Comment by carterschonwald | Hacker News Reader

Comment by carterschonwald | Hacker News Reader

Parent Full thread

carterschonwald

This is actually a pretty reasonable idea. There are several considerations that need to be worked out first. Specifically:

1) there are few programs in the entire US which have nuclear engineering degrees (MIT has one, I believe fewer than 10 others do too), a topic requiring a sufficiently broad range of skills that its almost like a fusion [sic] of solid state physics (for robust materials), electrohydrodynamics (to understand anything relating to fusion and plasma), numerical computing (analytic solutions to realistic PDEs is a funny idea sometimes), as well complete case analysis of all possible risk and failure modes via probabilistic model building (there is in fact dedicated software used to handle the complexity of these models). In short, do we have the educational capacity to train the people needed?

2) at least with how current nuclear regulation is structured (which would presumably have to change), technically any new reactor designs that get approved are "exemptions" to a general ban, so theres actually a lot of time spent finaggling approval via a somewhat insidery process.

3) I want fusion thats commercially viable! :). The idea of having energy become functionally free and what would then be possible boggles my mind, and I don't think anyone can fully comprehend the myriad subtle ways in which the world would immediately change. Also, while in practice dealing with fission products/waste isn't that bad (and its much more contained than waste from conventional power plants), it is logistically tricky to manage both the waste and all the people who worry about the waste.

Carter,

Where did you get the idea that fusion power is functionally free?

The fuel you burn is nearly free, but you also use up a ton of material and make it radioactive so it's hard to recycle. The structural and shielding elements of the reactor will likely degrade (for we haven't tested this) on a similar timescale to the degradation of wind turbines, solar collectors, and the other components in the balance of the plant.

The heat from the sun and at the core of the earth is functionally free, in the same sense: the question is how expensive it is to get it out! I've never seen a really reasonable and measured study of what commercial fusion might cost to produce power.

Indeed. Plus the thing about fusion power is that likely these plants will be absolutely massive, which raises issues about: * single point of failure * huge transmission needs * very large start-up cost, likely banning developing countries from using it.

All of these issues are much alleviated with distributed power like solar or wind (or the mini-reactors suggested by this article), which makes me think it's preferable.

One of the encouraging things about IEC fusion and dense plasma focus fusion is that they're supposed to be something you can do with smaller facilities, and construct incrementally. What you said certainly applies to more traditional fusion methods, though. We've already seen the economic problems of huge power plants with recent nuclear reactors: they're so darn big nowadays that financing them is a huge headache.

This is part of why I'm moderately excited about aneutronic fusion: the radiation damage you mention will be greatly lessened by reactions that release dramatically fewer neutrons.

As for the commercial cost of fusion, the reason you've never seen such a study is because nobody even knows yet how to get fusion to produce commercially usable amounts of power, and there are so many possible methods being looked at, with such different costs involved, that we can't really do a meaningful study of fusion economics yet.

I was working on aneutronic fusion for some time: my feeling is that there is a lot of research to go before we can know if it's even feasible; even then unless there are breakthroughs it will be difficult to do in small reactors. There are significant thermodynamic barriers to be overcome e.g. Todd Rider's thesis (I believe they can but they're difficult -- almost like no-go theorems for fusion).

As for whether one could do an economic study, I believe at least one could lower bound the cost, using the cost of the balance of the plant (steam generator, heat exchangers, cooling towers) and an expected cost and lifetime of the various structural and shielding materials.

Regarding #1 -- "educational capacity to train the people needed".

One of the additional benefits to source article's proposal: There's a large, highly trained base of Navy 'nukes' who are in the fleet now or out in the world. Many of the ex-nukes are building and operating non-nuke plants now. These are the folks who spent years (decades) training and operating the same scale of power plants in the proposal (on subs and surface ships).

While university nuke-e programs are light, there's no shortage of folks to build, manage and do the work.

Professor B. is on to something.

Regarding your first point, I believe the current dearth of programs is a function of lack of demand: When I was at Georgia Tech, we had a reactor on campus and still only had 3 Nuclear Engineering students.

There is demand from industry, but not from students. In my class at NCSU I was the only one; the years before and after there were less than 10.

It's still a small department at NCSU, but not that small (so I've been told). I know some guys graduating this year and looking at graduate opportunities.

OT: Hello fellow Wolfpack-er. I think there are a couple other NCSU grads on HN, too.

I'd rather see liquid flouride thorium reactors than fusion. They're much more viable (we've already got the tech - they had a working one back in the '60s), cheaper (we've got thousands of tons of thorium sitting buried in the desert ready to go), safer (meltdown-proof by design), don't contribute to proliferation (byproducts cannot be used to create nuclear weapons), and reduce waste (the waste is safe in tens of years instead of tens of thousands of years).

We still don't have a working man-made prototype for a fusion power reactor.

Reply on news.ycombinator.com