Is the DFR a thorium molten salt reactor?

Alvin Weinberg was working towards a thermal molten salt breeder reactor based on thorium. From this, a popular generalization was born: Nowadays the notion has become popular that every molten salt reactor is necessarily(!) a thorium reactor. But the DFR neither needs to be powered by thorium nor molten-salt-fueled. There are two possible variants: DFR/s (salt) and DFR/m (metal melt). Thorium can be fed in as fertile fuel, but any other actinide of higher atomic weight will do the job as well. Depending on the nucleus’ stability, it either gets fissioned itself or transmuted into a fissile isotope.

There are no inherent advantages of thorium compared with uranium 238, except in areas where it is a common natural resource, such as India.

Isn’t nuclear waste annihilation a completely novel concept which needs yet to be proven?

The behavior of many different transuranics under neutron irradiation was tested extensively, most of all at Idaho National Laboratory (Fast Flux Test Facility). Concerning th DFR, two doctoral theses were presented at Technical University of Munich affirming its properties — including fissioning of transuranics — through simulation.

Half-lives of different isotopes can be determined experimentalle with extreme precision. While an individual nucleus’ decay is a stochastic event, the decay of a macroscopic amount of atoms occurs strictly deterministically. The future behavior of fission products and transuranics is known with mathematical exactness.

It should be noted, by the way, that the term “proof” cannot, from the standpoint of epistemology, be applied to principles of physics or technology. Only mathematical and logical theorems can be proven: This means their derivation via implication series from axioms (basic tenets which are recognized under certain circumstances as evident without need or possibility of proof) or theorems that have already been proven. Physical laws, on the other hand, rest on observations of nature. Compared to the theorems of mathematics, they are ascertained to a lesser degree; they can be checked through experimental tests of their predictions, but not proven.

This is, notabene, not the case for the laws of thermodynamics: They are based on stochastics and not on observations. All other physical laws, including very basic ones such as the theory of relativity and quantum mechanics, are viewed as being, in principle, falsifiable; the thermodynamic laws, on the other hand, seem to be secured for now and all times by pure mathematics.

Didn’t the Hamm-Uentrop experiment fail?

At Hamm-Uentrop, a gas-cooled pebble bed reactor was tested. This technology has next to nothing to do with the DFR (except the possible use of thorium as an energy source). Pebble bed reactors do not use fluids, they work with thermal neutrons, are not able to breed and, most of all, stand out by their extremely low power density to enable passive safety. The DFR, on the other hand, is the first reactor to combine passive and inherent safety with high power density.

Don’t thorium teactors create long-lived waste, too, needing geological storage?

There are two misunderstandings in this question: First of all, while the DFR can use thorium, it is only one of many possible fuels (see above). Secondly, storage time depends only weakly on the nuclide fissioned, the determining factor is rather the efficiency of reprocessing extracting the different element groups from the fuel mixture.

All actinide nuclei create upon fission a similar (though not completely identical) spectrum of medium-weight, mostly unstable isotopes. This fission product mixture (“fission ashes”) radiates intensively at the beginning, but decay is rapid: After some decades, most nuclei have become stable; after 300 years, radiotoxicity is weaker than that of natural uranium. Additionally, neutron capture creates transuranics in the reactor. If left in the fuel, as with the “once-through-cycle” of the light-water reactor, waste decay time rises to hundreds of thousands of years.

The DFR retains all actinides until they have been fissioned: Only fission products leave the fuel loop! They need storage for no longer than a few centuries (many can, in fact, be extracted earlier and put to industrial use), irrespective of whether they come from uranium 233, plutonium 239 or another actinide.

Is the DFR a hypothetical futuristic technology like fusion?

Nuclear fusion and fission do not have much in common: While neutrons enter a nucleus effortlessly to trigger fission — and thus a certain arrangement of fissile nuclides will become critical automatically, and whether this takes place can be predicted with mathematical exactitude —, massive amounts of energy need to be expended to fuse to positively charged nuclei (they repel each other). For this reason, fusion reactors are extremely complex and can, as of 2017/18, not be used for net energy production.

The DFR, on the other hand, is based on established or at least tested technologies — fission in a liquid medium, lead cooling, breeding of fissile material, thermal distillation etc. The central idea is to re-purpose proven — and, in part, industrially applied — technologies and combine them into something new. Fusion still needs basic research to make it work.

Who says this reactor is working?

Whether a certain arrangement of materials creates a self-sustaining nuclear chain reaction (“going critical”, i.e. works as a reactor) is a purely mathematical problem, which is, similar to the problems of celestial mechanics, precisely solvable. This was done at Technical University Munich in two doctoral theses — with a clear answer: Yes, the DFR works as desired.

This is not surprising if one takes into consideration that the DFR is in fact nothing more than a metal-cooled fast reactor with a single fuel element. Connect the fuel rod hulls of a sodium- or lead-cooled reactor to each other and fill them with uranium-/plutonium chloride instead of the oxide pellets — and you have a DFR.

Is it a fast reactor?

Yes. The neutron spectrum is in fact very hard (i.e. the number of neutron with high energy is far greater than that of slower ones), thanks to the low moderating properties of lead and the absence of fuel dilution. The extremely hard spectrum is also responsible for the high neutron excess which can be used for breeding, transmutation or isotope production.

What kind of salt is used in the fuel?

Fluorine salts have still considerable moderating quality thus softening the neutron spectrum and deteriorating the neutron economy. Furthermore, many of the involved metal fluorides have a boiling point too high for an effective online reprocessing in the PPU. Higher halogens are more practical with respect to both properties. For the metals in the used fuel mixture, chlorine salts have sufficiently low boiling points yet still higher than 1000 °C as required in the DFR core.

How does reprocessing of the salt work?

Thanks to the separation of fuel and coolant loop, the fuel can be reprocessed online in the PPU. Dry high temperature processing can be used in combination with the fuel cycle. Due to the ionic nature of the bond, the used fuel salt is impervious to radiolysis and as such directly apts for physicochemical separation methods at high temperatures. Two such methods have been proven in the past:

  • Molten salt electrorefining as for the IFR project of Idaho National Laboratory;
  • MSR-style high temperature distillation.

Both are suitable for the DFR. The capacity of the pyrochemical facility can be reduced substantially as processing is performed online continuously. In a simple version, the electrorefining method is appropriate to purify the fuel salt by precipitation of a fission product mixture. For the purpose of specific transmutation, a more precise separation is required which can be accomplished by fractionated distillation / rectification going beyond the MSR principle. Thanks to the low boiling points (still over 1000 °C as required in the DFR core), a separation in a fractionated distillation facility alone becomes feasible.

Which size and shape does the reactor have?

It is cubical due to the durable metal alloys for the fuel ducts being difficult to process. Non-curved parts are easier to manufacture into the fuel duct assembly. The size of the reactor vessel is approximately 2 m edge length for a power output of 1 GWth.

What is a typical actinide mixture and concentration in the fuel?

Nominally, the fuel consists of undiluted actinide salt. Its composition is very flexible, though, and can be taylored to specific applications.

In any case, sufficient fissile material (i.e. U 233, U 235, Pu 239, Pu 241) needs to be contained in order to keep the reactor critical. Minor Actinides (which are fissionable) may contribute, too. The other fractions are fertile material (i.e. U 238, Th 232) and possibly a small amount of transmuted material like long-lived fission products.

A small system with 1 GWth working in the U-Pu cycle has a concentration of 35% plutonium and 65% (depleted) uranium. With increasing reactor size, the concentration of fissile material declines.

What happens to the fission products?

Fission products produce decay heat, the more the fresher they are. Therefore, they must be stored within a container that allows sufficient passive cooling. In solid-fuel reactors like the ones in use today, there is no way to avoid the accumulation of fission products in the fuel rods while the reactor is running. For this reason, active cooling is required, even after the reactor has been shut down. The requirement of active cooling is the biggest safety challenge for solid-fuel driven reactors and lead to the problems in Harrisburg and Fukushima. With the DFR, all this changes. Since the liquid fuel circulates through the PPU outside the reactor core, fission products can be continuously separated from the fuel and therefore can’t accumulate in the reaction zone. Outside the reactor core they can easily be stored according to the usual safety standards for radioactive waste treatment. When the DFR is shut down, no active cooling is required. Furthermore, since the DFR is able to transmute fission products in a very effective manner, medium-lived fission products like Sr 90 could also be de-activated in the system which would further reduce the waste storage size. Assessments for these possibilities are in progress.

How proliferation-resistant is the DFR?

Breeding additional pure Pu 239 (DFR running in breeder mode) for nuclear weapons is not possible, because there is no separated breeding zone containing pure U 238.

Using a fractional-distillation PPU with the U/Pu fuel cycle, you have to remove Np 239 (half-life about 2 days, decay to Pu 239) very fast to obtain high purity Pu 239. This is hardly possible.

Since the DFR’s nuclear part is fully capsulated and watched telemetrically by anti-proliferation authorities, there is no way to extract weapons-grade material. Using an electro-refining PPU makes things even more difficult for bomb makers, as it can only distinguish between actinide salts and fission product salts. The highly radioactive Np 239 must be chemically purified in a separate facility in a timeframe between hours and one day, which is nearly impossible as such facilities would be phased out — they are not necessary for the civil use of the fuel cycle.

When using the Th/U cycle, the U 232 isotope, synthesized via (n,2n) reactions, generates intense hard gamma radiation, that can be detected easily and causes serious damage to weapon electronics. The fractional-distillation PPU is secured in the same way it is in the U/Pu cycle, preventing the Pa 233 being captured by bomb makers. Compared to the well-known standard method for bomb making, the enrichment of weapons-grade uranium from natural uranium, both possibilities mentioned above are far more difficult to realize.

Is shale gas power not cheaper than nuclear power?

Cheaper than from the DFR? No way!

Long answer: Gas power plants have the lowest construction costs of all power plant technologies, since they are not much more than a turbine with a generator, a turbo-compressor and a comparatively tiny combustion chamber directly in front of the turbine. However the “combustion chambers” of coal plants are much larger and more complex — similar to a nuclear plant with additional abundant safety equipment. Therefore the investment costs for gas plants are low, but the effort to provide the gas is much higher than for coal plants let alone nuclear plants. This reflected in the EROIs which are similar for gas and coal plants and almost 3 times lower than those of nuclear power plants. Natural gas prices are artificially super-elevated by the cartel of countries which possess abundant easily extractable gas deposits. Thus, gas power plants have the highest production costs of all conventional power plants. Unconventional shale gas deposits are more evenly distributed on earth, and in spite of their considerable higher extraction costs they reduce the market prices of gas as they reduce the controlling power of the cartel. Despite the even lower EROI of a gas plant fueled with shale gas, it can be attractive to investors because the investment costs are low and the electricity production costs competitive though larger than those of coal plants. This means a shorter financial amortization time than for other power plants. The head start in the EROI of PWR plants is diminished by political costs like exuberant licensing procedures. It is therefore necessary to exploit the vast potential of nuclear fission to a larger extend than possible with existing nuclear technologies. That can be achieved by a high powered reactor core with an overall simplified design and inherently safe and simple operation. Also the costly fuel cycle industry needs to be abandoned. Precisely that is implemented by the DFR with an EROI on the order of thousands.

Isn’t corrosion at 1000 °C a great problem?

Using fast neutrons, the material palette for the DFR is markedly broader than for thermal reactors. (At higher energie, neutron interaction cross-sections tend to be smaller.) Most material problems were solved as early as the MSRE project. In the non-nuclear industries, a large selection of robust, heat resistant materials was developed, which can be be adapted to the DFR. These material are costly, but due to high power density and absence of fuel elements in the DFR, only small amounts are needed.