6 ms·

This is exciting. Magnetic field strength is a key component for enabling magnetic confinement fusion. This is because energy gain and power density scales to t

This is exciting. Magnetic field strength is a key component for enabling magnetic confinement fusion. This is because energy gain and power density scales to the 3rd and 4th power with magnetic field strength but only ~linearly with reactor size. See following equations for more details: https://youtu.be/xJ2h3vbOag4?t=306 https://youtu.be/xJ2h3vbOag4?t=306 So, why is this particular announcement exciting? There are 3 factors: 1. This is a high temperature superconductor. I can't find any references, but as far as I remember the substrate they are using needs to be cooled to (WRONG, it was cooled to 20degK, see reply by MauranKilom) 60-70 degK to achieve super conductivity. Compare to magnets used in ITER which need to be cooled to 4degK. This is the difference between using relatively cheap liquid nitrogen vs liquid helium. 2. Field strength of 20 Tesla is significantly higher than 13 Tesla used in ITER. Given that magnetic confinement fusion scales significantly better with field strength vs reactor size, this will enable much smaller reactor to be power positive. See following links for more details on ITERs magnets: https://www.newscientist.com/article/2280763-worlds-most-powerful-magnet-being-shipped-to-iter-fusion-reactor/ https://www.newscientist.com/article/2280763-worlds-most-pow... https://www.iter.org/newsline/-/2700 https://www.iter.org/newsline/-/2700 3. Finally, the magnet was assembled from 16 identical subassemblies, each of which used mass manufactured magnetic tape. This is significantly cheaper and more scalable than custom magnet design/manufacturing used by ITER. The kicker is how 3 of the factors above interact with the cost of the project. Stronger magnets allow smaller viable reactors. High temperature superconductors + smaller reactors allow for a much simpler and smaller cooling system. Smaller reactors + scalable magnet design further drives down the cost. Finally, cost of state of art mega projects scales somewhere between 3rd and 4th power with the size of the device. Combining all of the above factors, SPARC should be here significantly sooner than ITER and cost a tiny fraction (I would guesstimate that fraction to be between 1/100 and 1/10,000). edit: typos + looked at the cost of ITER and refined my cost fraction guesstimate + corrected some stuff based on the reply by MauranKilom.

MauranKilom 5y ago
Appreciate the rundown of why this is important! > This is because energy gain and power density scale exponentially with magnetic field strength but only linearly with reactor size Nit: It scales polynomially, not exponentially. Specifically (according to those formulas) energy gain scales with the cube of field strength and power density with the fourth power. Still massive scaling indeed, but exponentially would be something else. > as far as I remember the substrate they are using needs to be cooled to 60-70 degK to achieve super conductivity The video in the article shows 20 K. Could of course be that higher temperature is feasible and they just played it safe (or the video is wrong).
- TrainedMonkey 5y ago
  You are right on both counts, I will update the comment. 2020 paper referenced 7degK magnets. I am not sure where I got 70degK number from... https://www.cambridge.org/core/services/aop-cambridge-core/content/view/DD3C44ECD26F5EACC554811764EF9FF0/S0022377820001257a.pdf/div-class-title-overview-of-the-sparc-tokamak-div.pdf https://www.cambridge.org/core/services/aop-cambridge-core/c...
- zardo 5y ago
  There is a temperature dependent maximum field strength the material can take and keep superconducting. REBCO superconducts at 70k, but it can't take much current at that temperature.
fabian2k 5y ago
I'm curious if they can push the magnetic field even higher in the near future. For smaller magnets in NMR spectrometers 20 Tesla has been commercially available for 20 years. Of course this is more difficult for larger magnets. The new superconductors that allow these larger magnets are also very recent, not in discovery but in actual mass production. So they don't have as much experience with using these as with the classical superconductors. So I hope there is still quite some quick improvement there on the table.
- elihu 5y ago
  I think so far the ReBCO tape is manufactured in small quantities by a handful of suppliers. If there were million-dollar orders coming in regularly, I suppose there'd be a lot of competition to develop the highest quality product. I expect it'd be like batteries; there's a lot of incremental improvements, and then once in awhile a major chemistry change. There's probably other high-temperature superconductors just waiting to be invented. If I remember right from one of the videos from the SPARC reactor folks, they were experimenting with not bothering with insulation between the magnet windings. The ReBCO film is bonded to a layer of stainless steel, and they figured the conductivity of the film is so much better than stainless steel that they wouldn't actually get much loss from current leaking through. That seems kind of crazy, but I guess there's a lot of things about superconducting materials that don't behave intuitively. Maybe manufacturers can make film that's bonded to a thinner layer of stainless steel or whatever, and thus allow for more windings in the same space?
  - fabian2k 5y ago
    I've visited a company that produces classical and high-temperature superconductors, though this was quite a few years back. I don't know the exact market size, but the MRI and NMR markets are probably not that small, though they use almost entirely classical superconductors right now. But they have hit the physical limits of classical superconductors in NMR, and the first NMRs with high-temperature superconductors are produced and sold now. So there might be some more development there even without the fusion angle. One purpose of the support material that isn't super-conducting is thermal protection. If your superconducter quenches, you have to dissipate the energy contained in it without destroying the magnet. In classical ones they use copper wire around them as far as I remember, and the high-temperature ones are a very thin film of ReBCO deposited on metal tape, so the actual superconductor is always a small part of the material.
choeger 5y ago
ITER plans first plasma for 2025 - do you think it is a coincidence that SPARC is planned for 2025 as well in that release? I think both projects will hit delays, but ITER is much further in construction, so I wouldn't bet on SPARC to win that particular race. But they don't need to, do they? If their claim is sound, they could as well just optimize the magnets and wait for ITER to complete to offer an ITERation (pun very much intended) on the design. The fact that they focus on this weird race against an international research project makes me wonder if SPARC is mostly a vehicle to attract investors.
- TrainedMonkey 5y ago
  I think biggest downfall of ITER is also why we must do it. The downfall is thus - ITER is huge and that generally implies lots construction delays and cost overruns. But, being huge also means it will be able to study sustaining high volume of plasma for long durations. ITERs plasma density will be comparatively low, and that is where SPARC with stronger magnets comes in. SPARC will produce data on lower volume and limited burn time, but significantly higher plasma density.

MauranKilom 5y ago

Appreciate the rundown of why this is important! > This is because energy gain and power density scale exponentially with magnetic field strength but only linearly with reactor size Nit: It scales polynomially, not exponentially. Specifically (according to those formulas) energy gain scales with the cube of field strength and power density with the fourth power. Still massive scaling indeed, but exponentially would be something else. > as far as I remember the substrate they are using needs to be cooled to 60-70 degK to achieve super conductivity The video in the article shows 20 K. Could of course be that higher temperature is feasible and they just played it safe (or the video is wrong).

TrainedMonkey 5y ago
You are right on both counts, I will update the comment. 2020 paper referenced 7degK magnets. I am not sure where I got 70degK number from... https://www.cambridge.org/core/services/aop-cambridge-core/content/view/DD3C44ECD26F5EACC554811764EF9FF0/S0022377820001257a.pdf/div-class-title-overview-of-the-sparc-tokamak-div.pdf https://www.cambridge.org/core/services/aop-cambridge-core/c...
zardo 5y ago
There is a temperature dependent maximum field strength the material can take and keep superconducting. REBCO superconducts at 70k, but it can't take much current at that temperature.

fabian2k 5y ago

I'm curious if they can push the magnetic field even higher in the near future. For smaller magnets in NMR spectrometers 20 Tesla has been commercially available for 20 years. Of course this is more difficult for larger magnets. The new superconductors that allow these larger magnets are also very recent, not in discovery but in actual mass production. So they don't have as much experience with using these as with the classical superconductors. So I hope there is still quite some quick improvement there on the table.

elihu 5y ago
I think so far the ReBCO tape is manufactured in small quantities by a handful of suppliers. If there were million-dollar orders coming in regularly, I suppose there'd be a lot of competition to develop the highest quality product. I expect it'd be like batteries; there's a lot of incremental improvements, and then once in awhile a major chemistry change. There's probably other high-temperature superconductors just waiting to be invented. If I remember right from one of the videos from the SPARC reactor folks, they were experimenting with not bothering with insulation between the magnet windings. The ReBCO film is bonded to a layer of stainless steel, and they figured the conductivity of the film is so much better than stainless steel that they wouldn't actually get much loss from current leaking through. That seems kind of crazy, but I guess there's a lot of things about superconducting materials that don't behave intuitively. Maybe manufacturers can make film that's bonded to a thinner layer of stainless steel or whatever, and thus allow for more windings in the same space?
- fabian2k 5y ago
  I've visited a company that produces classical and high-temperature superconductors, though this was quite a few years back. I don't know the exact market size, but the MRI and NMR markets are probably not that small, though they use almost entirely classical superconductors right now. But they have hit the physical limits of classical superconductors in NMR, and the first NMRs with high-temperature superconductors are produced and sold now. So there might be some more development there even without the fusion angle. One purpose of the support material that isn't super-conducting is thermal protection. If your superconducter quenches, you have to dissipate the energy contained in it without destroying the magnet. In classical ones they use copper wire around them as far as I remember, and the high-temperature ones are a very thin film of ReBCO deposited on metal tape, so the actual superconductor is always a small part of the material.

choeger 5y ago

ITER plans first plasma for 2025 - do you think it is a coincidence that SPARC is planned for 2025 as well in that release? I think both projects will hit delays, but ITER is much further in construction, so I wouldn't bet on SPARC to win that particular race. But they don't need to, do they? If their claim is sound, they could as well just optimize the magnets and wait for ITER to complete to offer an ITERation (pun very much intended) on the design. The fact that they focus on this weird race against an international research project makes me wonder if SPARC is mostly a vehicle to attract investors.

TrainedMonkey 5y ago
I think biggest downfall of ITER is also why we must do it. The downfall is thus - ITER is huge and that generally implies lots construction delays and cost overruns. But, being huge also means it will be able to study sustaining high volume of plasma for long durations. ITERs plasma density will be comparatively low, and that is where SPARC with stronger magnets comes in. SPARC will produce data on lower volume and limited burn time, but significantly higher plasma density.