IFK logo
Institute for Solid-State Nuclear Physics
Institut für Festkörper-Kernphysik
English flag
Main page
Dual Fluid Reactor
About us
German only

How the DFR works

Historical concepts

Nuclear Energy today

The discovery of nuclear fission was probably the most important event in science history. It offers a sustainable and extremely safe energy supply sufficient for all humans. Nuclear fuel resources are available in all countries in amounts that are sufficient for millions of years.

This is in stark contrast to the way nuclear fission energy is "produced" these days. Barely one percent of the original natural uranium mass is used up in the fuel rods for energy production in water-cooled reactors, the remainder is treated as radiotoxic waste, transported around the world, perhaps recycled, and eventually buried deep-underground (somewhere at some point in the future). Moreover, the nuclear fuel is burned at an inefficiently low temperature, which is too low for any interesting chemical applications.

However, despite its inefficient use, the energy density of nuclear fuel is so vast compared with other energy carriers (or compared with solar irradiation density on Earth) that the advantages of those reactor types still prevail, even after 60 years of basically unchanged technology, though safety has been improved as much as possible. However, the established fuel enrichment/recycling industry working hand in hand with the nuclear waste transportation and the nuclear reactor industries has resulted in stagnation of the entire field. These structures are so settled and well entrenched that there is no interest in other nuclear reactor systems. There is a closed enrichment/recycling/waste market that does not permit waste-free or enrichment-free nuclear concepts were proven to be reliable and more efficient half a century ago.

The Generation IV concepts, though feasible today, are postponed to the future by decades. In particular the MSR was tested in the late 1960s in the U.S., running for years with great success. The thorium high-temperature reactor THTR was developed in Germany in the 1970s, also running for years with great success. All these market-ready developments could not penetrate the well-established fuel enrichment/recycling/waste industry based on the ineffective pressurized water reactor technology.

The water-cooled fuel rod technology emerged essentially from the first military concepts of mobile nuclear reactors for the specific propulsion of naval submarines and carriers. This technology, together with the fuel enrichment and recycling industry, is not really made for civil applications, however, it is established all over the world and is now the only available reactor technology on the market. It is expensive, produces a lot of nuclear waste, and needs a complex infrastructure.

The MSR concept

The MSR experiment (MSRE), performed in the 1960s at Oak Ridge National Laboratory, proved that the molten-salt technology is reliable and easy to handle. All material problems were solved and the test reactor was running for years. However, it was still a thermal or epithermal type of reactor which typically produces more actinides than it burns. Recently, a calculation performed by members of the Generation IV forum has shown that the MSR works even better with no graphite moderator at all, making it a fast neutron type of reactor.

The efficiency of the MSR is reduced by the double function of the fuel to also act as coolant. As a result, the molten salt used had to be diluted in order to limit the power density, otherwise the heat could not be removed fast enough. Furthermore, salts with low melting point are necessary for the effective utilization of a heat engine. In addition, the salt has to circulate fast for efficient cooling and that, in turn, prevents any on-line reprocessing of the fuel. The fuel, thus, needs to be processed off-line (but still on-site) at regular intervals. Off-line processing of the fuel requires long shut-downs, further reducing the efficiency of the overall system. Several techniques exist that improve the fuel cycle by extending the intervals between shut-downs, however, these techniques worsen the neutron economy and thus reduce the transmutation turnover.

Lower power density means a larger reactor volume which allows the use of cheaper and well workable structural materials. However, these materials suffer from corrosion problems at high temperatures, aggravated by the unavoidable production of significant amounts of hydrofluoric acid during the regular operation of the reactor. This problem limits the operation temperature of the reactor core to 650 °C.

Liquid metal cooled reactors

Liquid metal as coolant simultaneously renders the neutron spectrum fast and allows for the heat removal of a high powered core which is implied by the low cross sections of fast neutrons. Contrary to water coolant a costly pressure vessel is not required. Historically, the envisioned application of a high powered metal cooled reactor for the propulsion of strategic bombers in the 1950s caused a lasting fixation on sodium as coolant in spite of its engraving disadvantages because of its low density which was mandatory for the airborne reactor project. Consequently, all experimental fast reactors up today were cooled with liquid sodium. A fast reactor that was actually deployed powered the soviet Alfa class submarines and used a liquid lead bismut alloy as coolant. The Alfa submarines were in operation for 2 decades and were decommissioned at the collapse of the USSR. Both reactor types joined in the generation IV canon.

Nowadays concepts for larger power plants place the reactor core inside a multiple times larger pool of the liquid metal coolant in order to remove by convection the residual heat from the decay of fission products accumulated in the fuel rods in case of a loss-of-power accident . Sodium, besides its well known aggressive reaction with air and water also has disadvantageous neutronic properties. In comparison to lead it has a considerably higher neutron absorption cross section with resulting activation and still sizable moderation quality. A softer neutron spectrum also reduces the neutron economy since the number of released neutrons per fissioned nucleus inclines with the incident neutron's energy. The high neutron absorption of sodium in combination with its low boiling point (883 °C) may enable a temporary positive void coefficient with a power excursion when sodium vapour bubbles form and reduce the neutron absorption (a density decrease in lead reduces foremost the neutron reflection). As a countermeasure a pressure vessel for the reactor pool and an intermediate sodium coolant loop is necessary because of the high radiation intensity of activated sodium and its flammability in case of a leakage in the steam generator is necessary. In summary, all of these measures render SFR's uneconomical in comparison to water cooled reactors.

Absorption of neutrons by nuclei in the liquid-lead coolant produces most frequently stable nuclei again. So even after long operation the radioactivity of the coolant is low. Together with the chemical inertia of lead, an intermediate coolant loop is obsolete. The excellent neutronic properties of lead offer many options for reactor design, including long refueling cycles, since neutron poison from fission products is less impairing, and maximized transmutation performance. The high boiling point of lead (1749 °C) enables operating temperatures in the very high temperature range, suitable for process chemistry. However, the LFR concepts remain stuck at temperatures below 700 °C because the employed exchangeable fuel elements require cheap structural materials which are dissolved by liquid lead at elevated temperatures. The dissolving properties of liquid lead towards steel are to be reduced by aluminium plating and regulated oxygen admixture to the lead.

The DFR concept

The DFR incorporates elements of the MSR and liquid-metal cooled fast reactors and extends their advantageous properties.

Figure 1. Dual Fluid reactor core. The number of tubes, actually ten thousands, is reduced for illustration.

Dual Fluid reactor core. The number of tubes, actually ten thousands, is reduced for illustration.

The key feature of the DFR is the removal of the heat by a separate coolant loop, contrary to the MSR where the heat has to be removed by the fuel itself.

Using separate loops for fuel and cooling has many advantages while it simplifies the overall concept. It allows for a very high power density going along with inherent safety thus rendering the DFR very economically competitive. A main advantage is that the DFR works with very fast neutrons and a very high neutron excess. The high neutron excess can be used to dispose of long-lived nuclear waste produced by todays thermal reactors, in parallel to its own power production. The DFR can be built at low costs without the need for extensive material research.

The fuel can now circulate as slowly as necessary to provide an optimum burn-up rate while the coolant can circulate as fast as necessary to provide the best heat removal. As a consequence, undiluted molten-salt fuel can be used and the amount of fissile materials in the fuel can be increased significantly. This not only boosts the power density but also raises the fuel's boiling point and therefore allows for a very high operational temperature. This combines the advantages of the Generation IV concepts of the MSR and the VHTR.

As the choice of coolant materials is not restricted by the fuel anymore, liquid lead can be used which not only has the advantage of an excellent heat transfer but also makes it possible to operate the DFR as a fast reactor. Fast reactors have a net neutron excess which can be used for breeding or for the deactivation of long-lived fission products. This is further amplified by the employment of chlorine salts (preferable 37Cl) which are much less neutron moderating and corrosive than fluorine salts.

The material which separates the two fluids must have sufficient heat conduction and resistance to corrosion by the salt and the liquid metal. In comparison to the conditions in thermal neutron reactors the choice of nuclides for the structural wall material opens widely because of the low neutron capture cross sections for fast neutrons. Appropriate materials were developed decades ago; however they contain rarer and more expensive chemical elements. Contrary to solid fuel elements which are changed regularly, all DFR materials remain in the core and so even rare noble metals might be used as ingredients for the alloys with low impact on the total costs of the plant. Examples for such materials are the alloys of the refractory metals which are being used more and more in heavy duty applications in industry over the past decades.

In this manner the DFR solves the material problems of the MSR as well as the LFR which both must rely on cheaper materials in order to remain economically feasible.

Physical control loops

Figure 2. Dual Fluid reactor physical control loops

Dual Fluid reactor physical control loops

The two loops, for the fuel (red) and for the coolant (green), are shown in the figure. The liquid lead coolant leaves the reactor core at a high temperature and moves into the heat exchanger where it transfers heat to another medium. It leaves the heat exchanger at a lower temperature and enters the reactor again.

When the molten-salt fuel moves through the reactor the chemical composition changes by transmutation, fission or breeding. The chemically modified fuel leaves the reactor core and enters the pyroprocessing unit (PPU). Here, short-lived fission products, medical isotopes, or bred actinides are separated out while natural or depleted uranium or thorium, used fuel pellets or long-lived waste from conventional nuclear reactors can be mixed in. The PPU provides the right mixture that becomes critical in the reactor core.

The circulation speed of the fuel loop can be adjusted for various purposes such as maximum burn-up, transuranium element incineration, isotope production, fertile material conversion (aka breeding), specific deactivation of fission products, etc.

A very important property of a nuclear reactor is the temperature coefficient. If positive, the fission rate increases with temperature which can lead to dangerous power excursions as it happened with the Chernobyl reactor. The temperature coefficient of the DFR is negative, so deeply that the reactor is self-regulating: The fission rate follows the power extraction. If the lead circulation slows down, the fission rate slows down as well.

The melting fuse plug (already proven for the MSR) is a fuel duct section that is constantly being cooled so that the molten-salt freezes in it and seals it up. The DFR is shut down by simply stopping the power supply for the fuse plug which melts it, allowing the molten-salt fuel to flow into the subcritical storage tanks where the low residual decay heat can be removed passively. The same happens in case of a power outage or in case of a core temperature that is too high. There is no difference between a regular and an emergency shutdown. Therefore, for all known typical dangerous reactor accidents like 'loss of power accident', 'loss of coolant accident', 'criticality accident', 'decay heat' the DFR behaves well mannered like for a regular shut down.

Those safety features originating from the MSR are combined with the excellent neutronic properties and the high heat transport capacity of liquid lead coolant into a compact reactor core with a very high power density and all the same inherent safety. Thus the technical prerequisites are in place to render the DFR the most cost-effective power generation system.


In contrast to present reactors the transmutational power of the DFR can be maximized because of its optimal neutron economy. This is achieved by the design and the continuous removal of fission products (neutron poisons).

0.2 neutrons per fission are already sufficient for the continuous processing of the self-generated long-lived burn-up; thermal reactors do not have the necessary neutron excess. In the DFR fissioning plutonium 0.6 neutrons or more are available for that, thus at least 0.4 neutrons may be used for the transmutation of external fission products.

Inside a 1 GWth-plant up to several kilograms of additional fission products would be transmuted annually, leading to a reduction in amount of long-lived isotopes in the interim storage facilities. In the pyrochemical facility valuable materials for other applications can be extracted from the stable fission products. In addition, up to 400 kg of transuranium elements may be fissioned.

According to the German Federal Office for Radiation Protection (Bundesamt für Strahlenschutz) 100 tons of transuranium 'waste' will have been produced by the existing nuclear power plants until the year 2040. Through transmutation of the transuranium elements the storage time for the radioactive waste is reduced from several 100,000 to some 100 years. Simultaneous transmutation of long-lived fission products limits the necessary storage time to 300 years.

If the neutron excess is employed for the transmutation of fertile material (238U, 232Th) the DFR operates in the breeder mode. With the U-Pu cycle the DFR has a doubling time for the initial inventory of another reactor down to 4 years. Hence the doubling time is comparable to the total construction time of a power plant and not the limiting factor for deployment. For comparison, SFR's (like the french Superphénix and the russian BN) together with PUREX-reprocessing plants have doubling times of 30-40 years. Utilizing the Th-U cycle in water cooled reactors with fuel elements would exceed even these long doubling times. The thermal thorium MSR (aka liquid fluoride thorium reactor - LFTR or "lifter") has a doubling time of 25 years. This is because the fission of 233U yields considerable fewer neutrons. Though in the very hard neutron spectrum of the DFR even the neutron yield of 233U would improve.

Heat Utilization

The heat Utilization module for electricity production consists of a closed supercritical water cycle with turbine generator pairs. The electric efficiency of the system would be about 50%. The residual heat content after the turbine needs to be discharged. This can be accomplished in the usual way as for any other power plant such as sea water cooling at coastal locations, wet cooling tower, and dry tower cooling. The last option is most wearing on the efficiency but does not rely on a constant water supply while for the first option it is parenthetically possible to cheaply desalinate sea water since the coolant temperature is still high enough.

DFR systems can be equipped with specially manufactured heat exchange modules for the transduction of chemical process heat for different industrial purposes. Here, peripheral technology being developed for the VHTR can be employed. In combination with electricity producing DFR's any industrial plant can be energized completely nuclear.

The technical properties of the DFR like high operating temperature and liquid metal coolant enable further development options for the conventional part. One option is a noble gas turbine instead of supercritical water. Another option is the installation of a magneto-hydrodynamic generator in a liquid metal topping cycle reducing the size of the turbo-generators and the overall costs while increasing the efficiency. A similar possibility is an AMTEC (Alkali-metal thermal to electric converter) which has no moving parts, too.

DFR Applications

Figure 3. Applications of the Dual Fluid Reactor

Applications of the Dual Fluid Reactor

The heat can be used to provide electricity with an efficiency of 50% or higher. The gas turbine outlet temperature is still high enough to drive processes like water desalination on a large scale as it is desired in many desert countries. Alternatively, the high temperature can be used for the production of petrochemical products, an essential part of the modern chemical industry, more cost-effectively.

Most interesting is the production of hydrogen, the base material for many chemical procedures. Current steam reforming and similar processes are CO2 intense and consume fossil fuels. At the high temperature of the DFR, hydrogen can be produced from water by high-temperature catalytic thermolysis at high efficiency. Ammonia and synthetic fuels like hydrazine can therefore be produced completely CO2 neutral. Energy density and chemical toxicity of hydrazine are similar to gasoline [1], though with the DFR it can be produced for 1/3 of the current pre-tax gasoline market price.

Low hydrogen costs also make the hydrogen reduction of iron for steel production very attractive. Replacing the coke reduction reduces the CO2 footprint of the steel industry remarkably.

Automotive fuel production

With respect to the high and conceivably rising costs of natural gas and petroleum the nuclear synthesis of fuel is a worthwhile business proposition, especially since it can end the dependency on producer countries in a timely manner.

These synthesis processes are well-tested in chemical engineering. However, they are not very energy efficient. Thus, for the past several years more sophisticated processes have been being developped to achieve significant improvements in this respect. In particular, researching catalysts leads to substantial efficiency enhancements. Thus, ruthenium for example can increase the efficiency of the Haber-Bosch process for nitrogen fixation considerably. The required reactant hydrogen peroxide for the alternative hydrazine synthesis of the Pechiney-Ugine-Kuhlmann process can be produced with higher energy efficiency via the NxCat process. Also for a hydrazine fuel production these developments will lead to a reduction in costs.

A recently developed new method, the "Solid-State Ammonia Synthesis" SSAS, bypasses the hydrogen production and synthezises ammonia directly from water and nitrogen. This would reduce the capital and maintenance costs by a factor of 3, and the energetic production costs by 25%, therefore further halving the total ammonia production costs. SSAS has not yet been applied on a large scale but it provides interesting prospects, in particular when combined with the DFR.

Synthetic Fuels

The operating temperature of 1000 °C provides for the highly efficient production of hydrogen from water through combined electrolysis and thermal decomposition. Such a process, the HOT ELLY process, was developped for the high-temperature reactor at the Jülich Research Center in Germany. Alternatively, the sulfur-iodine cycle process produces hydrogen employing only thermal energy of >830 °C. Hydrogen gas, even highly compressed or cryogenic liquid, has a low energy content and is difficult to handle. Moreover, both processes utilize a significant amount of the energy content of the hydrogen, that cannot be regained. Hence, hydrogen as a fuel for vehicles or aircraft is impractical. Therefore, it seems to be more useful to concentrate the hydrogen in fluid chemical compounds, that are easy to handle.

These synthetic fuels are also known as XtL-fuels, where the substance 'X' is converted to Liquid.

The relevant industrial scale synthesis processes are well-developped and used commercially - mainly for chemical engineering in the primary production. They all have in common, that these synthesized fuels made from fossil fuels are obviously wasteful compared to petroleum products. Therefore, they are only used for special applications (rocket propellant) or if the absolute rule of the state takes priority. With the use of 'renewable' energy this becomes completely uneconomical. The situation changes, if nuclear thermal energy is utilized.

Nuclear thermal energy combined with the previously generated hydrogen allows to hydrogenate coal via the Bergius process, by which synthetic benzine or diesel could be formed. (CtL-fuel: Coal to Liquid). Instead of hydrogen, the Fischer-Tropsch process may employ syngas.


If the exploitation of coal is to be avoided, because it is not available or the generation of carbon dioxide is not wished-for, atmospheric nitrogen may be used instead of carbon. Here, the synthetic fuel of choice would be hydrazine (N2H4), a liquid fuel with similar properties to benzine (including toxicity). Hydrazine has been used as a rocket propellant for 80 years. Produced by nuclear energy it becomes an affordable alternative to petroleum products for the use in transport. As liquid fuel it can be combusted in piston engines of vehicles and in turbines of aircraft with only minor modification. These adaptions are similar to those for other alternative fuels and affect mostly the quantities of injected fuel and the ignition points. Here, the combustion is a very clean process producing water and nitrogen. Similar to the combustion in air, the production of nitric oxides only depends on the temperature. Hydrazine congeals at freezing temperature, which can easily be avoided by seasonal admixture of antifreezes, such as methylhydrazines or alkanols. In contrast to benzine, it decomposes quickly, if being released into the environment perhaps by accident. In fuel value it is equivalent to LPG (liquefied petroleum gas) and half as much as benzine. This disadvantage in fuel value can be compensated by simply admixing water, since heat engines depend on the difference in pressure and not on the temperature. By adding water the energy content of a fuel can be utilized more efficiently, because the vaporization of the water increases the pressure. This has been tested in the past, but never converted into commercial application, because benzine doesnt't mix with water, thus an additional tank for the water would have been required. In contrary, hydrazine smoothly blends with water, which in addition simplifies the ease of use. Thus, hydrazine may be considered an NtL-fuel: Nitrogen to Liquid.

With hydrazine, fuel cells may be produced at a lower price and be operated more efficiently than with hydrogen, because the power density and the level of efficiency are significantly higher, and no rare, expensive metals such as platinum for hydrogen fuel cells are needed for their production. In the past, hydrazine fuel cells have already been used successfully in spacecraft.

Taking nuclear produced hydrogen combined with atmospheric nitrogen and further nuclear energy ammonia is produced via the Haber-Bosch-process. Being the most important precursor in nitrogen chemistry today, ammonia is mainly used for the production of fertilizers. Today, more than 130 million tons of ammonia are produced via the Haber-Bosch-process, already consuming 1.4% of the global fossil energy sources. In this prospering market nuclear produced ammonia may provide a cost advantage, too. Among other things, it can be used for the production of hydrazine via the Olin Raschig process. The additional energy required for the production of hydrazine is released to a great extent in the combustion, as are the reactants water and nitrogen, thus the process may be regarded as a sustainable closed loop.

Pure ammonia itself does not burn in air under normal conditions because the flame temperature is lower than the ignition temperature. However, under certain conditions it may serve as a fuel. These are present in Diesel engines and in turbines with very high compression ratios. The advantage is the lower production cost of ammonia in comparison to hydrazine. Disadvantages include: Ammonia is a gas and can be liquefied under pressure having similar physical properties like LPG hence likewise requiring a pressurized gas tank. The fuel value is somewhat lower than hydrazine but due to the difficult flammability the efficiency boost by water admixture is hardly possible. The utilization of pure ammonia in Otto engines is not possible albeit a mix with hydrazine may work.


Silane is another synthetic fuel. Silanes are the silicon homologues to carbon-based alkanes. Like hydrazine they are produced endothermically, thus being capable of re-emitting part of the synthesis energy during the combustion. They are characterized by very high energy densities. Starting with heptasilane (Si7H16) they are stable and easy to handle. Heptasilane is liquid from -30 to 227 °C. It is produced from the reactants water and silicon dioxide (silica) via the Müller-Rochow synthesis, which can be performed at lower cost with nuclear energy from the DFR system. Water and silica re-emerge by the combustion. At combustion temperatures above 1,400 °C silane burns exothermically with atmospheric nitrogen to water and silicon nitride (Si3N4). The combustion products silicon dioxide and silicon nitride are solids. This is a substantial problem for conventional piston engines and turbines. Because of its high energy content it is perfectly suitable for ramjets at hypersonic speeds (RAM - SCRAM Jets) and spacecraft engines. Modified Wankel engines and external combustion turbines would allow to use silane in vehicles as well. Thus, silane would be a StL-fuel: Silicon to Liquid.

[1] Pure hydrazine has half of the energy density of gasoline. However, contrary to gasoline it mixes well with water which boosts the efficiency in combustion engines while reducing the toxicity and flammability.