Fuel Cycle System

A fusion reactor does not use its fuel just once and throw it away. Only a few percent of the fuel injected into the plasma actually burns; most of it comes back out as unburned fuel. Recovering this unburned fuel, cleaning it, and feeding it back in: this circulation scheme is the fuel cycle system. On this page we work step by step from the intuition of circulating and reusing fuel to the research frontier of procuring tritium and achieving self-sufficiency.

Start with Intuition (High School)

The fuel of a fusion reactor comes in two kinds, both relatives of hydrogen: deuterium (symbol D) and tritium (symbol T). When these two collide and fuse, they produce a large amount of energy, helium, and a neutron. This is the D-T reaction.

What matters here is that not all of the injected fuel burns. As an image, it is like adding firewood to a campfire: it does not all turn to ash at once, and some unburned logs are left over. In a fusion reactor, only a few percent of the fuel actually reacts in the plasma, and the remaining 90 percent and more leaves the reactor without burning.

If we threw away the unburned fuel, it would be far too wasteful, and there simply would not be enough fuel to begin with. So we sort out just the fuel from the exhaust gas that comes out and return it to the reactor. We keep the fuel going round and round: inject it, burn it, recover it, inject it again. This is what the term fuel cycle means.

Another intuitive problem is that tritium barely exists in this world. Deuterium is abundant in seawater, and you can extract as much of it as you like just by scooping up seawater. Tritium, however, is radioactive and halves in about 12.3 years, so whatever formed naturally long ago has long since disappeared. Across the entire Earth there are only a few kilograms. In other words, one of the two fuels is effectively impossible to “buy.”

So what do we do? In a fusion reactor, we use the neutrons that fly out of the reaction to make tritium ourselves inside the reactor wall (the blanket). The idea is to make fuel while burning it at the same time, in effect achieving fuel self-sufficiency. Whether this self-sufficiency can be sustained is a major key to whether a fusion reactor becomes practical.

Understand the Physics (Undergraduate)

The D-T reaction can be written as follows.

$$\mathrm{D} + \mathrm{T} \rightarrow {}^{4}\mathrm{He}\,(3.5\ \mathrm{MeV}) + n\,(14.1\ \mathrm{MeV})$$

A single reaction releases a total of $17.6\ \mathrm{MeV}$ of energy. Of this, the helium (alpha particle) carries $3.5\ \mathrm{MeV}$ and the neutron carries $14.1\ \mathrm{MeV}$. To get a sense of the scale of the energy, consider the fuel consumption: in a reactor whose fusion power (not electrical power) is $1\ \mathrm{GW}$, the tritium consumed is roughly 56 kg/year. Given that the entire Earth’s stock is only a few kilograms, you can see how essential in-reactor production is.

The fuel cycle system is broadly divided into an inner cycle and an outer cycle. The inner cycle is the fast loop that runs from injecting the fuel, exhausting the unburned fuel, purifying it, and injecting it again, with a residence time of a few to a few tens of minutes. The outer cycle is the loop that extracts the tritium bred in the blanket, purifies it, and sends it to storage, and this takes from several hours to several days.

There are two main methods of putting fuel into the plasma (fueling).

Gas puffing blows fuel gas directly in from near the wall of the vacuum vessel. The mechanism is simple and the response is fast, but the gas is almost entirely ionized at the outer edge of the plasma and struggles to reach the center. Its fueling efficiency generally stays around a few tens of percent.

Pellet injection cools the fuel and freezes it into ice grains (pellets) a few millimeters in diameter, then fires them in at several hundred meters per second up to more than 1 km per second. Because the solid grains fly deep in while gradually melting from the surface, they can deliver fuel close to the center of the plasma. Its advantages are high fueling efficiency and easier control over where the fuel is placed.

The burn-up fraction is the proportion of the injected fuel that actually reacted. In current designs it is expected to be on the order of a few percent. Conversely, since most of the injected fuel is exhausted without burning, the throughput needed to recover and reuse it determines the scale of the fuel cycle system.

Deepen the Theory (Graduate)

The exhausted unburned fuel contains, besides the fuel D and T, helium produced by the reaction (the ash) and impurities coming off the walls. From this we need to extract just the fuel, and moreover separate it by isotope.

First, in the stage of roughly sorting out the hydrogen isotopes, methods such as permeating only hydrogen through a palladium-alloy membrane are used to obtain a high-purity hydrogen isotope stream. The problem lies beyond that: D and T, which behave almost identically chemically, along with light hydrogen (protium, H), must be separated from one another. Here we exploit the tiny difference of their mass.

A representative method is cryogenic distillation. Because the hydrogen isotopes have very slightly different boiling points, cooling them to the liquid-hydrogen temperature range (around $20\ \mathrm{K}$) and running them through a distillation column makes lighter components evaporate more readily while heavier components stay in the liquid. By stacking this difference over many stages inside the column, H, D, and T are separated. Because the difference is small, many stages and much time are needed, the equipment is large, and the amount of tritium held inside (the inventory) is not negligible. When separation with a fast response at a small scale is desired, another method such as TCAP, which uses temperature-swing adsorption, is also used in combination.

The core of the outer cycle is the tritium plant, which handles all of these together. It is responsible for recovering the exhaust, removing impurities, isotope separation, storing the tritium, and even recovering tritium that has been taken up into water and the like (water detritiation). Storage uses metal hydrides: for example, an alloy such as ZrCo absorbs and fixes the hydrogen isotopes, and heating it releases them for extraction. This is safer and easier to handle than storing the gas at high pressure. However, since tritium itself undergoes beta decay, turning into helium-3 at about $5.5\ \%$ per year, degradation of the storage material and management of the decay products become challenges.

The ITER fuel cycle is the first major attempt to make this whole sequence of processing work at machine scale. Fuel is supplied by gas puffing and pellet injection, exhausted by cryopumps (pumps that draw gas down by freezing it onto cryogenic surfaces), and separated and purified in the tritium plant before being returned. Because helium has an extremely low boiling point of $4.2\ \mathrm{K}$ and is hard to freeze out, adsorbents such as activated charcoal are combined to exhaust it. As a key safety measure, an upper limit is placed on the amount of tritium present inside the vacuum vessel. How much tritium is taken up into the walls (retention) and how much can be recovered are studied in detail as quantities that bear directly on whether the whole cycle can be sustained.

Research Frontier (PhD)

The biggest research theme is whether tritium self-sufficiency can truly be achieved. Because a fusion reactor cannot buy tritium from outside, it must keep breeding more than it consumes inside the reactor, or operation will stop. This balance is expressed by the tritium breeding ratio (TBR), the ratio of the amount bred inside the reactor to one unit consumed. In principle $\mathrm{TBR} > 1$ is required, but in practice an effective value somewhat above $1$ is needed, accounting for losses in the fuel cycle system, decay, storage, and even the surplus needed to start up new reactors.

What governs the TBR is the neutron economy: whether one or more tritons can reliably be produced from a single neutron. Neutron multiplication by beryllium or lead, enrichment of lithium-6, the coverage fraction of the blanket, and how much the neutrons are eaten up by structural materials and piping all come into play. This is a region where the fuel cycle and the blanket are inextricably linked. The physics of breeding is treated in detail on the Blanket page.

Another focus is the start-up inventory. To bring a reactor online and start circulating tritium on its own, a substantial amount of tritium must first be prepared from outside. The main current supply source is recovery from CANDU-type reactors that use heavy water as a moderator, and the total amount is limited. For this reason, how small the start-up inventory can be made, and how short the tritium doubling time (the time until surplus tritium can start up the next reactor) can be made, are studied as problems that govern the deployment of multiple reactors. These are discussed together with designs that squeeze the inventory of the whole fuel cycle as tightly as possible.

The low burn-up fraction also feeds directly into the cycle. A burn-up fraction of a few percent means that many tens of times the consumed amount of fuel must be circulated every day, which increases the amount of tritium held in the system and raises the burden of loss and exposure management accordingly. Efforts to raise the burn-up fraction, or to reduce the inventory and processing time of the cycle, are pursued from the standpoints of both safety and economics. For the safe handling and confinement of tritium, see Tritium Management, and for the physics of the D-T reaction itself, see Fusion Reactions.

The place where these challenges are integrated and demonstrated is the prototype reactor (DEMO) that follows ITER. Whether it can simultaneously satisfy continuous operation, machine-scale throughput, demonstration of $\mathrm{TBR} > 1$, and self-sufficiency with a small start-up inventory will be the dividing line for whether fusion can be made to work as a power source.

Check Your Understanding

Q1. Why does a fusion reactor circulate and reuse the fuel it injects?

A Because only a few percent of the fuel actually reacts in the plasma, while most of it is exhausted as unburned fuel B Because circulating the fuel naturally raises the plasma temperature C Because circulating it chemically transforms the fuel into a different substance D Because injecting a large amount of fuel at once would make the reactor explode

Q2. Deuterium and tritium differ greatly in how easily they can be procured. Which is the correct reason?

A Deuterium can only be made artificially, while tritium can be mined from ore B Both can be extracted equally from seawater, and only the purification cost differs C Deuterium is stably present in seawater, but tritium is radioactive with a short half-life and barely remains in nature D Tritium is a stable isotope that keeps well, while deuterium decays quickly

Q3. When gas puffing and pellet injection are compared by fueling efficiency and how easily they reach the plasma center, which is correct?

A Gas puffing reaches the center easily and has high fueling efficiency, while pellet injection can only place fuel at the outer edge B Gas puffing ionizes at the outer edge and struggles to reach the center with low fueling efficiency, while pellet injection reaches near the center with high fueling efficiency C Both have the same fueling efficiency, and the only difference is the size of the equipment D Pellet injection is only faster in response, and its ability to reach the center is no different from gas puffing

Q4. Why can cryogenic distillation be used to separate hydrogen isotopes?

A Because H, D, and T have different charges and can be pulled apart by an electric field B Because H, D, and T differ greatly in chemical reactivity and can be sorted with chemicals C Because H, D, and T have different colors and can be told apart optically D Because H, D, and T have very slightly different boiling points, so when cooled to the liquid-hydrogen temperature range the lighter components evaporate more readily

Q5. Why is it considered insufficient for the tritium breeding ratio (TBR) to be only slightly above 1?

A Because in practice there are losses and radioactive decay in the fuel cycle system, and a start-up inventory for new reactors must also be secured B Because the TBR cannot be measured, so a larger value is just assumed for safety C Because a TBR above 1 breeds too much tritium and becomes dangerous D Because the TBR changes with temperature, so its average always falls below 1

Related Topics

• Blanket: Covers the breeding mechanism that produces tritium from neutrons, and the neutron economy that determines the TBR.
• Tritium Management: Covers the confinement of tritium, exposure management, and safety measures as a facility.
• Fusion Reactions: Covers the physics and energy balance of fusion reactions, including the D-T reaction.