Fusion Reactor: Blanket
The blanket is the material surrounding the plasma inside the vacuum vessel. It has three functions: heat conversion, shielding, and fuel breeding.
You might ask yourself, why are we talking about the outside wall of the reactor, the blanket? Isn’t the interesting part about a fusion reactor the fusion process? Well… true and and to make matters worse, we’ll later on discover that inside the blanket, no fusion but the inverse process (nuclear fission) is taking place. Yet, the blanket is an integral part of the fusion reactor design. So what do we need it for?
Once excess energy has been created by the fusion process we have not yet generated any heat nor electricity for that matter. In the following, we will discuss that the energy of the fusion process is carried by a neutron, converted into heat in the blanket, and then transported away from inside the reactor vessel to a heat exchange system, similar to the ones in other power plants.
Basics
Inside the reactor we have a hot plasma. Surely we can simply convert the heat of that plasma to electric energy. We do this for all sorts of power plants. Well, a fusion reactor is a bit different here.
Gas-fired plants heat up water and the steam drives a turbine that converts the thermal energy into electric energy. Nuclear fission reactors, both pressurized and boiling reactors, create steam that rises through a turbine to create electricity. The water is heated mainly by three processes: kinetic energy, gamma rays, and neutrons.
This is different in a fusion reactor. The most efficient type of fusion uses the heavier isotopes of hydrogen, deuterium (D) and tritium (T) in the fusion process where helium (He) and a neutron (n) is created. We write the equation so that the energy gain is visible for each individual product:
From this, we already see that around 80% of the 17.6 MeV energy gain is carried away by the neutron. It follows that our main job is to convert the neutron’s kinetic energy into heat. (The unit MeV refers to mega electronvolts, a unit that shows how much kinetic energy an electron gains when accelerated in a 1 volt potential)
Function
The blanket in a fusion reactor surrounds the fusion chamber and serves three purposes:
- It absorbs almost all of the neutron energy (~99%) and, therefore, heats up. The heat of the reactor is transported away by a coolant that flows through the blanket.
- Fuel breeding, the blanket can breed the necessary tritium that the reactor uses as fuel.
- It also acts as a shield. This requires strong materials and additionally those in which no long lived radioactive isotopes can be created by the neutrons.
Material
The structural material of the blanket is not directly exposed to the plasma. The first wall typically made out of tungsten or other high-temperature withstanding materials are attached to the plasma-facing side. Still the blanket has to cope with temperatures of 500–950°C depending on the reactor design. Few materials can cope with such extreme heat and neutron bombardment for a long time. Moreover, there can be corrosion from the coolant or the breeding material used inside the blanket.
“the blanket has to cope with temperatures of 500–950°C”
One possibility currently preferred by several ITER blanket proposals is Reduced Activation Ferritic/Martensitic (RAFM) steels. Some flavours are the European EUROFER97, the Japanese F82H and the Chinese CLAM. These steels undergo heat treatment to have exceptional toughness and strength. The high activation elements (Mo, Nb, Ni, Cu and N) are replaced by low activation elements (e.g. W, V and Ta). Blankets with this type of material are limited to around 500°C.
Somtimes vandium-alloy is proposed because of its compatibility with lithium. However, it’s unclear whether large scale components could be manufactured with V-alloys.
A more costly alternative is silicon carbide fibre composites (SiCf/SiC). These materials are nowadays used in the space industry and for jet engines and have extremely good attributes as structural material for the blanket. The main advantage is the heat capabilities (usable up to about 1200°C) and its low degradation under the neutron bombardment. Another advantage is that its density is three times lower than that of RAFM steel. If you remember that ITER (23,000 t) weighs more than twice the Eiffel tower you quickly see why a blanket on the order of 10,000 t is unwieldy. A commercially available fibre is e.g. Tyranno Fiber.
A general concern when liquid metals are used as the flowing material inside is magnetohydrodynamic (MHD) effects created by the fusion reactor magnets. The magnetic field strength is larger than 10 T (10 Tesla is more than 2000 times stronger than your average refrigerator magnet). ITER’s central superconducting solenoid delivers 13 T but with HTS magnets we will see larger field strengths soon (MIT showed off a 20 T magnet for their next fusion reactor design). In other words, one has to design the blanket in a way that the magnet allows the liquid to move without problems, e.g. drops in the pressure at certain points that hinder the flow.
Coolants and Heat Exchange Cycle
Several coolants are currently being considered. The primary goal is to bring the converted kinetic energy of the neutrons via the coolant to the power-producing turbine. At the same time, the immense power output creating a heat flux on the order of 3 MW/m² can damage the structural material.
We can build the blanket either with one cycle or with two coolants (dual coolant) separating the functions of fuel breeding from the coolant. Both ways have their own challenges and mainly depend on whether the coolant can contain lithium needed for the breeding.
Materials like lithium-lead eutectic (LiPb, though sometimes “PbLi” is used, 17 at% Li and 83 at% Pb) (at% refers to the atomic ratio) can fulfil both functions. Eutectic here means that the melting point of the LiPb is low (235°C) and in a local minium as a function of the mixing ratio. Blankets designed in this way are called self-cooled blankets. One such design is the SCYLLA© blanket by Kyoto Fusioneering.
Other choices are molten salts, e.g. fluoride-based FLiBe. FLiBe has good breeding qualities because of the use of beryllium and high lithium content but is expensive due to the beryllium use and the technology surrounding (such as redox control and tritium extraction) it is at low technical readiness as of now.
For dual-coolant designs, helium is often used for the coolant function. The helium in the system has to be kept at high pressure (around 8 Mpa or roughly the pressure used in a high pressure washer). If a pressurised water system is utilised instead, the steam could be operated at 15 Mpa.
A general problem is that the bred tritium does not stay where it’s supposed to stay and will permeate through all materials. It is slightly radioactive and it’s not a good idea to mix those with your coolants.
Fuel Breeding Blankets
ITER’s DEMO reactor will need about 300 g of tritium per day to produce 800 MW of power. Tritium has no natural abundance and only a few sources produce it at a very steep price. Tritium is produced as a by-product of nuclear fission reactors or with accelerators via spallation. CANDU-type nuclear reactors produce globally around 50 g worth of tritium. This type of nuclear reactors will likely phase out over the coming decades and with a half-time of tritium of 12.3 years it becomes evident that our stash of tritium will shrink very quickly once we get fusion reactors online. What to do? Yes you got it, we have to breed our fuel within the fusion reactor itself.
“We have to breed our fuel within the fusion reactor itself.”
Lithium is abundant and fairly cheap. If neutrons hit lithium, fuel for our reactor is created. Deuterium is anyway available in sea water so we don’t need to worry about that. A fusion reactor creates lots of 14.1 MeV neutrons so the best way is to make use of them in the reactor walls.
The most important quantity is the tritium breeding ratio:
As long as it’s larger than one, the reactor breeds more tritium than it uses up. However, it’s expected that values TBR>1.1 are needed to build a storage and combat leakage.
The two stable isotopes of lithium, ⁶Li and ⁷Li, both produce tritium. However, the cross section (probability that the neutron interacts with the lithium) is far larger for ⁶Li. Unfortunately, ⁶Li is the in nature less abundant isotope. Consequently, to increase the TBR, one often has to enrich the breeding material with ⁶Li. Depending on the blanket type we want as much as 40% of the good ⁶Li.
The energy balance of the two tritium production reaction, i.e. Li-6(n,α)t and Li-7(n,n’α)t, for both lithium isotopes looks like this:
Interestingly enough, the neutron of the latter interaction can further interact with the lithium but it doesn’t give us a better TBR than ⁶Li. The reason is that ⁷Li is a threshold reaction and requires a minimum neutron energy of 2.8 MeV whereas ⁶Li can make use of the high cross section of lower energy neutrons:
A good way to achieve a high TBR is to increase the number of neutrons by adding a multiplier, e.g., W Tungsten plates, in which a neutron showers and creates multiple lower energy neutrons. Lastly, designs exist that use neutron reflectors attached to the far side made out of, e.g., graphite.
For the fuel breeding as a separate cycle we could use lithium in the form of a pure molten lithium, a molten salt (e.g. FLiBe), or the previously mentioned eutectic LiPb. In the latter case, LiPb would flow very slowly at a velocity of less than 5 cm/s and be recirculated about 10 times/day to move the tritium outside of the vessel where it can be extracted.
Extraction is fairly straightforward and has been tested already with several systems with tritium or sometimes with the non-radioactive deuterium for easier handling. We will have a separate article on tritium extraction so I will only show one such example on the left. The vacuum sieve tray pumps the LiPb fluid containing the tritium through tiny nozzles through the vacuum chamber. Via diffusion the tritium is released from the LiPb and can be pumped out. This works because tritium has a low solubility in LiPb. Problems occur due to the materials used. For example corrosion at the nozzles. Other extraction devices include gas-liquid contactors (GLC) or permeation in vacuum.
Maintenance
In fact it’s still very much unclear how to run the maintenance. Of course no breeding blanket concept has been realised in a reactor yet and we can only estimate what needs to happen. We know though that the structural material can be subject to corrosion from breeding carrier/coolant, that the structural integrity is destroyed by heat and neutrons and that impurities in the materials can create radioactive isotopes that need to be treated and stored.
All in all I don’t really have an answer how often the blanket needs to be replaced. Presumably every few years. The only good news I can offer here is that with designs using LiPb carriers you can just pump more lithium into the fluid outside of the reactor. So at least we don’t have to worry about that.
Conclusion
“There are many good and valid designs for breeding blankets on the market but none is being used in a reactor just yet.”
There are many good and valid designs for breeding blankets on the market but none is being used in a reactor just yet. We have seen that there are still many things that need clarification, measurements and more data to verify that the proposed concepts hold up in commercial fusion reactors. It is clear though that the best design always depends on the other components of the reactor (magnetic field, plasma shape, heat load, accessibility and others).
