Overview: the Thorium Molten Salt Reactor
What is a molten salt reactor?
In essence, a molten salt reactor [MSR] is a vessel that contains a hot liquid salt which a nuclear reaction takes place. The salt consists of the nuclear fuel, and several other compounds that optimize the reaction, the heat transfer and the stability of the salt. This means the salt mixture is both the fuel and the coolant. This allows for very high heat production. Unlike the solid fuel in LWR’s, the liquid fuel can be kept in optimal condition at all times. (Hargraves & Moir, 2010, pp. 307,308), (Leblanc, 2009, pp. 1644,1645), (Hart, 2011, p. 17), (Serp, et al., 2014)
Molten salt reactor technology was developed in the United States mid-20th century at Oak Ridge National Laboratories (ORNL). This R&D programme culminated in the Molten Salt Reactor Experiment (MSRE)) during which the reactor’s principles were successfully demonstrated. Despite promising results, politics decided to prioritize the main competitor at the time: sodium cooled Liquid Metal Fast Breeder Reactor (LMFBR). Several reasons have been pointed out for this. The main reason was that with the choice for the LWR, the world had chosen for the U/Pu fuel cycle, and the LMFBR at the time seemed a logical next step within that fuel cycle. Oak Ridge’s proposal not only involved developing a new reactor, but also a new fuel cycle. In comparison, the development of the LMFBR seemed to be well underway. (MacPherson, 1985).
ORNL director Alvin Weinberg marks 6000 full power hours of MSRE operation
There are three main reasons for the superior performance of thorium MSR’s.
The first is the excellent heat transfer capacity of the molten salt, that allows it to quickly transfer large amounts of energy out of a relatively small reactor core.
Secondly, the liquid state makes it possible to keep the fuel in optimal condition at all times. In solid fuel reactors, the fuel can only be partly used because waste products build up within the material. In an MSR the waste products can be taken out of the fuel, even during the process. This allows for a higher ‘burnup’ of the fission fuel.
Thirdly, MSR’s greatly facilitate the use of thorium, which enables an abundant resource with hardly any long-lived nuclear waste production. Thorium needs to be activated by neutrons first, to become Uranium233, a kind of uranium that is a particularly good nuclear fuel. In an MSR this process can be managed effectively, because the salt can be continuously optimized for this process online. In solid fuel reactors this is not possible.
The fourth reason for the high energy production is the MSR’s high operating temperature (550-700 degrees Celcius) that yields higher efficiencies of electricity production.
Thorium MSR is safe
Three large historic nuclear accidents (Harrisburg, Chernobyl, Fukushima) have shown that the major risk during and after nuclear accidents lies in the spreading of radioactive iodine, caesium and strontium. These components are the product of the nuclear reaction and are formed in any reactor, thorium molten salt reactors included. And although other radioactive elements like molybdenum, ruthenium, krypton and xenon are formed as well, iodine, caesium and strontium are the troublesome ones because in today’s light water reactors they are present in volatile form, can become airborne, and can be picked up and partly remain inside living tissue.
In light water reactors in usual operation, these elements remain safely trapped inside the fuel elements and their cladding. But if an accident leads to their release, iodine and caesium may get into the atmosphere and can travel away from the reactor, over large distances.
In molten salt reactors iodine and caesium – and other fission products – are ionically bound. Ionic binding is an incredibly strong chemical bonding – it’s the reason why you can safely use kitchen salt, without having to worry about poisonous chlorine gas coming out of it, even though roughly half of your kitchen salt is chlorine.
In molten salt reactors, this ionic bonding makes sure that all radioactive components that provide a key radiological hazard are safely bound to the salt and are unable to travel by air.
Contrary to most of today’s reactors, the molten salt reactor is not pressurised and contains no water: there is nothing that could cause an explosion. Molten salt reactors therefore also have no ‘driving mechanism’ that would be able to spread the ionically bound radioactive components. Of course, a molten salt reactor will need adequate shielding, including protection from outside impact. These aspects will have to be included in a design with appropriate protection levels.
The danger of nuclear meltdown, which is generally viewed as a major concern in nuclear reactors, is simply not present in molten salt reactors because the fuel is not in a solid state. Meltdown occurs when the solid uranium fuel rods overheat to such an extent that the material melts, which can have dire consequences if the material then escapes its containment. In the MSR the fuel is expected to be in a liquid state and the structure is engineered to safely accommodate this.
Yet another boundary of safety in MSR’s is established by the reactive behaviour of the salt. When the salt is cooled (because the pumps are ‘on’), the nuclear reaction intensifies. When the salt heats up (the pumps ‘off’) the nuclear reaction slows down or even stops. This ‘load following’ behaviour is a convenient operating principle, but also serves as a fool-proof safety mechanism. It means that if for whatever reason the cooling pumps fail, the reactor heats up to a calculated maximum, then simply stops.
If for whatever reason the reactor heats up further, another safety mechanism gets activated. This is the so called ‘freeze plug’, also called ‘melting plug’. Both names apply to the same simple mechanism that consists of a section of salt in a drain pipe that is actively kept frozen by an electric fan. If the power fails, the fan stops, the plug melts and gravity makes the salt drain away to the safety of specially designed storage in which the decay heat is released by passive cooling.
The big difference with earlier solid fuel designs here is that instead of power being needed to shut down a reactor safely, power is needed to prevent the safe shutdown of a reactor. Therefore, in case control is lost the only logical outcome is automatic shutdown.
The properties described above combined can be summarized in the statement that molten salt reactors are inherently safe, meaning their safety does not depend on additional mechanisms that may require adequate handling in order to function properly. Molten salt reactors have also been described as ‘walk away safe’. This means that the reactor will shut itself down safely, even without any human intervention.
Finally, because of their small footprint, molten salt reactors can be built underground. Even though this will not always be necessary, building them underground provides extra shielding from external impact. (Hargraves & Moir, 2010, p. 310), (Leblanc, 2009, p. 1645), (Juhasz, et al., 2009, p. 4), (Kloosterman, 2016).
A clean alternative
Nuclear waste production will be very limited in the Thorium-MSR, due to the superior fuel utilization in the MSR, that results in smaller waste streams, that will be easier to manage.
An important difference between the waste stream of LWR’s and MSR’s is that in the former, the spent nuclear fuel is in fact a mixture of fission products, actinides and unused fuel, that make the waste stream long term hazardous, which causes expensive handling. In solid fuelled reactors, the fuel rods have to be removed after having only partially consumed the fuel rod’s energy potential. Replacing the rods is necessary because over time, due to the reactions that take place, the composition of the rod and its structure degrade. Whereas in a well-designed molten salt reactor, the liquid fuel can be kept in optimal condition at all times. Transuranics that are formed can remain in the fuel until they too undergo fission and are burned up, while the short-lived fission products that change the salt behaviour too much, or potentially deposit in the MSR system or decrease the reactor efficiency can be removed online.
It should be noted that options exist that would lead to better fuel utilization in LWR’s as well. These however require either fuel reprocessing, as is the present practice in France, or the use of fast reactors. These technologies have significant challenges however, most notably on the policy level.
In essence, one could say, the waste stream of MSR’s consists of ‘waste only’ – it contains virtually no actinides, comes in very small amounts and will be relatively easy and affordable to dispose of responsibly.
Of the MSR’s waste, 99,99% is stable within 300 years, instead of the controversial tens of thousands of years for conventional nuclear waste. The remaining 0,01% has no significant impact on the overall radiotoxicity (Hargraves & Moir, 2010, p. 308), (Hart, 2011, p. 17).
To put these numbers into perspective: the yearly consumption of electricity of the average affluent person (western standards of living) requires one gram of thorium per year. This thorium will be turned into one gram of fission products. Of this gram, about 83% only requires storage for 10 years. The remaining 17% needs storage for about 300 years, after which the radiotoxicity of this fraction is lower than that of uranium ore.
No added proliferation risk
There has been considerable debate about whether or not molten salt reactors add to the proliferation risk or reduce it. In many cases, these debates are conducted through the exchange of opinions on whether or not certain technical details are correctly displayed. Although useful, these debates are not conclusive on the proliferation resistance of future reactors.
It is good to notice that the authorities that decide on the allowance of nuclear technology take their task very seriously. If they are not convinced of the adequacy of the proliferation resistance of certain designs, licenses for such designs will not be granted.
An important aspect that is taken into consideration is whether or not a certain design adds to the proliferation risk. The countries with the ability to develop MSR-programs are countries that presently already master the uranium fuel cycle. Countries that have developed those skills and infrastructures already have the possibility to produce weapon material, and developing MSR technology does not add proliferation risk.
Some MSR designs are specifically made to offer the highest resistance to proliferation risks. Reactors based on those designs may be suitable to export to countries that are presently ‘non-nuclear’.
Malicious outsiders form a different type of risk, and MSR’s will probably make it as difficult to obtain any weapon grade material as it is to obtain such materials from today’s nuclear power plant’s – and nearly impossible to do so unnoticed.
MSR designers have new issues to deal with in regard to proliferation resistance – and have options to choose from in order to reach the protection levels that are required. For certain designs, concerns have been expressed about the so-called protactinium route, but here also holds that proliferation risk is not increased.
On the other hand, many protective mechanisms have been suggested that contribute to the molten salt reactor’s proliferation resistance. A well-known example is the thorium decay chain that produces uranium-233, which is always accompanied by uranium-232, of which the decay product produces strong gamma radiation, which is highly destructive to ordnance components, circuitry, and personnel, making it very difficult to work with. For usual operation this is no problem as the reactor will be adequately shielded, but this strong radiation makes unauthorised access virtually impossible.
This is just one example of many protective mechanisms in molten salt reactors that are available to designers, and need to be adequately combined into safe designs. For instance, due to the low fissile load of MSR’s, any removal of fissile material from the core would instantly show in its reactivity. This can easily be observed with remote monitoring. Remote monitoring could also be combined with remotely operated ‘fuel poisoning’ that instantly renders the fissile material useless as weapons material (Leblanc, 2009), (Hargraves & Moir, 2010).
MSR can cooperate with wind and sun
In the past, nuclear power plants have often been designed for ‘baseload’ supply and have been optimized for constant full power output. Other nuclear powerplants have been designed with the ability to ‘load follow’, so this specific ability of MSR’ is not new. That being said, the MSR can in principle automatically load follow due to some specific properties it has by nature. This will allow for flexible cooperation with intermittent energy sources like sun and wind (this same property has also been described as a safety feature). This means that molten salt reactors could provide a long sought solution to what many consider to be the Achilles heel for renewable power sources: the sun doesn’t shine at night, the wind doesn’t always blow. If the power generated by these sources falls, the molten salt reactor can compensate by generating more power – this is even possible without operator intervention.
This load following principle of an MSR-power plant is regulated by the laws of nature. If the salt is cooled (because the generator uses the heat to produce power), the nuclear reaction intensifies. If the salt heats up (because the power demand decreases and less heat is used) the nuclear reaction slows down or even stops. This is caused by changing density of nuclear fuel, which can change because a molten liquid can expand or shrink freely, and the fact that at higher temperatures, there is more absorption of neutrons by fission products and fertile elements. Both strongly influence the level of fission reaction in the salt.
Next to the fact that this avoids overheating of the salt, it also allows convenient load following, automatically, depending on the amount of heat extracted from the salt.
Molten salt reactors can be built in different sizes, and because of their inherent safety they can be built close to areas with high energy demand like cities. The ability to cooperate with sun and wind, and a high power production with a very compact design would offer a highly sustainable energy solution (Hargraves & Moir, 2010).
MSR’s can deliver heat for applications that would otherwise continue to rely on fossil fuel
A graph presented by the Lawrence Livermore National Laboratory indicates that industrial heat applications are virtually untouched by renewable energy sources. Nevertheless, these applications have a 24,5 per cent share of the US’s total energy consumption. For the EU, these figures will not be much different.
The primary energy delivered by MSR’s is heat that the molten salt delivers at 550-700 degrees Celcius. This high temperature allows for higher efficiency in electricity generation by employing so called “Brayton cycle” turbines. At the temperature range indicated, the Brayton cycle may yield efficiencies of around 44%. However, it may not be useful to apply Brayton cycle turbines as these will need to be further developed. The alternative is to use ‘standard’ Rankine turbines, which is the basic technology that is in use for both coal power plants and conventional nuclear power plants.
The high temperature heat can also be used directly for industrial applications – the 24,5 per cent pink rectangle of the above graph. MSR’s will also allow for co-generation, in which both electricity and heat for other applications is produced, and many other industrial applications, see In depth: clean industrial heat.
There is reason to believe Thorium MSR electricity will be significantly less expensive than electricity from current nuclear technology. Factors contributing to a lower cost profile are simplified construction including less expensive safety systems, modular construction, simplified fuel and waste handling, higher energy efficiency, and the potential for additional processes and earnings made possible by the thorium MSR’s high operating temperature, among others. Additional construction relating to components unique to the MSR such as the fuel salt processing system have to be accounted for as well however (Moir, 2001, p. 94), (Leblanc, 2009), (Hargraves & Moir, 2010).
Although the principle has been proven in the 1960s, large scale deployment of thorium MSR’s will require substantial upfront funding, necessary for materials research, design, numeric modelling, validating tests and development of a licensing framework. Cost competitiveness should be part of the development process: the operational costs of MSR’s will strongly benefit from standardized designs and a modular approach, suitable for industrial production of reactors and components.
The Molten Salt Reactor Experiment showed no fundamental obstacles for large scale deployment of the basic principle. Materials issues that showed after extended periods of operation were largely resolved during and shortly after the experiment. However, additional research will be needed: present day allowance demands will not be satisfied by referring to the original Oak Ridge National Laboratory experiments of the mid-20th century, as the data and information is outdated and not supported by current experience or safety standards.
Among the main challenges are development of the processes to continuously clean and manipulate the molten salt fuel, research and development into (a) structural material(s) suitable for containing the molten salt fuel for long operation times, limited graphite lifetime under molten fuel salt conditions, the availability of sufficient start-up fuel for large-scale thorium MSR deployment, and total development time which is likely to be on the order of 2 decades (Kasten, 1969, p. 1), (Macpherson 1985), (Le Blanc, 2009), (Delpech, et al., 2009), (Serp, et al., 2014), (Sietsma, 2015), (Kloosterman, 2016).
The key challenge however is the lack of a licensing framework for MSR systems. This framework will need to be developed from the ground up, but requires finalized designs. This dilemma may be overcome by developing numeric models for salt behavior, neutronics, materials behavior and salt/materials interactions, accompanied by assumption validation by representative testing.