CATALOG OF NUCLEAR REACTOR CONCEPTS
Part I. Homogeneous and Quasi-homogeneous Reactors Section 111. Reactors Fueled with Molten-salt Solutions
Charles E. Teeter, James A. Lecky, and John H.Martens
Technical Publications Department
The reactor concepts described in this section utilize a fluid fuel consisting of a fissionable material dissolved in a carrier of molten salt. Some, such concepts also call for a molten salt as a primary coolant.
H. G. Maceherson has reviewed the technology of molten-salt reactors.
The fissionable compound usually chosen is uranium tetrafluoride.2 Uranium fluorides other than the tetrafluoride have the disadvantages of higher volatility, instability, or corrosivity. Chlorides or fluorides as solvents have been given the most consideration because of the need for radiation stability and high solubility. The chlorides are used for fast reactors, and the fluorides, because of their low cross sections for thermal neutrons, are best for thermal and epithermal reactors.
Compounds other than fluorides and chlorides have been suggested, including phosphates, sulfates, sulfide, hydrosulfides ,and hydroxides, Molten fluoride mixtures, e.g. LiF-NaF, have the most desirable properties: they dissolve adequate amounts of fuel; they have satisfactory heat transfer properties; they resist radiation; they can tolerate an accumulation of fission products; the melting points are low enough that corrosion problems of excessively high temperatures are avoided; and the vapor pressures are low enough to permit low-pressure operation. Although, fluorine itself has some moderating properties, a better moderator must be present, either in the molten-salt mixture or as a separate structure, to obtain a thermal reactor of reasonable size. The alkali metal fluorides have been especially considered as solvents because they have low melting points. Beryllium fluoride may be added to the mixture as a moderator, and thorium tetrafluoride can be added for conversion.
Graphite and beryllium are commonly used as structured moderators. Discussion of structured-moderated reactors in this section may appear anomalous, in that most reactors in this first part of the catalog have completely homogeneous cores. The fuel itself, however, is a homogeneous solution, and such structure-moderated reactors are otherwise closely related to the more homogeneous ones.
Molten-salt reactors can be classified in several ways–by the purpose of the reactor, by the use or nonuse of a separate moderator, by the cooling method (internal or external), or by the core arrangement (one- or two-region). For this catalog, the classification will be by one- and two-region reactors. Chapter 2 will cover the first and Chapter 3
the second. According to MacPhers~nt,h~e most attractive types are the one-region, graphite-moderated reactor: and the two-region reactor. The one-region reactor is simpler, and it is cheaper to construct and operate for small power stations. Most of the one-region reactors discussed in Chapter 2 are burners. The two-region reactor has a better neutron economy, is best for breeding, and, in larger installations, gives higher con- version ratio and lower fuel-cycle costs.
The origin of the molten-salt concept is not clear. Early in the 1950’s, however, molten salts were considered as reactor fuels to satisfy the need for high temperature and extremely high power densities needed for reactors intended for nuclear aircraft propulsion. Development work, particularly at Oak Ridge National Laboratory, resulted in several con- cepts, and in 1954 the Aircraft Reactor Experiment (ARE), part of the Aircraft Nuclear Propulsion Project (ANP), was operated. Since the cancellation, as of October 1957, of work on circulating-fuel reactors for aircraft work has continued aimed at developing power and breeder reactors that utilize molten salts.
Molten-salt reactors are attractive concepts for several reasons. They provide high temperatures in a low-pressure system to produce steam at temperatures high enough to give high thermal-cycle efficiencies. They are versatile because of the range of solubilities of different compounds of fissionable elements in salts. The simple ionic salts are stable under irradiation. Such reactors also have the advantages of other fluid-fueled reactors; for example, they have high nega- tive temperature coefficients of reactivity, fission products can be re- moved continuously, fuel elements need not be fabricated, and make-up fuel may be added as needed. According to Weinberg,” a great advantage is that the fissionable materials can be consumed at very high thermal efficiencies and with extremely high burnup. The major problem with a salt-fueled reactor is that all the salt in the system must be kept molten at all times.
In 1959, the Fluid Fuel Reactor Task Force compared the aqueous homogeneous, molten- salt, and liquid-metal fueled reactors The prin- cipal conclusions were: the molten-salt reactor had the best chance of achieving technical feasibility; only with the homogeneous aqueous reactor was there a possibility of achieving a reasonably short doubling time;
and the total power costs for the MSR were between those for the aqueous reactor and the liquid-metal-fueled reactor.
Salt-fueled reactors are advanced concepts, and there is as yet no adequate experience for building large- scale power plants, although many concepts have been developed for such plantse7 Current development is represented by the Molten Salt Reactor Experiment, which achieved criticality at ORNL in mid 1965.