The Wendelstein 7-X concept proves its efficiency
ipp.mpg.de
ipp.mpg.de
It is for me a wonderful example of how a large engineering project to build a new thing should be approached. Lay out all of the questions that are currently unanswered for which the answer will affect the next step. Then start building with the goal of answering the questions in dependency order so that the next question/build can incorporate your new understanding given the answer to the previous question. Iterate until you've answered all the engineering questions and you're sitting there looking at a fully functional device that does this new thing.
One of the most interesting parts of that episode was how the described their relationship with the hardware manufacturers. The deep level of co-working on a very hard task and jointly developing the tech needed was really inspiring.
[1] http://omegataupodcast.net/312-the-wendelstein-7-x-fusion-ex...
Also, something older from the middle of 2014, again in german:
[2] https://resonator-podcast.de/2014/res032-der-wendelstein-7-x...
Most fusion projects seem to take years between major steps. Do you think a rapid-iteration approach (like what SpaceX is doing to develop Starship) could work for fusion research?
My impression is that a Manhattan Project for fusion would be successful. We already have all the technology needed to make a successful reactor. The reason things like HTS coil winding are moving so slowly to scale up is because there is no market for it. How could anyone invest in something with no return?
If a lot of money did suddenly become available then innovative and radical platforms could be seriously considered. Things like stellarator platforms designed with adjustable geometry coils to explore many different optimizations without needing to build a new machine every time would be on the table.
Issues related to first-wall and divertors require large machines to develop. They are not unsolvable problems, but those are the ones that need a serious amount of money to test and develop. Theory is way ahead of experiment here. I think it would be easier to find the money for these big machines if the performance metrics of smaller machines continued to increase. The problem is there isn't enough money to keep these new small-to-medium machines coming. Look at the history of FES spending in the DoE budget.
SPARC might be the community's best chance at convincing the public that fusion is worth investing in.
After working with a 50 kW gyrotron and 1 m major radius machine, the raw industrial scale of the project is awe-inspiring [0]. ITER has 20x 1 MW gyrotrons and a 6 m major radius. Every system in the machine is fascinating.
All that said, it is a slow and expensive project. The issues are mostly political. The design is very old at this point. All of this is fine, but in the context of potentially scaling up HTS supply chains to make a reactor that is much smaller, it becomes clear that the ITER path is not the one the first generation reactors will take. It's an important stepping stone in terms of plasma science and public opinion. It's going to be a lot of fun to follow its successes.
This seems like an odd question. You seriously don't think people invest in businesses with no profits, or even revenue? In 2021? One of the financial trends recently has been investing in blank check companies that don't even have a business let alone revenue or profits. Also the biggest investing celebrity recently, Cathie Wood, is known for buying "innovative" companies based on their far future prospects, and has reaped spectacular returns so far.
There's a joke about two economists, where one of them sees a twenty dollar bill on the ground and says so, and the other says, "can't be, someone would've picked it up already".
Practical fusion power is like a multitrillion dollar bill on the ground and it needs some really creative explanation why it wouldn't be picked up.
Setting aside all the biotech IPOs:
"SpaceX has raised a total of $6.6B in funding over 46 rounds. Their latest funding was raised on Jul 30, 2021 from a Secondary Market round."
https://www.crunchbase.com/organization/space-exploration-te...
Searching crunchbase for companies in the energy industry that have raised > $1B returns 24 results, too.
And fusion is hardest of the wares.
This 45-year old analysis [0] says that 15 Bn USD 1978 (63 Bn USD 2021) stands between us and a first gen reactor. Afaict that is still accurate as we have been on the Logic I path since then. How quickly you want to spend that 63 Bn USD dictates what logic (II-V) you're on.
* W-7X uses 3T magnets (traditional type-II superconductors), so it has to be big (to give more room for plasma instability to not hit the walls?)
* ARC can be smaller because it uses 21T HTS magnets, so the plasma can be more tightly confined.
* LTS magnets have an abrupt phase transition. Imperfections have resistance, which causes heat, which can quench the magnet if current gets too high. (?)
* HTS magnets for - reasons? - don't quench like this. The magnet just gets hot and slowly degrades.
* HTS magnets are really brittle and finicky to work with, so winding them without causing imperfections is hard.
Fusion aside, wouldn't 21T magnets be valuable in a lot of areas? Like couldn't you do really high res MRIs with such a magnet? I don't see why there's not more commercial interest in such a general technology.
One of the selling points of stellarators is that they can be operated without risk of disruption, but W7-X has done campaigns that push its operation to disrupt. A machine with dimensions similar to W7-X could be made with higher field magnets.
* Yes, higher field means larger machines are not needed to hit the same performance metrics, pushing down the minimum size (and cost) of a reactor. REBCO (HTS material of choice) is quite an expensive material to procure and work, so even modest sized research devices are more expensive than Uncle Sam is willing to gift right now. Private investment into SPARC is changing my opinion about what the future of fusion research will look like.
* The details of working HTS into confinement coils is an active area of discussion. HTS coils will be operated at liquid helium temperatures but won't stop superconducting until liquid nitrogen temperatures. The critical current limit decreases as temperature increases, so the risk of quench is still there when operated at near-boundary conditions. One thing that is discussed is whether to insulate the HTS tape or not. Conventionally, coils have insulation between the windings. In the case of superconductors it isn't strictly necessary. Superconductors are embedded in large copper jackets. When the core is superconducting the copper acts as a virtual open. When the core quenches the coil turns into a one turn copper ring and can dump all of the stored energy through the leads quickly without boiling cryogenics because the inductance is much lower.
High field magnets are useful, just not to many areas. MRI is the singular commercial application and they are more interested in lower cost (ie being able to run a superconducting magnet without liquid helium) than they are a higher field. NMR is the other big (but actually quite small) application for high field magnets. There are commercially available NMRs today with fields of 50 T.
They've been working on their ARC reactor design, and decided they wanted a smaller, easy-to-make prototyping platform that they can run tests with. If I understand correctly, SPARC isn't designed for long-term use, or practical energy extraction. If they destroy one, they'll just make another.
If I remember right, they're working on re-mountable magnet coils for ARC so that you can just de-solder the windings and take the whole thing apart and replace the inner lining when necessary. I don't think SPARC has that.
Different types of engines have different details of where that energy gets lost, but ultimately they are devices that if left to run indefinitely would equilibrate in some way and the way that they idle tells you a fair bit about where the energy flows are.
Turbines are pretty impressively efficient, but something can always be better. They often utilize the low grade waste heat in steam heating though, so a lot of times the efficiency is extremely good when you include free heating or process heat (in a plant they might use that steam to heat another piece of equipment).
It hits the sweet spot on fiscal, engineering, safety, and science fronts. Energy extraction is very challenging, we're likely hundreds of years away from aneutronic fusion, for example. The research area you are asking about is direct energy conversion, of which aneutronic fusion is one small branch (though within it, there are many scientific and engineering branches to explore).
We're only a bit over a hundred years from developing quantum mechanics and relativity, how can we possibly say anything about hundreds of years from now?
If something is well enough understood to accurately predict the timeline, we could do it much sooner.
If we have no idea how to do something, then "hundreds of years" means nothing, except maybe "not proven impossible yet".
- deuterium-tritium fusion, which produces lots of high-energy neutrons
- the neutrons help to keep the plasma hot, but being neutral, they escape the magnetic bottle; there is a lithium blanket around the reactor to catch them
- the neutrons that hit the lithium produce tritium for the reactor, and steam for the turbines
"Dense Plasma Focus device with aneutronic hydrogen-boron (pB11) fuel"