Machinery of an Energy Dream: How to Keep Fusion Going Long Enough | Hacker News Reader

Machinery of an Energy Dream: How to Keep Fusion Going Long Enough

nytimes.com

23 pointswikiburner13 comments

The question of sustainable fusion receives a lot of press regardless whether it is a Tokamak or laser-ignited fuel pellets. However, the question I'd always love to see get some attention is the engineering required to make it a successful power plant -- will it use heat transfer and neutron thermalization like traditional fission reactors to heat a coolant? How will that interface look, and what kind of maintenance issues are going to go along with that?

I'm waiting especially eagerly in the case of laser-ignited fuel pellets, as to me it's like harnessing the power of a multitude of miniature discrete hydrogen bombs.

The tentative plans for DEMO (the follow up to ITER) address the power generation side of things:

http://en.wikipedia.org/wiki/DEMO

Laser-ignited fuel pellets is all about nuclear weapons testing to validate computer models not net energy generation. It's even paid for by "the 'Nuclear Weapons Stockpile Stewardship' program, and the goals of the program are oriented accordingly" It's possible to scale the idea to the point where it would generate energy but Tokamak's are far cheaper and much further along. As in we could build one that generates net energy today it's the material science stuff as in how long do the wall last under heavy neutron bombardment there still working out.

"ITER's goal is to produce 10 times as much power as is required for break-even" but what there going to be doing with it is testing different material blankets and magnetic control schemes plasma heating methods etc. There is also some concern about trying to stabilize as self heating plasma. (Of interest it's supposed go generate 500MW of thermal energy for ~10+ minutes so dissipating that is rather important.)

After that DEMO is supposed to be a working prototype fusion reactor.

PS: Jet reactor built in 1982 was reasonably close to break even. It was assumed we could build a working reactor next but it was simpler and cheaper to scale ITER down as much as possible which removed the energy generating side of thigns which is both expencive and well understood. http://en.wikipedia.org/wiki/Joint_European_Torus

"we could build one that generates net energy today". That's a pretty broad claim, considering the article linked itself notes that we've only hit a Q factor (output/input) of 0.7 with a test reactor measuring all the available energy. We still need a Q of 5, or a little under a factor of 10 (7.14) jump to be able to even have sustained fusion, as we specifically need the energy of the alpha particles, and need an estimated Q of 10 (factor of 14.3 jump, "Indispensable Truth") to be able to extract useful power. Any engineer will tell you jumping the efficiency of almost any system enough to get a 14.3x jump is not trivial.

We are making steady progress, and we're at least getting near break-even, but we're still a ways from rolling out Tokamak's as the one true solution. Frankly, I think the final solution may be a combination of strategies (magnetic confinement, inertial confinement, electrostatic confinement, and / or maybe one of the other ones, like sonofusion or the z-pinch).

Q is only the first problem. Tokamak's are considered the front runner because they scale ridiculously well so getting something as small as JET to 0.7 briefly means something with 10x the volume is easy to hit a Q of 10. Jet's was design work started in 1973 we simply have never built another device that large again.

However, the Tokamack confinement record is only 6 minutes 30 seconds. The neutron bombardment at scale is insane and grid power needs a steady and dependable energy supply, nobody has a ready supply of Tritium so they need to breed it somehow etc etc.

Worse even if we built something that was clean and net energy positive so are windmills so the reactor also needs to be cheap. Also, while far safer and cleaner than fission fusion still generates a fair bit of radioactive waste due to all that neutron bombardment.

I'd like to see someone, somewhere address the accounting realities of fusion, even if the science and engineering can be made to work.

The various scaling laws have pushed the minimum viable power plant size higher and higher. Now we are at the point where even if ITER and DEMO work out exactly as thier designers hope we will have "commercially viable fusion" that involves a monolithic multi-gigawatt plant that requires the bulk of its TCO to be made up front before any electricity is produced. That means profitability will be highly dependant on financing rates, construction time, and above all stable or increasing demand and prices for decades after construction begins.

> I'd like to see someone, somewhere address the accounting realities of fusion, even if the science and engineering can be made to work.

Yeah, well, that's part of the problem with fusion. When it works once, it sucks all the profit out of the energy market.

That's why we do need governments to step in and fund things like this.

Nothing to see here. The story is just a "where we are now" update with nothing new or significant to add for anyone casually following progress on controlled fusion.

Of all the various fusion approaches out there, I believe that dense plasma focus fusion (dubbed by one group pursuing it as 'focus fusion') is the most promising. Considering the ridiculously small budget they've had to work on it, the Lawrenceville Plasma Physics team has worked wonders. If I had the money I'd show no hesitation in investing. http://www.lawrencevilleplasmaphysics.com/

“Awesome science story, but there’s a zero percent chance that a fusion laser scientist named Dr. Hurricane isn’t a supervillain”

I got a pay wall...

10 free articles per month.

You could either pay or bypass it. https://www.google.com/search?q=bypass+nyt+paywall

PG needs to automatically add a little "Paywall?" tooltip to Times submissions, because this comes up every time.

Just zap your cache, use Incognito, an alternate browser, search for the article in Google News, or, Gasp!, buy a subscription, and you'll be able to view it. They limit you to 10 a month.

Reply on news.ycombinator.com