Hacker News Reader

Comment by nharada | Hacker News Reader

There's a good talk here about why ITER is so large: https://www.youtube.com/watch?v=L0KuAx1COEk

The idea is essentially that to get the parameters needed to make net energy with tokamaks, you need either very strong magnets or a large device. At the time of ITER's design, they used the strongest magnets they could find and then made the thing big enough to get the energy gain they wanted.

The speaker of this talk argues that it's size that stalled progress in tokamaks, since they'd become so big that building them became a massive, multinational project.

View on HN

I wonder how much the possible magnet strength affects the design of a tokamak. There are very clear limits on the field strength you can achieve with classical superconductors, and I know e.g. in Nuclear Magnetic Resonance those limits had been almost hit maybe a decade ago or so. But very recently spectrometers based on new high-temperature superconductors have been delivered, so the technology seems to be far enough for actual production use now. I'm not sure how big the possible increase in field strength will be. Right now it's 23.5 Tesla for the largest NMR spectrometer with conventional superconductors compared to 28 Tesla for the new ones with high-temperature superconductors (actually it's a hybrid with high-temperature superconductors on the inside, and conventional ones on the outside), but I suspect there is more room with the new ones for higher fields.

It would be interesting whether the availability of better superconductors would change the design of a fusion reactor much, and allow significantly smaller ones.

Yes that has a big impact. The quality of a tokamak (as meassured by the triple product density, temperature, and confinement time) goes up like B^3 (IF I remember correctly, I do plasma physics but not fusion stuff) and the fussion power goes up like B^4 or something like that. So larger magnetic fields help a lot. But back when ITER was designed we did not have such strong superconductors yet.

PS: Google found DOI 10.1088/0029-5515/56/6/066003 but I didn't read it carefully.

Is there a successor to iter already being worked out on paper that takes into account new developments?

There is things such as SPARC by Commonwealth Fusion Systems (CFS) using rare-earth barium copper oxide (REBCO) superconductors. Not quite the scale of ITER, but it should give us some first experience working with that material. If that works out the next step would be ARC [1].

[1] https://en.wikipedia.org/wiki/ARC_fusion_reactor

Yes - the sequence is ITER, DEMO, then PROTO:

https://en.wikipedia.org/wiki/DEMOnstration_Power_Station

https://en.wikipedia.org/wiki/PROTO_(fusion_reactor)

Yes, the video above is basically about this, and is really worth watching -- it's a fantastic talk.

I'm pretty sure that Commonwealth Fusion Systems[0] is the entity affiliated with MIT that has been doing work to prove out the recently available higher magnetic field strength superconducting materials and apply them to tokamak construction to bring size down dramatically. They had a bunch of press in 2018/2019 when they first got underway[1], and it looks like they've received a lot more investment over the last few years and likely made quite some progress since then[2].

0 - https://cfs.energy/ 1 - https://www.bostonglobe.com/opinion/2018/03/09/new-approach-... (op-ed in the Boston Globe by a Vice President at MIT) 2 - https://cfs.energy/press/

Cambridge Fusion Systems[1] is a private company spun out of MIT that is building a proof-of-concept reactor based on these new magnets within the next 5 years.

Breakthrough in Nuclear Fusion? - Prof. Dennis Whyte (2016) - https://www.youtube.com/watch?v=KkpqA8yG9T4

Timeline (in case you want to skip over some parts):

  00:01:00 - introducing Dennis Whyte, MIT department head for nuclear science
  00:04:24 - presentation starts
  00:06:00 - identifies breakthrough with REBCO magnets
  00:07:25 - explains deuterium-tritium fusion
  00:12:30 - basic metrics for reactor performance
  00:17:15 - energy output of other previous fusion experiments
  00:19:00 - examines ITER and the problems of its approach
  00:22:00 - problems solved by high energy magnetic fields
  00:28:15 - full scale reactor concept, teardown of REBCO magnets
  00:37:00 - design limits and margins
  00:39:00 - fixes plasma instabilities found in weaker magnetic chambers
  00:40:00 - maintainability, lifespan, component replacement
  00:45:00 - solution to neutron damage and energy capture
  00:50:30 - cost and profitability
  00:54:00 - full graph of field strength vs reactor scale (and thus funding requirements)
  01:01:50 - Q&A
  01:30:00 - question about the biggest risks

He gave another talk in 2019 with more numbers and even more confidence: https://www.youtube.com/watch?v=rY6U4wB-oYM

1: https://cfs.energy/technology

And then you go to the arxiv paper for ARC and learn: the power density is 40x worse than a PWR primary reactor vessel, and their projected cost is $29/W(e) (vs. < $1/W(e) for PV). Also, the vacuum vessel likely wouldn't survive a disruption (although maybe they've fixed that in the years since).

Compact high field tokamaks have better power density that ITER or DEMO would, but they still are very inferior to fission reactors. And fission reactors are far out of the running economically.

^ That is a _fantastic_ talk for someone who know's some physics get a quick grasp of the state of the field, the different approaches being taken from a set of first principles, and how to evaluate them.

Also check out this UK company which is instead going down the path of using stronger magnets: https://www.youtube.com/watch?v=hWYnbYsp3i8

I didn't watch the talk, maybe it's the same one, but I recall watching a talk about Tokomaks that said thanks to the development of high-temperature superconducting wire ITER is already an obsolete design.

Awesome talk. Thank you for the link!

Reply on news.ycombinator.com