“The dream of clean energy is closer”.

“Nuclear fusion takes a big step forward”.

Those are just two recent headlines concerning the latest developments in nuclear fusion. There is little doubt that progress has been made in recent years with new results at international facilities and the ITER nuclear fusion experimental reactor in Cadarache, France, inching closer following years of delays and cost overruns. The past few years have also seen significant private investment with countless start-up companies looking to commercialize fusion in the coming decade.

With these developments, one might think that fusion is about to finally deliver clean energy following decades of promise. Media articles certainly give this angle yet little attention is given to the problems that commercial fusion devices might face when scaling up. Nuclear fusion reactors will not only need to fuse the hydrogen isotopes deuterium and tritium (D-T) to produce power but also be fuel self-sufficient.

Obtaining enough deuterium, which occurs naturally, is not a problem, but tritium only exists in trace amounts with limited stocks available as a by-product from fission reactors. The 14 MeV neutrons produced in the D-T reaction must therefore not only produce useful heat but also interact with lithium “blankets” to produce tritium. This tritium will then be recycled to provide later input into the reactor with any excess stored for a subsequent reactor. In other words, reactors will need to have a tritium breeding ratio (BR) greater than one.

Natural lithium has two naturally-occurring stable isotopes: lithium-6 at 7.6% abundance and lithium-7 at 92.4% abundance. The different interaction of the two lithium isotopes with neutrons to produce tritium is well known with lithium-6 being by far the more important in terms of producing tritium. Although there are suggestions that natural lithium can provide a sufficient breeding ratio, others say that it is essential to isotopically enrich the lithium-6 content.

Ian Chapman, chief executive of the UK Atomic Energy Authority (UKAEA), has suggested that the lithium-6 to lithium-7 ratio would need to be more like 40/60. This raises two questions: how might this be done and how much lithium is required? Chapman states that a commercial enrichment method would need to be set up and the UKAEA is now exploring several possible routes. (Although one method already exists, the COLEX process, it uses large quantities of mercury and has been banned since 1963 in the US).

With regard to the amount of lithium required, this can be roughly calculated for a tokamak using the requirement that a thickness of about 1 m of lithium is needed to slow and stop the fast neutrons. The exact amount of lithium will, of course, depend on the exact blanket structure but it must be sufficient to ensure that as many neutrons as possible interact with it to breed tritium.

For a simple spherical reactor with a chamber diameter of 1 m, a 1 m thick surrounding shell of lithium could weigh about 7.5 metric tonnes – and more if the machine were bigger. To obtain a 40/60 ratio of lithium-6 to lithium-7, a starting quantity of some 40 tonnes of natural lithium would need to be enriched, a significant target for whatever process emerges.

There have been several suggestions that the breeding material could either be pure lithium, lithium-lead or ceramic lithium compounds. Whatever the exact breeding material, the efficient retrieval of the tritium in the light of its fast diffusion in the hot reactor environment will be crucial.

Might ITER use the total available inventory of tritium worldwide in their experiments?

An important reactor issue at start-up will be the quantity of tritium available to fuel the reactor until the breeding system allows the reactor to become self-sufficient. One would need to know how quickly the tritium can be recycled and how much would be needed as a back-up in case problems arise. The amount of tritium being trapped in the fusion reactor also needs to be taken into account. Clearly a reactor is unlikely to be built unless there is certainty that it will not run out of the essential tritium, both at the start and in the breeding cycle.

A key factor in tritium discussions must be its availability, and this is often quoted to be between 20–30 kg extracted from heavy water reactors such as the CANDU reactors in Canada. Of interest here is that ITER stated in March 2023 that it will use the total available inventory of tritium worldwide in their experiments. Even if this claim is exaggerated, it is not clear how the owners of tritium stocks will distribute these to potential customers. A recent agreement on tritium collaboration between the Canadian Nuclear Laboratories and the UKAEA could benefit the Spherical Tokamak for Energy Production – a demonstration nuclear-fusion power plant – that is planned to switch on in the UK in the 2040s.

Read more

Fusion industry outlines ambitious plans to deliver electricity to the grid by 2035

Much is also made about how fusion is cleaner than fission. This is true, but both the use of large amounts of reactive lithium and the creation of radiologically unpleasant tritium means that it can’t be compared to greener options such as renewables. Although the radioactivity of tritium decays with a relatively low energy 5.7 keV beta particle, its rapid interchange with hydrogen to give tritiated water (HTO) could contaminate ground water, which can then enter the food chain. This has led to very strict limits for the release of HTO from heavy water reactors and other nuclear installations. One assumes similar limits would equally apply to a fusion reactor.

The fusion community still has many issues to iron out before it can become a viable energy source. Fusion has already been studied for over six decades and the following two will see whether the fusion dream can put energy on the grid economically, or whether it will remain just that – a dream.