Safety Challenges of Molten Salt Reactors

Garrett LeCroy
February 18, 2021

Submitted as coursework for PH241, Stanford University, Winter 2021

Background on Molten Salt Reactors

Fig. 1: Schematic of a Molten Salt Reactor with a two-stage heat exchanger. (Source: Wikimedia Commons. Courtesy of the DOE)

Molten salt reactors (MSRs) have received significant attention from both private investors and government agencies in the past few decades. Proponents of these reactors tout that MSRs could pave the way for reactors that are less expensive to operate and inherently safer than current reactors. [1,2]

Operating Principles

MSRs have fundamental similarities to the light and heavy water reactors (LWRs and HWRs) currently in active use in that all reactors use the thermal energy released from nuclear fission events to ultimately generate electricity in steam-powered turbines. [3,4] MSRs differ substantially from LWRs and HWRs in both design and operation though. LWRs and HWRs use solid rods as fuel sources and water as both a moderator and heat transfer medium to generate steam; whereas MSRs use molten salt solutions of dissolved fissile material (U and/or Th) to act as both a fuel source and heat transfer medium. These salts are typically lithium halide based salts, designed to achieve reasonable operating temperatures (700°C) and suitable fissile material solubility. [5,6] Fig. 1 shows a generic schematic for a MSR power plant design. Purified fuel-salt flows into the reactor core where fission events will take place. This heats the fuel-salt that then eventually exchanges its heat with water to generate steam.

Safety Features of MSRs

Safety features are built into the reactor and materials design of MSRs. The first, and potentially most crucial safety feature, comes from the formulation of the fuel salt itself. The fuel-salt is designed with a low enough concentration of fissile material that the only time a fission reaction can be sustained in the fuel-salt is when the fuel-salt nears a moderator (graphite in MSRs). [7] This only occurs in the reactor core and means only a fraction of the total nuclear fuel is undergoing a sustained fission reaction at any given time, a stark contrast to LWR and HWR designs where the all the fissile material is present in the core and reactor safety is maintained with control rods. [3] The second safety feature comes from physical properties of the fuel-salt. The fuel-salts are known to physically expand on heating to such an extent that temperature increases in the core will push fuel-salt out of the core and decrease the amount of fissile material in the core. [6] The fuel salts are also known to decrease neutron production on heating which further decreases the amount of fission events taking place in the core in the event of a temperature rise. [6] The combination of these two features lends a unique safety advantage to MSRs compared to LWR and HWR designs and essentially precludes the possibility of a fission run-away event - what researchers at Oak Ridge National Lab termed a "major accident". [5] A core temperature rise due to increased fission activity (which can only occur in a small amount of fuel) would self-correct and decrease the neutron population of the core. The design and material properties of MSRs provides impressive safety features, but also presents other safety concerns.

Safety Concerns

An important concern comes from the method of removing contaminants and fission by-products from the fuel-salt. Fission reactions generate products that are strong neutron absorbers (neutron poisons) that can destroy the neutron population in a reactor. A well-studied example of this is the generation of the strong neutron absorber, Xe-135, as a decay daughter of various species. Xe-135 can be removed by operating the reactor carefully such that the buildup of Xe-135 is balanced by its removal so-called burning of the Xe-135. However, other neutron poisons can not be removed in a reasonable time scale during the reaction and are only eliminated by physically removing fuel rods that contain these poisons. [8] This presents an interesting design challenge for MSRs that do not have individual fuel rods, but instead contain a continuously circulating fluid. This was a recognized problem as early as the first MSR experiments performed at ORNL, and the proposed solution involves an onsite processing facility (chemical processing plant in Fig. 1) for cleaning the fuel-salt after it has spent time in the reactor and generated neutron poisons. [5] However, this solution leads to a general design challenge of moving radioactive waste safely from the reactor core to the processing facility and handling radioactive contamination in the structural components of the processing facility (e.g. pumps, pipes, etc.). The structural components are primarily nickel-based alloys that withstand corrosive attack from the fuel-salt, and radioactive isotopes of nickel, such as Ni-59 and Ni-63, are long-term storage issues that present radiation hazards for greater than 1000 years. [9,10] Handling contaminated structural components does not necessarily present an impossible challenge: reprocessing of spent fuel and structures around the spent fuel already does occur on the order of ~4000 metric tons of material per year, and structural components of MSRs would in principle be designed for durability. [5,11,12]

A more difficult challenge arises for handling the fuel-salt waste from the reactor. Extensive work was performed in the 1970s at ORNL to identify fission by-products in the MSR waste and to develop processes for "cleaning" the fuel-salt of certain neutron poisons and contaminants. [13,14] However, cleaning the fuel-salts still generates hazardous waste that must be dealt with. The ORNL scheme for cleaning the fuel-salts of a MSR highlights the kind of waste products that exist. For the 2250 MWth MSR developed at ORNL, the fuel-salt cleaning process was estimated to generate 2 m3 of waste every 220 days. [15] This particular process was designed to recover U-233 present in the waste stream. The final waste composition was 76.3-12.3-9.8-0.64 mole % LiF-ThF4-BeF2-Zr4 and 0.96 mole % rare earths (including Sr-90 and Zr-95). [15] Though modern MSR designs will have varied waste streams from the ORNL experiments, common waste elements remain, including lithiated fluoride or chloride salts, left-over fuel (e.g. U or Th isotopes), and a host of fission by-products. [16,17] This is high level radioactive waste that must be managed carefully. Though management of high level waste already exists, MSR waste presents a challenge in that some form of containing the salt waste must be developed, since the MSR waste is corrosive (even in the solid state highly corrosive gases are produced) and is not the conventional fuel rods from LWRs or HWRs. [12,16] Current research in containing MSR waste is focused on sequestering the MSR waste in glasses, taking inspiration from the way high level waste in current spent fuel reprocessing is stored diluted in chemically inert glasses. [16] There is a lack of experimental studies on the long-term (>1 year) durability of chloride or fluoride containing glasses though, so developing strategies for containing MSR waste is still an active area of research. [16]

All nuclear reactors generate gaseous waste (mostly Xe-135 and tritium). [8] MSRs generate relatively large quantities of tritium, and this presents an important safety challenge. [16] Tritium readily exchanges hydrogen in with water vapor to form tritiated water vapor that condenses to become tritiated water, which is difficult and costly to separate from ordinary water. [18] This tritiated water is a potential environmental health risk that must be managed carefully. Tritium is a low energy beta emitter that has a relatively short half-life of ~14 days. [19] The potential for uptake in the human body though is worrying nonetheless, as tritiated water absorbs readily through skin. [18] Tritium is formed in MSR reactors primarily from the decay of Li-7. Though Li-6 is the dominant isotope of Li used in the MSR salt-fuel solutions, Li-6 has a large neutron absorption cross section (940 barns) and converts to Li-7 readily. [5,20] Concerningly, tritium was observed to diffuse through the nickel-based alloys that make up the majority of MSR components. [5,13] While estimates from ORNL in the 1970s place MSR tritium production less than HWRs of the time, more modern estimates MSR tritium production significantly higher than HWRs. [5,16] Any modern design of a MSR must have some system in place to capture and sequester tritium. There is some understanding for handling tritium capture in current HWRs, so this again does not necessarily present an impossible challenge. [18,21]


Research and interest in MSRs has persisted for more than 6 decades, but commercialization and practical use remains to be seen. While there are design and material features of MSRs that do provide safety benefits, such as the elimination of possible fission run-away events, there are still safety issues involving material handling that must be mitigated before commercialization of these reactors begins. Nonetheless, there is considerable interest in MSRs as an important reactor design for the future. [1]

© Garrett LeCroy. The author warrants that the work is the author's own and that Stanford University provided no input other than typesetting and referencing guidelines. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] D. LeBlanc, "Molten Salt Reactors: A New Beginning For an Old Idea," Nucl. Eng. Des. 240, 1644 (2010).

[2] B. M. Elsheikh, "Safety Assessment of Molten Salt Reactors in Comparison with Light Water Reactors," J. Radiat. Res. Appl. Sci. 6, 63 (2013).

[3] B. Zarubin, "Introduction to Light Water Reactors," Physics 241, Stanford University, Winter 2015.

[4] M. Cooper, "Molten Salt Reactors," Physics 241, Stanford University, Winter 2019.

[5] "An Evaluation of the Molten Salt Breeder Reactor," U.S. Atomic Energy Commission, WASH-1222, September 1972.

[6] J. A. Lane, H. G. MacPherson, and F. Maslan, Fluid Fuel Reactors (Addison-Wesley, 1958), pp. 569-577, 640-643.

[7] K. Furukawa et al., "Thorium Cycle Implementation Through Plutonium Incineration by Thorium Molten-Salt Nuclear Energy Synergetics" in Thorium Fuel Utilization: Options and trends, International Atomic Energy Agency, IAEA-TECDOC-1319, November 2002, p. 123.

[8] "Nuclear Physics and Reactor Theory, Volume 2 of 2," U.S. Department of Energy, DOE-HDBK-1019/2-93, January 1993.

[9] J. Sunde, "Material Corrosion in Molten Salt Reactors," Physics 241, Stanford University, Winter 2017.

[10] A. G. Croff, M. S. Liberman, G. W. Morrison, "Graphical and Tabular Summaries of Decay Characteristics for Once-Through PWR, LMFBR and FFTF Fuel Cycle Materials," Oak Ridge National Laboratory, ORNL-TM-8061, January 1982.

[11] "Nuclear Technology Review," International Atomic Energy Agency, GC(64)/INF/2, September 2020.

[12] E. D. Federovich, "Technical Issues of Wet and Dry Storage Facilities for Spent Nuclear Fuel," in Safety Related Issues of Spent Nuclear Fuel Storage, ed. by J. D. B. Lambert and K. K. Kadyrzhanov (Springer, 2007)

[13] R. E. Thomas, "Chemical Aspects of MRSE Operations," Oak Ridge National Laboratory, ORNL-4658, December 1971.

[14] J. R. Engel et al., "Molten-Salt Reactors for Efficient Nuclear Fuel Utilization Without Plutonium Separation," Oak Ridge National Laboratory, ORNL/TM-6413, August 1978.

[15] L. E. McNeese, "Engineering Development Studies of Molten-Salt Breeder Reactor Processing No. 9," Oak Ridge National Laboratory, ORNL-TM-3259, December 1972.

[16] B. J. Riley et al.,"Identification of Potential Waste Processing and Waste Form Options for Molten Salt Reactors," Pacific Northwest National Laboratory, PNNL-27723, August 2018.

[17] Technical White Paper, V 1.0.1," Transatomic Power, March 2014.

[18] J. E. Phillips and C. E. Easterly, "Sources of Tritium," Oak Ridge National Laboratory, TM-6402, December 1980.

[19] V. P. Singh et al., "Estimation of Biological Half-Life of Tritium in Coastal Regions of India," Radiat. Prot. Dosim. 142, 153 (2010).

[20] V. F. Sears, "Neutron Scattering Lengths and Cross Sections," Neutron News 3, 26 (1992).

[21] "Management of Tritium at Nuclear Facilities," International Atomic Energy Agency, Technical Report Series No. 234, 1984.