Comment by Robotbeat | Hacker News Reader

Comment by Robotbeat | Hacker News Reader

Parent Full thread

Magnetic confinement fusion like ITER is no less of a boondoggle. Maybe even more so because the progress is intentionally slow in spite of not having a dual-role for “stockpile stewardship.” ITER is being funded not just by the US but by many countries, started development in 1985, detailed design in 2001, and construction in 2013, but it’s not even PLANNED to get full fusion until 2035. 2035!

Plus, it won’t even generate electricity at all. That’s planned for the DEMO reactor that won’t start operation until 2051 at earliest. It is depressingly slow if you think one of the main reasons we should be developing alternative energy sources is to address climate change. It’s so bad as to qualify as a waste and maybe even a negative investment as it’s pulling a bunch of researchers toward a project that literally has no hope of being relevant to fighting climate change (as its first possible kilowatt-hour of electricity won’t start until 30 years from now, well after we’ve exhausted our carbon budget for 2 degrees C of warming).

I also understand that fusion in general is unlikely to be an economically-viable energy source anyway, since the reactor and any surrounding material will be relatively quickly (~few years) be made brittle by the neutron bombardment, while also becoming radioactive - meaning any fusion plant will have to be carefully and constantly torn down and rebuilt, and materials from the old plant securely stored for large amounts of time (not as large as fission waste, but still in the order of decades or centuries). There are other concerns with hydrogen escape etc, but this one seems completely fundamental.

Some methods of fusion solve this in varying ways, with liquid metal blankets, etc, or using non-neutron is fusion fuels. But that’s missing the point. There’s no path to these more viable methods of economically producing power that don’t run through the path of generating more fusion energy than it absorbs, so we start with the easier fuels first to prove we can do sustained fusion before worrying too much about neutron embrittlement.

Even if we had unobtainium that was free of radiation degradation, fusion reactors would still be unlikely to be competitive -- they're just too large and complex, and hence expensive.

The problem is that ITER is funded as a science project, and the researchers want to get as much research as possible.

So they are going to spend a lot of time studying plasma before they irradiate the vessel with fusion byproducts and it's no longer safe to take apart (for example, to add new sensors).

It's the only facility of this size so the research program is completely sequential.

We could have fusion, we just need to spend $20 billion a year for 10 years. Not $1 billion a year for 200 years.

They could spend the same amount per year and get results in 5-10 years if it were being run competently.

There’s no point in babying a facility if it means taking decades too long to get useful results!

The first job of ITER is to show that disruptions can be controlled. This is absolutely necessary, and requires access to the machine to repair it when disruptions occur. So this had better be done without tritium (or possibly even deuterium). And if they can't do it, they will never be allowed to operate the machine with tritium.

What about "magnetic mirrors"? https://www.llnl.gov/archives/1980s/mirror-fusion-test-facil...

I recall hearing a scientist from the lab say that was the way to go and they mothballed it because they wanted to focus on weapons research.

Mirrors are interesting. They hit very high performance metrics in a small budget. However they have this pesky issue of requiring an electrostatic field. Conduction losses are a killer, scale up in nasty ways, and ablate material quickly.

In terms of inexpensive neutron sources: they're perhaps some of the best we have.

Reply on news.ycombinator.com