Stellarator is Germany’s Devilishly Complex Nuclear Fusion

You may not have heard of a Stellarator before, but if all goes well later this month in a small university town in the far northeast of Germany, you will. That’s because the Wendelstein 7-X is finally going to be fired up. If it’s able to hold the heat of a fusion-capable plasma, it could be a huge breakthrough.

So what’s a stellarator? It’s a specific type of nuclear fusion containment geometry that, while devilishly complex to build and maintain, stands a chance at being the first fusion generator to achieve break-even, where the energy extracted from the fusion reaction is greater or equal to the energy used in creating the necessary hot plasma.

There’s an awesome video on the W7-X, and some of the theory behind the reactor just below the break.

Most fusion reactors are tokamaks, which are doughnut-shaped reactors wrapped in superconducting magnets that contain the plasma inside the loop. The problem with tokamaks is that they don’t have an intrinsically uniform magnetic field inside because, put simply, the coils on the hole of the doughnut are closer together than those on the outside. This makes the plasma want to wander outward and hit the walls, which isn’t good.

Tokamak_(scheme)To compensate, a tokamak induces a circular current within the plasma as well, which pulls it inwards. The problem with this is the heat in the parts that induce the current as well as instabilities in the current as it runs through the plasma itself. This means that tokamaks can’t run for very long: we’re talking a few minutes at most. This makes it very hard for tokamaks to produce more energy than it takes to fire up the plasma in the first place. (Tokamak schematic courtesy Max Planck Institute.)

Stellarators get around this by twisting the path of the plasma. The original stellarator design, invented by all-around physics-badass [Lyman Spitzer], twisted a loop into a figure-8 shape. The point is that the plasma that’s on the inside of the track in one half of the eight is on the outside in the other half, and there’s no need for the tokamak’s toroidal plasma current. The W7-X twists the plasma band five times as it moves around what’s essentially a circle.

And this means that there’s a chance of containing a plasma for a much longer time in a stellarator, and eventually breaking even on the startup energy. The only problem is that making one is devilishly difficult, even with the benefit of a billion Euros and the best German engineering. Just looking at the number of access ports on the W7-X gives you a bit of insight into how incredibly complex this machine is. 

stellarator_assemblyAccording to this coverage in Science magazine, the W7-X project was given the green light in 1994, and was due to be completed by 2006 at the cost of €550 million. But the reactor involves 425 tons of specifically-shaped superconducting magnets and the subsequent cooling apparatus. They ran into manufacturing troubles with the magnets, with a third of them failing inspection and one of the firms producing them going bankrupt.

Not that delays or cost overruns are uncommon in fusion reactors — the ultra-large ITER tokamak is probably going to cost around thirteen billion Euros (three times the original budget) and be eleven years late by the time it’s firing a plasma in 2020. Creating stable nuclear fusion just isn’t easy.

But the W7-X looks to be ready for a real plasma firing any day now. They’ve already pre-tested the containment on (cold) electrons, and everything’s reported to be perfect. If they can keep a hot plasma stable for a long(er) time, it may pave the way to actual fusion power generation. Keep your fingers crossed!

82 thoughts on “Stellarator is Germany’s Devilishly Complex Nuclear Fusion

  1. this is exciting time for nuclear fusion research. At least three ideas are being tested at roughly the same time: inertia confinement at National Ignition Laboratory, tokamak at ITER, and stellarator at W7-X in this article

        1. Care to elaborate on the Pollywell being infesable? From all the research I’ve read on the various fusion technquies this seems to be the most promissing. Plus, I like that if it uses fuels capable of aneutronic fusion like D-3He or p-11B, it can use Direct Energy Conversion, removing the need for a thermal conversion system.

        2. Care to elaborate on the Pollywell being infesable? From all the research I’ve read on the various fusion technquies this seems to be the most promissing. Plus, I like that if it uses fuels capable of aneutronic fusion like D-3He or p-11B, it can use Direct Energy Conversion, removing the need for a thermal conversion system.

        1. And power density is very low. The fusion core produces about as much power per cubic meter as an efficient compost pile. The only reason it’s overall power output is high is because it’s so big. But a 1GW fusion reactor at that power density would have to have a contained plasma volume of about 1000 acre-feet.

    1. NIF was and will never be a serious fusion research tool, it was designed for testing our Nuke stockpile and the fusion energy research bit was thrown in to make it more palatable to the public.

      Tokamaks are a dead end, they just keep beating that dead horse. Plus there is no real good way to use the neutrons emitted by it to get useable energy. Running steam engines off it is just silly. And then there is the problems of the neutrons destroying the chamber and leaving everything activated.

      1. how nice it must be to know the future,

        NIF is just a testing facility, you are correct there, but that doesnt prevent it from reaching break even, repeating that x times a second for power generation is an engineering issue after that.

        1. NIF will never break even because the amount of energy needed to for power generation can’t be sustained. The gain medium and optics would be fried. Even if it could produce the energy required, the device would incur unacceptable losses in amplification.

          The only place laser fusion makes sense is in the context of Chirped Pulse Amplification. A lower power wide-band laser source is split using diffraction gratings, stretching the pulse into a series of pulses along the timescale. The resulting beam is more suitable for amplification and recombination into an extremely short, but high power beam that can be used to heat the fuel directly.

          This is the goal of the Japanese/French FIREX program. In 2015 they completed the upgrades to the GEKKO XII laser as part of the LFEX experiment and succeeded in producing 2000 trillion watts for a picosecond. Reactors that produce energy will require a somewhat more sustained heating period, on the order of 10-20ps, but we’re getting extremely close. Now that we know it can be done we can improve the laser designs and start talking about test reactors that break even.

  2. The Science magazine article talks about .550 million Euro, which is a weird way to say, that around 500 million were scheduled for 2006.

    A look at the project page here: reveals a projected cost of 320 million DM in 1994, roughly 160 million Euros.

    The cost for the facility in 2014, were slightly more than a billion Euros. The cost of the experiment adds another 370 million Euros. Source:

  3. Just a quibble, but shouldn’t it be “where the energy PRODUCED from the fusion reaction is greater or equal to the energy used in creating the necessary hot plasma”, not extracted. Since, technically, NO energy is going to be EXTRACTED from this reactor.

    1. If you are not extracting energy, and it is producing more than you put in, the build-up would destroy the machine. So even if you are extracting it into a dummy load that is still extraction. Otherwise the device would be a bomb.

      1. They are not extracting anything. They’re simply limited in how long they can run before things start to overheat.

        They’re not expecting the plasma to stay lit for more than a handful of seconds anyways. The ITER would break even if it could hold the plasma for 60 seconds. For that sort of timescale, the sheer mass of the device is enough to absorb any energy it produces, and then they just kinda fan it down for a couple days while they go over the data.

      2. I think you’re overthinking this. There’s no way this machine is going to produce a sufficient amount of excess energy for a sufficient amount of time to destroy itself or become a bomb. Any EXCESS energy it MAY produce should easily be handled by whatever cooling system that must be in place to remove the heat generated by the machine during normal operation.

        Although technically, I suppose, the cooling system would qualify as EXTRACTING any excess heat “greater or equal to the energy used in creating the necessary hot plasma.”

        1. I would guess that, since they’ve only managed to get one running for 6.5 seconds in a row, heat build-up is not yet a problem they need to deal with. Indeed the thing’s full of electric heaters to help get it up to temperature.

          I’d also guess that cooling the plasma by much would stop it from being a plasma, and thus no longer conductive and able to be manipulated by magnets. If it were powerful enough, and (I guess) it’s not, it would vapourise whatever it comes into contact with.

          So presumably most of the heat given off by the plasma is radiated (isn’t there a vacuum in there?). Presumably into the walls, where you can remove heat however you like, including dumping the whole thing in a pool of water as Torcue suggested.

          Still I suppose this is all somebody else’s problem, for our scientists, who are still working on getting much fusion at all. To have constant fusion, for minutes or hours would be great. I wonder what’s stopping it now?

          1. That was explained in the article. The ions in the tokamak design drift towards the outer wall when they go around because of the non-uniform magnetic field. There they cool down, stop being plasma and drop out of the magnetic containment.

            An additional current is induced in the plasma to recirculate the ions back the other way, but this current reacts with the magnetic contaiment and becomes unstable and eventually breaks down into oscillations because the entire magnetic field starts to wobble and wiggle, and eventually all the plasma flies off to the walls and the reactor shuts down.

            It’s been described as trying to bind a doughnut made of jello with rubber bands.

  4. This post is another technology ‘news’ item that belongs on other techie websites, but not hackaday. You are deviating too far from your original charter. The more you dilute your content on slow ‘hacking news days’ by posting such material, the less relevant you are becoming. Just sayin’…..

      1. wee oo wee oo, call yourself “” if you keep posting non-hack articles, because half the new non-hack stuff is basically verbatim from their old articles, and the other half is something lifted out of

        1. Point being that if you become a regurgitator of topics and articles already covered by other regurgitators (news aggregators), you might as well just post a bunch of links to them and save yourself the effort.

          When you stoop down to the level of a web-crawler bot because you’ve got to fill some content quota, you’re doing yourself a disservice.

        1. That’s basically cargo-cult science.

          Simply “writing it down” results in garbage, like Cold Fusion and Overunity, Ley Lines, Chiropractice, Phrenology, Em-Drive type of garbage, because what you write down depends on what you think is happening and what you think you are doing, which is not the same thing as what IS happening.

          There’s a whole philosophy of science you should try to to adhere to, to ensure you aren’t fudging your own results and simply writing down figments of imagination:

          But Mythbusters and Adam Savage is TV Science with a capital S. It’s Michio Kaku talking about the singularity on the History Channel as if nothing he says is simply pulled out of a hat, or Neil DeGrasse Tyson pretending to be Carl Sagan while presenting historically inaccurate interpretations of Galileo Galilei as a martyr of Science and leaving half the story untold (Hint: Galileo was a crank scientist by modern standards). It’s geared towards selling a monolithic picture of “Science” as an infallible unfaltering cornucopia of knowledge that has Facts and Evidence you can slam on the coffee table with impunity, because you’re absolutely sure you know where the line between what you know and what you don’t know is drawn.

          1. Though the reason why this TV Science is being promoted despite its inaccuracies and omissions is because of a social need. The moment you expose that the “Science” isn’t actually as settled as you make it out to be in as many areas and topics that you pretend, the religious and the woo people creep in and start to insert their stuff like jews sticking prayer notes into the wailing wall. Soon you can’t see the wall.

            And also because simple people like simple facts.

            So Adam Savage can’t say on television that the difference between science and fooling around is keeping appropriate statistics and sample sizes, null hypotheses, experimental controls and falsifiability, and actually having some basic idea of what the fuck you’re trying to do, so you don’t go around explaining phenomena with pixies. That would undermine the authority of the Scientists by opening them to the question, or whether they’ve done their jobs properly – and that’s bad TV.

    1. Oh that? That’s easy, just drop it into water like a fission reactor. Clearly boiling water and then using that to spin a turbine is the most efficient way possible to convert heat to electricity.

      1. That was one thing that hit me as a huge disappointment when I was first learning about nuclear energy back in school. I was so excited to find out how they converted it into usable power ……. aaaand they effectively boil water….. I guess I was expecting something a little more exotic.

        Sorta had the same thing happen with computers, they went from magical boxes of infinite mystery to inserting tab A into slot A, molex molex errywhere :P

    2. Sadly they don’t extract heat from the plasma. The process produces a lot on neutrons and the idea is that they will produce heat from the total neutron flux and capture that energy through traditional means.

  5. this is why im a polywell fan. its a simple machine than anyone can afford. tokamaks are huge, complicated and expensive. stellarators, as cool as they look, seem to be going in the direction of more complex, more expensive than tokamaks. fusion is worthless if a power plant is less cheap than current fission plants.

  6. I worked on a magnetic gun when I was a laboratory assistant at harwell, and later on a flowing Z pinch as a lecturer at Liverpool University. Combine the two with a magnetic gun coil arrangement after creating the flowing plasma, and keep hitting the plasma to compress it as it travels along the tube. It is a lot cheaper than a tokomac, solves the wall loading problem, and a sodium blanket to steam output would give a cheap reactor.

    1. So, am I guessing right that all that is stopping you from creating your cheap reactor is your inability to write up the scientific details and getting the proper funding for developing it?

  7. Well, complex to operate is debatable. Compared to a Tokamak it apears a lot simpler.

    And most of those access ports will be gone in a operational reactor. It’s a test bench.
    (Some now external tech will be moved inside the chamber too. Things are not only taken out.)
    IIRC Wendelstein does not yet breed it’s own Deuterium. But it’s a technologically solved problem.

  8. “They ran into manufacturing troubles with the magnets, with a third of them failing inspection and one of the firms producing them going bankrupt.”

    If this bankruptcy delayed the project with 9 years, it makes you wonder why they did not support the company. In the end, if they could have finished this device 9 years earlier, at an extra cost of, say 100 million, they might have had a commercially viable product 9 years ago. Everyone would have been a millionair.

    I guess that intelligence in smart people does not stop them from making stupid decisions.

  9. I came from this little City in the north of Germany callrd Greifswald ;-)

    In The Time were the reactor ware unfinished I have had the possibility to take a lock inside by my own. Pretty impressive!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.