Skip to content

Vacuum Vessel

The vacuum vessel is the large, innermost container that holds the fusion plasma. The magnetic field that confines the plasma can be created without a vacuum vessel, but to actually burn a plasma you must first make the inside of the vessel thoroughly free of air. On this page we will build up an understanding, step by step, of why ultra-high vacuum is needed, how that vacuum is created and maintained, and what structural demands are placed on the vessel itself.

In a single phrase, the vacuum vessel is the “ultimate clean, empty room” for holding the plasma.

Why does it need to be empty? To cause fusion, we heat the fuel, deuterium and tritium, to about 100 million degrees. But if air remains inside the vessel, the molecules of that air collide with the plasma and steal its heat, so the plasma you worked so hard to heat cools down right away. Fusion does not happen in a cold plasma. So the first thing we do is drive out the air, the troublemaker.

Let’s get a feel for how empty it needs to be using everyday numbers. The air we breathe every day contains, in one cubic centimeter (about the volume of a sugar cube), roughly 25 quintillion molecules. In a fusion reactor’s vacuum vessel, we reduce this to less than one ten-billionth of that. As vacuums go that we can create on the ground, it is a state nearly as thin as outer space.

Another important thing is “impurities.” If even a tiny amount of heavy atoms other than the fuel (for example, metal atoms flaked off the vessel wall) mix into the plasma, those heavy atoms radiate light strongly and throw away heat. Just as dropping a single pinch of salt into clean water clouds the whole thing, a small amount of impurity cools the entire plasma. So the vacuum vessel is not just emptied of air, but thoroughly “cleaned” so that no debris comes off the walls. This cleaning, done by heating and by scrubbing with electrical discharges, is baking and wall conditioning.

Furthermore, the vacuum vessel is not only the container that holds the plasma but also a “safety wall” that keeps radioactive material from leaking out. Radiation flies around inside, and the fuel tritium is radioactive, so the vessel is designed as a sturdy metal shell that reliably confines these.

The quantity that expresses how good or bad a vacuum is is pressure. The target base pressure aimed for in a vacuum vessel is roughly 105 Pa10^{-5}\ \mathrm{Pa} or lower, which is the regime called ultra-high vacuum. Since atmospheric pressure is about 105 Pa10^{5}\ \mathrm{Pa}, the target is a thinness of one ten-billionth of that.

The pressure pp is related to the gas temperature TT and the number of molecules per unit volume (number density) nn by the following relation.

p=nkBTp = n k_{\mathrm{B}} T

Here kBk_{\mathrm{B}} is the Boltzmann constant (1.38×1023 J/K1.38 \times 10^{-23}\ \mathrm{J/K}). This equation shows that pressure is proportional to “how densely the molecules are packed.” Substituting p=105 Pap = 10^{-5}\ \mathrm{Pa} at room temperature (T300 KT \approx 300\ \mathrm{K}), the number density comes out to about n2.4×1015 m3n \approx 2.4 \times 10^{15}\ \mathrm{m^{-3}}, that is, down to about 2.4 billion per cubic centimeter. The earlier intuition of one ten-billionth of the atmosphere is quantitatively confirmed from this equation.

Why impurities are so disliked can also be explained by physics. Every time an impurity ion in the plasma collides with an electron, it emits light (radiation) and loses energy. The power of this radiation loss depends strongly on the impurity’s charge number ZZ, roughly as the square of ZZ or more. In other words, heavy elements such as iron (Z=26Z = 26) or tungsten (Z=74Z = 74) radiate orders of magnitude more strongly than hydrogen (Z=1Z = 1), and cool the plasma with only a trace mixed in. This is exactly why ultra-high vacuum and wall conditioning, which suppress the intrusion of heavy elements from the walls, are indispensable. This radiation loss is also a central issue in the selection of plasma-facing materials (see First Wall for details).

The main player that creates ultra-high vacuum is the vacuum pumping system. First a roughing pump brings the pressure down from atmospheric to medium vacuum, and from there a high-vacuum pump takes over. The two types of high-vacuum pumps commonly used in fusion reactors are as follows.

The turbomolecular pump uses rotor blades spinning at tens of thousands of revolutions per minute to mechanically flick gas molecules away and pump them out. It works efficiently in the molecular-flow regime, where the mean free path of the molecules is longer than the vessel size, and it can pump a wide range of gases stably.

The cryopump captures gas molecules by freezing them onto a surface cooled to cryogenic temperatures (about a few K to 20 K), a pump that, so to speak, “reduces the air by frosting up a cold surface.” Its pumping speed is extremely large, making it suitable for the main pumping of fusion reactors. ITER also adopts large-capacity cryopumps. However, the captured gas must eventually be regenerated (warmed and released), and in a fusion reactor that handles the fuel tritium, an operational design that links this regeneration with fuel recovery becomes important.

The essence of what makes achieving ultra-high vacuum difficult lies not so much in the gas leaking in from outside (leaks), but rather in the gas coming out from the inner surfaces of the vessel. The base pressure is determined by the balance between the effective pumping speed SS and the total flow of gas release QQ, and the equilibrium pressure is given roughly by

p=QSp = \frac{Q}{S}

This equation shows that no matter how large you make the pump to raise SS, if QQ is large the vacuum will plateau. In a vessel with a vast inner surface area, as in a fusion reactor, the main cause of QQ is gas release from the surface at room temperature (outgassing), especially water molecules adsorbed on the metal surface and diffusive release of hydrogen dissolved inside the material.

This is where baking comes in. When the entire vessel is heated to around 200 degrees, the desorption of water molecules adsorbed on the surface accelerates exponentially, so a large amount can be released and fully pumped out while it is heated. After cooling, the surface is clean, so the outgassing flow QQ drops greatly, and only then can ultra-high vacuum be reached. Lowering the base pressure is engineering that not only raises SS but also lowers QQ; that is the key insight of this field.

Even after baking has removed the moisture, wall conditioning is needed to create a wall state suitable for plasma operation. A representative technique is glow discharge cleaning, which strikes a weak discharge to lightly tap the walls, knocking off adsorbed impurities and pumping them out. Furthermore, boronization forms a thin film of boron on the wall surface. Boron has the function of trapping oxygen, which suppresses oxygen intrusion into the plasma and makes it easier to control hydrogen recycling from the walls. These are operational techniques routinely performed in magnetic confinement devices, and they govern plasma performance.

From the viewpoint of structural design, the vacuum vessel must simultaneously satisfy several conflicting requirements. First, strength as a pressure vessel that withstands the pressure difference between inside and outside (atmospheric pressure outside, vacuum inside). Second, resistance to the enormous electromagnetic forces that arise during a disruption, the phenomenon in which the plasma current suddenly vanishes. During a disruption, eddy currents are induced in the metal walls of the vessel, and they interfere with the strong magnetic field to produce large forces. In addition, halo currents that flow from the plasma into the walls are also a cause of electromagnetic force. These loads are borne by increasing the rigidity of the vessel. ITER’s vacuum vessel adopts a double-wall structure with an inner wall and an outer wall precisely in order to connect the walls with ribs and achieve both rigidity and shielding performance. Third, it serves as radiation shielding that protects the superconducting coils from neutrons and gamma rays, and shielding material fills the space between the walls. For the coils on the protected side, see Superconducting Coils.

Furthermore, as a safety boundary that confines radioactive material, the vacuum vessel is required to have seismic design that maintains its integrity against external events such as earthquakes. The structural material uses nitrogen-added stainless steel that achieves both strength and low activation (at ITER, SUS316L(N)-IG), and for future power reactors the application of reduced-activation ferritic/martensitic steel (RAFM) is being studied.

The vacuum vessel may look like a mature engineering item, but it holds many unsolved challenges on the road to realizing fusion power.

ITER’s vacuum vessel is about 11 m tall, has a major radius of about 6 m, and a mass of about 5,000 tonnes for the body itself, reaching the 8,000-tonne class including shielding material and port structures, making it the largest fusion vacuum vessel ever built. Manufacturing such a large vessel in several sectors and welding them together on site into a single torus is itself a major engineering challenge. To suppress welding deformation while meeting the strict leak-rate requirements that guarantee ultra-high vacuum, welding technology and precise dimensional control are subjects of research. The overall standing of ITER is summarized in ITER Project.

One of the greatest research challenges is remote maintenance. After a fusion reactor has continued operating, the inside of the vessel is activated by neutrons and reaches a high radiation dose that people cannot enter. Therefore, all maintenance work, such as replacing, inspecting, and repairing the blanket and divertor, must be carried out by remotely operated robots. Developing remote-handling equipment that accurately handles heavy objects inside the narrow, complex torus, ensuring the reliability of the work, and establishing replacement procedures that avoid radiation exposure are being pursued as important themes that govern a power reactor’s availability.

On the vacuum-pumping side too, how to recover and recirculate the fuel tritium after capturing it with a cryopump, and how to integrate large pumping capacity with the fuel cycle, are challenges. Also, since power reactors are exposed to a higher neutron environment over long periods, material selection that withstands the changes in properties due to irradiation of structural materials, and vessel design premised on it, are being studied. When reading papers, English keywords such as vacuum vessel, ultra-high vacuum, outgassing, baking, glow discharge cleaning, boronization, disruption, halo current, eddy current, and remote maintenance appear frequently.

Q1. Why must the air be pumped out of a fusion reactor's vacuum vessel first to reach ultra-high vacuum?
Q2. Using the pressure equation p = n kB T, how can you explain the meaning of ultra-high vacuum (10 to the minus 5 Pa)?
Q3. When the base pressure is determined by p = Q/S, which quantity does baking mainly act on?
Q4. What is boronization done for?
Q5. Why does the vacuum vessel need remote maintenance?