salts by on-site fluorination, and 233U was
added to the fuel salt for the next phase of the operation. Plutonium
produced during the 235U-238U operation remained in the salt during
the 233U operation. Several fissile additions consisting of PuF3 were
made15 for fuel makeup to demonstrate that
capability. The plutonium additions were made by adding capsules of
PuF3 in the solid form to the reactor salt and allowing the plutonium
salt to dissolve. Thus, plutonium from two sources was burned in the
MSRE: the added plutonium and the plutonium that was bred from the
238U in the initial operations.
The MSRE, without changes in design, operated successfully on all
of the major fissile fuels: 235U and 233U, and plutonium mixed with
uranium. This property provides the ultimate flexibility in the
utilization of fissile fuel.
By mixing the fuel with adequate proportions of fertile material,
conversion to either plutonium or 233U is possible. Calculations have
indicated promising conversion ratios (near 0.9) for a variety of
conditions, and values above 1.0 may be achievable under carefully
controlled conditions with on-line processing to remove fission
product poisons. With an appropriate fuel cycle, one fissile material
can be burned off almost completely or burned and "converted" into
another. As an example, one could burn plutonium and produce 233U.
Such a conversion will transform a fuel, plutonium, particularly
suitable for weapons, into a fuel, 233U, that may be less suitable
for weapons but more neutron productive in nonfast spectra.
Furthermore, while plutonium could be separated from the salt (or
other additives) by chemical means, uranium will contain substantial
amounts of 232U, which is considered a strong deterrent to
proliferation. The very strong radioactivity emanating from the 232U
decay products makes any direct handling prohibitive only a short
time after chemical purification . The choice of fissile material in
MSR fuel salt does not seriously affect the salt properties. Hence, a
given reactor plant would be capable of using fissile materials in
arbitrary combinations for high-temperature, high-efficiency power
The fuel supply from the dismantled nuclear devices could be augmented at any time or totally displaced by fuel from other sources. By adjusting other components of the fuel, the conversion ratio can be controlled within rather wide limits. This further ensures uninterrupted continued operation of MSRs for support of the overall energy economy. The fact that no substantial design changes are required to accommodate fissile supply changes acts as a damper on the propagation of interruptions, changes in schedule, or plans. This flexibility also moderates any costs that might result from changes and interruptions.
DISMANTLED WEAPONS FUEL
Fissile fuel from dismantled weapons is either highly enriched
uranium or plutonium. While there are only rough estimates available,
it is assumed that quantities that would become available in the
foreseeable future are sufficient to fuel one to a few reactor
lifetimes. ' It is further reasonable to assume that the fuel will
become available on a continuous, rather than batch, basis. It is
desirable to degrade the fuel to nonweapons grade immediately by such
means as denaturing, diluting, or spiking. This will reduce the
concern of diversion, the need for control and accounting, and the
extent of security provisions. To reduce cost, it is required to
degrade the fuel one time, preferably at the time and location of
dismantling. There should be no need to reverse any of these steps
later, for example, for the manufacture of fuel elements. As
discussed earlier, the MSRs are particularly well suited to
accommodate these needs.
The quantity and supply rate of dismantled weapons fuel poses a
dilemma. If a minimum number of reactors is dedicated to using this
fuel, then the fuel must be accumulated, protected, stored, and
monitored for long periods of time. If a large number of reactors is
utilized, then the probability of operations disruption becomes very
high. Furthermore, relatively large facilities and a large number of
reactors need to be modified to accommodate a short spurt of fuel
supply. Such an effort can be expensive and would require much
detailed advanced planning and an intense commitment to a detailed
schedule. The MSRs, as discussed earlier, require no design changes
and can readily switch between fuels on an ad hoc basis.
Solid fuel reactors with no reprocessing and fuel recycling leave a large percentage of the original fuel in the spent fuel. This constitutes an indefinite commitment for guarding and storing the spent fuel. Eventually, it adds a burden on the solution for the disposal of the waste.
FLUID FUEL REACTORS
The MSRs are fluid fuel reactors, and as such, they differ from
all the current, common, solid fuel reactors. Fluid fuel can be
transferred remotely by pumping through pipes connecting storage or
reaction vessels (e.g., a reactor core). The relatively simple remote
handling allows even the fresh fuel to be highly radioactive, which
provides a strong diversion inhibitor. Also, highly radioactive fuel
can be detected easily. If the temperature of the fuel is allowed to
drop, the fuel solidifies and again is difficult to manipulate,
providing additional diversion protection.
The fluid fuel at operating reactor fissile concentrations
provides inherent protection against criticality accidents during
handling. In thermal designs, the graphite moderator is required for
criticality so that criticality can occur only in the core. For other
concepts, the design would have to exclude vessels that are not
criticality safe for credible fuel mixtures.
Fuel prepared for an MSR can be conveniently shipped as a cold
solid and remelted just before it is added to the reactor system. For
small additions, the reactor can be designed to accept the fuel in
the frozen state, as in the MSRE. With a fluid fuel, the entire fuel
element fabrication process is avoided. This saves a significant part
of the head-end effort and cost. The absence of a solid fuel
manufacturing phase provides for enormous flexibility. The fuel can
be blended into the reactor exactly as needed at any time. The amount
of fuel added will depend on the type of fuel, its isotopic makeup,
and concentration. There is no need for exact longrange planning that
may be upset by variations on either the supply or the demand side.
There is no need for long lead times and interim storage. These
advantages are particularly important for fuel derived from weapons.
The rate and exact kind of fuel that becomes available can be
accommodated by the reactor. The fine-tuning of the composition can
be done on an ad hoc basis at the site.
One possibility for the process of converting weapons fissile material to fluid fuel for a reactor is to do it at a dismantling facility. At that facility, the fissile material could be converted into a salt, denatured, spiked, diluted, or whatever else may be deemed desirable for safety, security, economy, or practicality. The denaturing and spiking can render the fuel unattractive for proliferation or diversion. Being designated for MSRs allows shipment in quantities and form as optimized for safety and security, again inhibiting diversion and also possibly reducing potential public objection.
MOLTEN SALT REACTORS
Molten salt reactors are unique in many ways. One of the major
advantages of the fluoride-based MSRs is the potential for an
integrated fuel recovery capability. The processing is based on the
high volatility of UF6. By sparging the salt with fluorine, uranium
can be removed essentially quantitatively as UF6, which can then be
converted back to UF4 and recycled into a fresh batch of fuel salt.
The residual salt, now free of uranium, could be subjected to any of
a number of processes to remove fission products and concentrate
them. The carrier salt components (lithium, beryllium, fluorine)
could also be isolated and recycled if that were economically
desirable. All of these steps could be made independent of the
The feasibility of the various steps for on-line processing has
been calculated and individually demonstrated at ORNL (Refs.
16 and 17). In addition, the
uranium recovery step was demonstrated in the MSRE when the fissile
material was changed from 235U to 233U The process involved 47 h of
fluorine sparging over a 6-day periods to produce a uranium product
pure enough for cascade reenrichment.
Molten salts can operate at high temperatures and low pressures, and they possess favorable heat transfer properties. These properties result in high thermal efficiencies for the reactor and absence of safety hazards associated with high pressures, such as ruptures or depressurizations. The salts are chemically stable and nonflammable, averting fire hazards, and there are no energetic chemical interactions between the salts and water.
Safety of MSRs
The MSRs can potentially achieve almost any degree of safety
desirable at a cost. Some extreme degrees of safety were summarized
in the proposal for the Ultimate Safe Reactor (U.S.R) (Ref.
18). The MSRs possess many inherent safety
properties. As an MSR uses a molten fuel, a "meltdown" is of no
particular consequence. The fuel is critical in the molten state in
some optimal configuration. If the fuel escapes this environment or
configuration because of relocation, it will become subcritical thus,
recriticality in any reasonable design cannot occur.
Fluid fuel has inherently a strong negative temperature
coefficient of reactivity because of expansion of the fluid that
results in removal of fuel from the core. This property is in
addition to other spectral contributions to the negative reactivity
coefficient. At the very extreme, the fuel would cause failure of the
primary coolant boundary (without a serious pressure rise) in which
case the fuel would be returned to a critically safe configuration.
Further, the ability to add fuel with the reactor on-line strongly
limits the amount of excess nuclear reactivity that must be available
in the system.
On-line processing reduces the amount of fission products retained
in the system. This reduces both the risk of dispersal of
radioactivity and the amount of decay heat that must be contended
with during an accident. The fission product inventory, in an earlier
concept of the MSBR, was planned to be a l0-day
accumulation.7 A more recent proposal, the U.S.R
(Ref. l8), suggests reducing the fission products
to a level where the entire afterheat can be contained in the salt
without reaching boiling. There is a limit to the reduction of
fission product inventory in the reactor. The limit is determined by
several factors two of which are economics and concentration of the
fission products. The practical limit has not yet been determined and
is not known. In practically all MSR concepts, the fission gases and
volatiles are removed continuously, reducing significantly the
potential radioactive source term.
Fluid fuel also allows shutdown of the reactor by draining the
core into subcritical containers from which any decay heat can be
readily removed by conduction and natural convection.
Proliferation resistance and other safety attributes are described elsewhere in this technical note. The MSRs can be designed in an extremely safe manner with inherently safe properties that cannot be altered or tampered with. These safety attributes make the MSRs very attractive and may contribute to their economy by reducing the need for elaborate safety measures.
Nuclear waste is an important issue affecting the acceptability of
any nuclear-related system and reactors in particular. There is no
way that a reactor that utilizes the fission process can eliminate
the fission products. The MSRs, with their continuous processing and
the immediate separation of the residual fuel from the waste,
simplify the handling of the waste and contribute to the solution and
acceptability of the waste issue.
The on-line processing can significantly reduce the transportation
of radioactive shipments. There is no shipping between the reactor
and the processing facility. Storage requirements are also reduced as
there is no interim storage for either cooldown or preparation for
shipment. The waste, having been separated from the fuel, requires no
compromise to accommodate the fuel for either criticality or
diversion concerns. The waste shipments can be optimized for waste
concerns alone. The actinides can be recycled into the fuel for
burning and thus eliminated from the waste. While further work is
required to fully analyze this possibility, several proposals to burn
actinides have been made. The MSRs with on-line processing lend
themselves readily to recycling the actinides into the fuel.
Eliminating the actinides from shipments and from the waste reduces
the very long controlled storage time of the waste to more acceptable
and reasonable periods of time.19 The on-site
on-line processing allows for inclusion of some selected fission
products along with the recycled actinides for transmutation in the
reactor. For example, the long-lived iodine could be removed from the
waste and retained in the core .
The fission products, already being in a processing facility and in a fluid matrix, can be processed to the optimal form desired. That is, they can be reduced in volume by concentration or diluted to the most desirable constitution. They can be further transformed into the most desirable chemical state, shape, size, or configuration to meet shipping and/or storage requirements. The continuous processing also allows making the shipments to the final disposal site as large or small as desired. This can reduce the risk associated with each individual shipment to an acceptable level.
The MSRs are suitable for the beneficial utilization of fissile
material from dismantled weapons for efficient and economical energy
production. The MSRs can utilize all three major fissile fuels: 233U,
235U, and plutonium, as demonstrated in the MSRE. This flexibility is
achieved without reactor-core design modifications. Additionally, MSR
fuels can be fed continuously on-line and can come in a variety of
combinations. The fuels can be made proliferation and diversion
resistant during preparation at the head end, by dilution,
denaturing, spiking, and/or controlling the size of shipments.
The MSRs are expected to be generally attractive because they have inherent safety attributes that reduce the risks to low levels. These safety attributes include reduced probability for an accidental criticality or for recriticality and freedom from core meltdowns. The on-line processing potential can reduce the fission product inventory and, with it, any risks of radioactive dispersal and the risks associated with the inability to remove the afterheat. Reluctance to consider on-site processing is usually associated with solid fuel processing that requires head-end and tail-end remanufacturing and is therefore much more elaborate. On-site on-line processing is often opposed because of the introduction of chemical processing to the reactor site. However, for MSRs this aspect is more than compensated for by the simplified reactor design and reduced safety design requirements. On-line processing may also enable treatment of the waste by recycling and burning the actinides so that long-time controlled storage is not required. The bulk of the waste can be reduced in volume and brought into shape, size, form, chemical composition, and shipment and disposal size that are the most acceptable. Power production need not be interrupted by fissile supply fluctuations from the dismantled weapons. A particular fissile type can be burned completely and, if desired, converted into another fissile isotope. Fuel recycling and fabrication are not necessary. Fissiles can be treated completely at the head-end dismantling facility. Fuel shipment sizes may be controlled to minimize risk during fuel transportation.
This technical note was authored by a contractor of the U.S. government under contract DE-ACO5-840R21400. Oak Ridge National I laboratory is managed by Martin Marietta Energy Systems, Inc., under contract with DOE.
1. L. C. HEBEL, "Limiting and Reducing Inventories of Fissionable Weapon Materials"' presented at American Association for the Advancement of Science Mtg., February 18, 1991; published in Fissile Material from Nuclear Arms Reduction: A Question of Disposition, W. G. SUTCLIFFE, Ed.., CTS-3 1 92, Lawrence Livermore National Laboratory.
2. J. J. TAYLOR, "Disposal of Fissionable Material from Dismantled Nuclear Weapons, presented at American Association for the Advancement of Science Mtg., February 18, 1991; published in Fissile Material from Nuclear Arms Reduction: A Question of Disposition, W. G. SUTCLIFFE, Ed., CTS-31-92, Lawrence Livermore National Laboratory.
3. J. A. LANE, H. G. MacPHERSON, and F. MASLAN, Fluid Fuel Reactors, Addison-Wesley, Reading, Massachusetts (1958).
4. M. W. ROSENTHAL et al., "Molten Salt Reactors," Proc. Int. Conf. Constructive Uses of Atomic Energy, Washington, D.C., November 1968, American Nuclear Society (Mar. 1969).
5. P. N. HAUBENREICH and J. R. ENGEL, "Experience with the Molten Salt Reactor Experiment"' Nucl. Appl. Technol., 8, 118 (1970).
6. R. C. BRIANT et al., "The Aircraft Reactor Experiment"' Nucl. Sci. Eng., 2, 797 (1957).
7. R. C. ROBERTSON, Ed., "Conceptual Design Study of a Single-Fluid Molten Salt Breeder Reactor," ORNL-4541, Oak Ridge National Laboratory (June 1971).
8. "Nonproliferating Alternative Systems Assessment Program Plan," U.S. Department of Energy, Assistant Secretary for Energy Technology, Nuclear Energy Programs, Office of Fuel Cycle Evaluation (Apr . 1978) .
9. J. R. ENGEL et al., "Molten Salt Reactors for Efficient Nuclear Fuel Utilization Without Plutonium Separation", Nucl. Technol., 46, 30 (1979).
10. J. R. ENGEL et al., "Conceptual Design Characteristics of a Denatured Molten Salt Reactor with Once-Through Fueling", ORNL/TM-7207, Oak Ridge National Laboratory (July 1980).
11 . P. R. KASTEN, "The MOSEL-Reactor-Concept", Proc. 3rd Int. Conf. Peaceful Uses of Atomic Energy, Geneva, Switzerland September 1964.
12. P. R. KASTEN, U. GAT, S. SCHULZE HORN, and H. W. VORNHUSEN, "Design Concepts for the Core Structure of a MOSEL (Molten Salt Experimental) Reactor," Nucl. Struct. Eng., 2, 224 (1965).
13. K. FURUKAWA et al., "Simplified Safe Small
Molten Salt Reactor: 'FUJI, for a Global Measure of Greenhouse
Effect", presented at 9th Int. Conf. Energy and Environment, Miami
Beach, Florida, December 11-13, 1989; also J. Nucl. Sci. Technol.,
27, 1157 (1990)
14. U. GAT and S. R. DAUGHERTY, "The Ultimate Safe (U.S.) Reactor", presented at 7th Int. Conf. Alternative Energy Sources, Miami Beach, Florida, December 9-11, 1985.
15. M. W. ROSENTHAL, P. N. HAUBENREICH, H. E. McCOY, and L. E. McNEESE, "Recent Progress in Molten Salt Reactor Development", Atomic Energy Review, IX, 3, 601 (1971).
16. R. B. LINDAUER, "Processing of the MSRE Flush and Fuel Salts", ORNL/TM-2578, Oak Ridge National Laboratory (Aug. I 969).
17. W. L. COSTER and E. L. NICHOLSON, "Design Cost Study of a Fluorination-Reductive Extraction-Metal Transfer Processing Plant for the MSBR", ORNL/TM-3579, Oak Ridge National Laboratory (May 1972).
18. U. GAT, "The Ultimate Safe (U.S.) Reactor: A Concept for the Third Millennium", presented at 4th Int. Conf. Emerging Nuclear Energy Systems, Madrid, Spain, June 30-July 4, 1986.
19. A. G. CROFF, C. W. FORSBERG, and S. B. LUDWIG, "A Re-Examination of the Incentives for Actinide Burning", Trans. Am. Nucl. Soc., 62, 76 (1990).
Uri Gat (BSc, mechanical and nuclear engineering, Israel Institute
of Technology, Haifa, Israel; Dr. lng., ceramic and nuclear
engineering, Rheinisch-Westfalische Technische Hochschule, Federal
Republic of Germany) is a researcher at Oak Ridge National Laboratory
(ORNL). He was responsible for the engineering of the Molten Salt
Epithermal Reactor at KFA Jülich. He developed the concept of
the Ultimate Safe (Molten Salt) Reactor. Also, he was the ORNL
manager for the gas cooled fast reactor and the liquid metal fast
breeder reactor. He is currently working on safety for the High-Flux
Isotope Reactor and Molten Salt Oxidizer for waste treatment.
J. R. Engel (BS, University of Toledo, l953; Oak Ridge School of
Reactor Technology, l954) was involved in the development,
construction, operation, and evaluation of fluid fuel reactors at
ORNL, including the Homogeneous Reactor Experiment II and the Molten
Salt Reactor Experiment. He also participated in the study of molten
salt breeders and the conceptualization and evaluation of other
advanced molten salt reactor concepts, particularly for
proliferation- and diversion-resistant applications. He is currently
H. L. Dodds (BS, l966; MS, l969; and PhD, l970, nuclear engineering, University of Tennessee-Knoxville) is IBM Professor of Nuclear Engineering at the University of Tennessee-Knoxville and consultant to ORNL. His current research interests include methods development and applications in reactor physics, reactor safety, and nuclear criticality safety.