Fusion Illusions
Michael Dittmar
Summary and citation · read the original at the source · none found
In one page
Michael Dittmar, a particle physicist at ETH Zürich who works at CERN, wrote the toughest audit anyone had then written of the deuterium-tritium fusion roadmap. His argument is not about plasma physics, which he grants is the part the field understands best. It is about everything wrapped around the plasma. A power-station-scale reactor would have to burn roughly fifty-six kilograms of tritium a year, a substance that does not occur naturally, decays with a twelve-year half-life, and in 2008 cost about thirty million dollars a kilogram. So the machine has to breed its own from a lithium blanket, and Dittmar works through the arithmetic of how much it must breed and what the simulations then delivered. He adds the first-wall materials problem, where fourteen-million-electronvolt neutrons displace every atom hundreds of times over five years, and notes that no facility could test a candidate material at that flux. His conclusion is that ITER is superb plasma physics and should be funded as such.
Why it matters hereChapter 12 argues that fusion is the energy substrate the rest of the thesis rests on, and this chapter is the sharpest statement of what the mainstream deuterium-tritium route has to deliver before that is true — a fuel cycle that closes, and a wall that survives. Chapter 1’s evidence ladder gains a worked example of how to score a large programme by its unsolved engineering rather than by its press releases.
What it claims
01Commercial fusion faces four barriers, not one: steady-state operation at power-station scale, a first wall that survives years of fourteen-million-electronvolt neutrons, safe handling of kilogram quantities of tritium, and a self-sufficient tritium breeding cycle. Only the first sits inside the scope of the ITER experiment.Section 5.1, the four numbered barriers, arXiv 0911.2628v1
What to watch02The fuel arithmetic is the heart of the argument. A one-gigawatt-thermal reactor burns about 55.6 kilograms of tritium a year. JET’s record pulses burned about three micrograms drawn from a twenty-gram inventory, a fractional burn-up near one part in a million, and ITER’s planned four-hundred-second pulse at half a gigawatt burns 0.035 grams.Section 5.1 barrier 1, and section 5.2, arXiv 0911.2628v1
Published and peer-reviewed03Tritium supply is finite and shrinking. The world’s civil stock comes from Canadian heavy-water reactors at roughly thirty million dollars a kilogram, was expected to peak near twenty-seven kilograms around 2025, and then decays at more than five per cent a year on a 12.3-year half-life — so a prototype starting in 2050 might have about seven kilograms to open with. Any machine after ITER must therefore breed more tritium than it burns.Section 5.2, arXiv 0911.2628v1
What to watch04On the breeding ratio, Dittmar follows Sawan and Abdou’s 2005 calculations: the minimum required ratio is about 1.15 under the most favourable assumptions — fractional burn-up above five per cent, reserve time under five days, doubling time above four years — and rises towards 1.5 when a one-year doubling time is demanded, while the best ratio the blanket simulations then achieved was itself about 1.15, and the few real experiments came in consistently about fifteen per cent below what the models predicted.Section 5.2, discussion of the required and achievable tritium breeding ratios, arXiv 0911.2628v1
What to watch05The materials requirement is quantified. At full scale the neutron flux would be ten to twenty times that of a modern fission plant and at higher energy, displacing each atom of the surrounding solid roughly 475 times over five years, with erosion of about three millimetres per burn year for carbon-like materials and about a tenth of a millimetre for tungsten — and no existing or then-planned facility could expose a candidate material to that flux for a test.Section 5.1 barrier 2, arXiv 0911.2628v1
Published and peer-reviewed06Dittmar’s recommendation is a reclassification rather than a shutdown: ITER is an experiment into fundamental plasma physics and should be funded on the same terms as other research, with the chamber, blanket and fuel-cycle questions raised to equal priority — a point he takes from Mohamed Abdou’s 2003 briefing to the United States Department of Energy, which called tritium supply and self-sufficiency a go or no-go issue as critical as demonstrating a burning plasma.Section 5.3, Ending the dreams about controlled nuclear fusion, arXiv 0911.2628v1
What to watch
The way in
https://doi.org/10.2307/j.ctt18fs7bj.25LICENCE. Published as a chapter of The Final Energy Crisis, second edition, Pluto Press, 2008, by Michael Dittmar of the Institute of Particle Physics at ETH Zürich, who works at CERN. The chapter is behind the publisher’s copyright on JSTOR, Crossref carries no licence at all for it, and Unpaywall and OpenAlex both report it closed, so no text is reproduced here. TEXT. The summary and the claims below are the site’s own. They were written from the author’s own condensed version of the same material, which is section 5 of arXiv:0911.2628 version 1, The Future of Nuclear Energy: Facts and Fiction, Chapter IV, posted 13 November 2009 and downloaded and read in full on 2026-09-08. Dittmar states there, at the head of that section, that it is a short version of his detailed article in the second edition of The Final Energy Crisis, and the section carries the same title, Fusion Illusions. That preprint is distributed under the arXiv non-exclusive licence, which is not a Creative Commons grant, so it too is linked rather than quoted, and every locator names a section of the preprint rather than a page of the book chapter. CONTEXT. The chapter is a 2008 audit of the deuterium-tritium roadmap and of the ITER project as it was then scoped. The site reports it faithfully and treats each of its barriers as a live engineering target with a named measurement attached; the closing note below records where the field has moved since.
How to cite it
Michael Dittmar (2008) Fusion Illusions. doi:10.2307/j.ctt18fs7bj.25
Where it sits in the curriculum