Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

ITER: World's largest nuclear fusion project begins assembly

bbc.co.uk

Post reply on HN
Print view View original

461–466 of 466 posts

Re: ITER: World's largest nuclear fusion project begins assembly

#461

Post by pfdietz »

throwawaysea wrote: post #30

ITER won't be commercially viable. After ITER comes DEMO, which is itself a decades-long project. DEMO won't demonstrate electricity generation until 2048: https://en.wikipedia.org/wiki/DEMOnstration_Power_Station#Ti... In my mind, that means we need to see deployment of Gen 4 fission reactors ( https://en.wikipedia.org/wiki/Generation_IV_reactor ) commercially, to bridge the gaps between a decline in coal-fired plan…

There probably has to be a machine before DEMO to solidify the engineering and materials.

Also, after ITER, there will not be enough tritium to do another large machine (the tritium comes from heavy water reactors, but those have lost in the market and will be shutting down in the next couple of decades.) So a machine to just make tritium may be needed.

Re: ITER: World's largest nuclear fusion project begins assembly

#462

Post by pfdietz »

lambdatronics wrote: post #435
nucpal wrote: post #346

Earlier quoted context omitted.

It's mainly the waste. Nuclear fission is a stable source of energy that does not produce carbon emissions in production. It's pretty much the answer to the climate crisis except that it produces dangerous waste that is horrendous to store and manage. Fusion mainly avoids this :)

Also, the accident scenarios are not as bad -- there's less total radioactivity in the device, and there's no meltdown risk.

If you had a fission reactor with a power density as low as a fusion reactor, it would be incredibly safe also. All that thermal inertia, accidents would go in extreme slow motion.

You might ask why fission reactors aren't designed that way.

Re: ITER: World's largest nuclear fusion project begins assembly

#463

Post by lambdatronics »

pfdietz wrote: post #462
lambdatronics wrote: post #435

Earlier quoted context omitted.

Also, the accident scenarios are not as bad -- there's less total radioactivity in the device, and there's no meltdown risk.

If you had a fission reactor with a power density as low as a fusion reactor, it would be incredibly safe also. All that thermal inertia, accidents would go in extreme slow motion. You might ask why fission reactors aren't designed that way.

Interesting question. NuScale (https://en.wikipedia.org/wiki/NuScale_Power) reactor appears to have a very low power density around 1.5 MW/m^3, by my estimate (2.7m dia, 20m tall, 60MWe -> 180MWth). They still rely on having coolant (water) present in the reactor to avoid a meltdown, although it should circulate automatically by boiling & condensing.

The trick is that something like 10% of a fission reactor's thermal output continues for many hours due to decay heat, whereas with a fusion reactor that can be orders of magnitude less, with appropriate material choices.

Re: ITER: World's largest nuclear fusion project begins assembly

#464

Post by lambdatronics »

pfdietz wrote: post #446
lambdatronics wrote: post #436

Earlier quoted context omitted.

Solar was 2 orders of magnitude out of the running around 1975 -- and now it's at grid parity. So maybe fusion is 45 years away? :P You're right that the cost is going to have to come down, and the power density up, before fusion can compete. High-field superconductors are one obvious route, and using plasma configurations that make better use of the magnetic fields are another. Either one of those approaches could d…

That approach doesn't work, because reactor power density becomes limited by wall loading (neutron and/or thermal), regardless of how good the plasma physics is. It's the square cube law. At a given wall loading limit the volumetric power density is inversely proportional to the linear dimensions. And a DT fusion reactor must be meters across, due to the fixed cross section of the neutrons with wall materials. In con…

>the entire first wall is thick flowing liquid lithium

Yeah, I agree on the liquid walls. I think the optimum might be to have a thin Li first surface flow that is relatively cool, then a shell of SiC, and behind it a PbLi breeding blanket that can be at higher temperature. There's a trade-off because the first-surface can't be made too hot or the evaporation will pollute the plasma, but getting high thermodynamic efficiency means using higher coolant temperature. The thickness of the liquid first-surface plays into the lifetime of the SiC shell -- more shielding in front of it means longer life (so lower maintenance costs & higher reactor up-time, and less radwaste), but lower thermal conversion efficiency on average.

There are problems of course -- corrosion by Li, splashing of droplets into the plasma, MHD drag & pump power requirements, incompatibility of lithium with many forms of sensors & actuators for plasma control.... (OTOH, most sensors and actuators can't tolerate radiation anyway, so fancy control techniques just don't stand a chance in a reactor anyway. We need boring, stable plasma configurations that just sit there and work.)

>What is the other reason?

Some people think they can make a fortune :D

Re: ITER: World's largest nuclear fusion project begins assembly

#465

Post by nucpal »

pl-94 wrote: post #459
nucpal wrote: post #346

Earlier quoted context omitted.

It's mainly the waste. Nuclear fission is a stable source of energy that does not produce carbon emissions in production. It's pretty much the answer to the climate crisis except that it produces dangerous waste that is horrendous to store and manage. Fusion mainly avoids this :)

OK but some models of gen iv of nuclear power plants can pretty much do the same. It sounds pretty dumb to me that journalists are super enthusiast about fusion but much less about gen iv (which is way more mature)

"some models of gen iv of nuclear power plants can pretty much do the same"

I'd be interested to know which fission plants make fuel recycling and waste disposal trivial, I could make a lot of money!

In reality, even the gen 4 plants that have interesting approaches to reprocessing still produce a lot of harmful radioactive waste. Not all the fuel can be recycled and the process itself is imperfect and messy.

Re: ITER: World's largest nuclear fusion project begins assembly

#466

Post by Mvandenbergh »

gclawes wrote: post #23

At this point I've got a lot more hope in the MIT/Commonwealth Fusion Systems approach with REBCO magnets: https://www.youtube.com/watch?v=L0KuAx1COEk At this point it looks like ITER is hampered by it's relatively old supeconductor technology (ultra low temp/moderate field strength traditional magnets vs high temp/high field REBCO magnets).

Luckily, since ITER and SPARC are being built with the same aspect ratio, all the learning about plasma control, materials, cooling, tritium, remote handling, etc. is fully transferable.

The tritium extraction and processing is a whole separate multi-story building full of first-of-a-kind equipment which will be 1:1 transferable to any breeding fusion reactor.

The work that ITER and IFMIF will be doing on material lifetime and handling under heavy neutron bombardment - a really substantial engineering problem - will be fully transferable.

The work on first-wall material which has to handle the neutron flux, very high thermal loads, and not poison the plasma when traces of it come off is also fully transferable.

Basically everything that's really new about ITER except for the size will work the same way on an HTS based machine.

I'd say about 2/3 for the science done at ITER would need to be done for any D-T fusion device, another 1/6 applies to all similarly configured tokamaks (i.e. it's less less relevant for stellarators or spherical tokamaks), and 1/6 is ITER specific (high estimate TBH).

If things run according to schedule (obviously questionable) then in 2025/2026 ITER will have first plasma and CFS will be on schedule to start building SPARC. CFS is being very clever in doing all their magnet design work first - investors are funding it because even if they don't get either SPARC or ARC funded and built, they will at least have some very useful IP on large HTS magnets which is bound to be worth something.

Post reply on HN