Hacker News Reader

Comment by mlindner | Hacker News Reader

She also overstated things and made things out to be worse than they actually are. Getting to fusion plasma power breakeven IS the hardest part. And it's always been so far away, setting and talking about realistic goals is better than trying to immediately jump to the far end goal.

This is the difference between ITER and projects like SPARC though, they both plan to get fusion plasma power breakeven, but ITER's design has zero hope of ever being economical.

View on HN

As I recall her point about the deceptive reporting of gain figures was spot-on. It’s something I’d observed throughout the 20 years that I followed the field fairly closely, being intermittently involved in several fusion-related projects. Especially in inertial confinement, claims of “break-even” are almost entirely devoid of meaning. So how, exactly, do you think she overstated anything?

Extracting useful energy from the plasma, and protecting the reactor (magnets, control etc) from neutron bombardment are also hugely difficult problems.

Not to mention, trying to fool public opinion about how far away we are from any chance of a fusion power plant is still ethically wrong. Let's not forget that even if ITER achieves all of its goals and milestones, we will be able to start on the power extraction problem in 2030.

And let's also not forget that inertial confinement fusion, while also being presented as a potential fusion power plant concept, with great fanfare occasionally, is simply a weapons research program with no imaginable way of progressing to economical power generation.

How about space propulsion? Maybe I'm naive, but inertial confinement fusion seems like a great candidate for a space engine.

If you have the gigawatts of electricity required to run the inertial fusion lasers, there are probably better ways to use that power to get around. Massive ion engines, or powering the pumps for a nuclear salt water engine, maybe, depending on how exciting you want your ride to be.

Inertial containment fusion relies on firing a laser at an extremely precisely machined piece of heavy metal (called a hohlraum), heating it until it emits X-Rays, and using the X-Rays to create 2 implosion shockwaves in a tiny pellet of tritium + deuterium that have to meet in the dead center, so that where they meet they can start a fusion reaction, and ideally have the fusion reaction consume the whole pellet before the shockwaves dissipate enough to stop confining the plasma. A few microseconds after the laser blast, you get a puff of helium (+hydrogen if the reaction wasn't complete) and a mangled piece of metal, and some heat that in principle you could capture somehow, though no one has yet tried. Then, you throw away the now useless hohlraum, get a new one and a new pellet and you fire again for another little burst of energy, over and over.

The hohlraum is the biggest problem here: the level of precision needed to achieve the exact geometry inside the pellet to actually ignite a plasma means that every hohlraum is (a) extraordinarily expensive (currently in the millions of dollars range), and (b) entirely useless after a single shot - while continuous operation for a 1500 MW plant is estimated to require ~20 hohlraums/second).

For a spaceship design, this would mean that your engine would have to include a smelter and high-precision machining bay to constantly create new micrometer-smooth hohlraums from spent ones. Not even close to a promising technology.

I'd like to learn nuclear science. Where would I begin? I know basic classical mech, quantum mech, and classical field theory.

I'm a software engineer who saw a few videos on YouTube, and read a few articles on this topic, I don't think I am the right person to ask about this, sorry.

NIF also has an expendable blast shield in from of each mirror of the final optics. These mirrors have to be in line of sight of the explosion. If you tried using something like this in space, without that shield the expanding plasma/hohlraum fragments would go right out though the vacuum to hit the mirrors. Even slight imperfections there will explode under high energy pulsed laser irradiation.

CFS's ARC reactor design will use a FLiBe molten salt blanket to shield everything except the vacuum chamber itself, which is a steel doughnut with 2cm thick walls. They plan to build a proof of concept reactor by 2025 that will have net energy gain (Q > 10), and then a full demo power plant by 2030. CFS believes they can move faster than ITER because the newer stronger magnets that they're using make smaller reactors viable. Smaller scale means cheaper to build, which means a nimbler private company can build it rather than a bureaucratic international project.

Let's see if it actually materializes. It's not hard to make such claims, it's much harder to deliver on them. I very much doubt the 5 year timeline from "net energy gain" (net plasma energy gain) to an actual power plant (actual net energy gain). Again, extracting energy from the plasma is mostly a theoretical idea at this time, and all the proposed mechanisms would likely require significant engineering challenges (the biggest temperature differential in the solar system; neutron bombardment).

> Getting to fusion plasma power breakeven IS the hardest part.

I believe the engineering after that is harder, and likely intractable.

Reply on news.ycombinator.com