Book Review: The Future of Fusion Energy | Hacker News Reader

Book Review: The Future of Fusion Energy

martin.kleppmann.com

55 pointsbschne37 comments

The problem is that even if they were to get it working, it would still be several times as expensive to operate as the equivalent fission plant, which is not now competitive, and gets less so each year. So, at this stage there is no expectation of practical energy output, and thus only academic interest.

Why? Fission is expensive partly because of the safety requirements- and for good reason. Fusion is far safer.

Fission is expensive because of the ungodly amount of red tape that needs to be cut before groudbreaking can happen.

Fusion will face much the same amount of red tape without changes to the law. Doesn't matter that it is safer, it's "nuclear" which is what most laws target.

Seems unlikely that if we had working fusion, unnecessary red tape owing to both fusion and fission being “nuclear” would be the main stumbling block. That seems like something that would be easily fixed.

In principle it should be easy, in practice (in the US at least) it's nearly impossible. The US senate is currently in nearly total gridlock. Passing new laws is crazy hard.

On the one hand you're right, but OTOH I have to imagine "super cheap energy!" would hopefully change some minds quickly...

Those minds are currently being heavily encouraged to keep the money train going on resource extraction. Not clear why that would change, where's the "cheap energy" lobby going to come from exactly?

Everyone paying electric bills.

Sadly voters don't make for much of a lobby these days.

They used to promote fission power as "too cheap to meter". They were lying then, too, and knew it.

There is no physical possibility of Tokamak fusion power being cheaper than the most expensive method in current use anywhere.

The UK has already announced regulations specific to fusion, taking into account the expected low hazard of fusion reactors.

https://www.gov.uk/government/news/uk-set-to-be-first-countr...

And I'd be willing to bet that China won't put up unnecessary barriers.

Fission is relatively expensive globally, even China has major cost issues. It’s not a question of red tape, people really did try to operate nuclear without massive safety systems for example, that repeatedly failed.

In one case the US army had people manually lifting control rods on a dead simple reactor, someone messed up and the SL-1 melted down killing 3 people. Dig into thing and you can find well over a dozen reactors that melted down and many others that suffered significant damage from one mistake or another. So, now everyone builds and maintains redundant safety systems, but that simply isn’t free.

It isn’t any one thing keeping costs high. I was personally hopeful the Advanced CANDU reactor would drive down costs by avoiding expensive enrichment, but apparently it was overall more expensive.

Enrichment became much cheaper with centrifuges (50x less energy needed than gaseous diffusion.)

Less energy, but no where near 1/50th the cost. Centrifuges simply don’t process that much material per year so you need a lot of them to supply a nuclear reactor.

PS: Number I saw per pound was between 1/8th and 1/30th the energy. Though that may not be up to date.

Fission is expensive because it's complex and must be highly reliable.

Fusion will also be complex and must also be highly reliable -- not because of public safety, but because repairing a fusion reactor will be very difficult. It will be far too radioactive for contact maintenance.

A Tokamak fusion reactor would destroy itself in short order: the superconducting magnet windings are blasted with hot neutrons whenever it is operating. The liquified metal neutron-absorbing "blanket" is in pipes that are being blasted with hot neutrons, and will weaken and need to be replaced, but they have become seriously radioactive, so you have to replace them using robots. Which will become radioactive, too. You are processing the molten stuff to refine out tritium you have been breeding in it, that you need for fuel. The power density is very low, so you need an absolutely monstrous machine to get as much power out as from a lump of uranium.

Because it has to be so big, it has to be very expensive, several times as much as a fission plant. And, take much longer to build. That much money with no expectation for any power delivered for many years is catnip for the corrupt, as for any big-ticket public works item where the costs of components are far from obvious. (Cf. NASA STS and SLS, Boston's "Big Dig", California's "Bullet Train to Nowhere", F-35, New York's 2nd Ave subway.)

What is the practical advantage of fusion power over newer fission reactor designs? If we desire nuclear energy, why not just build out, say, a bunch of molten salt reactors?

It seems that the primary reasons we don't see more nuclear power are 1) it's highly capital intensive (big upfront investment for uncertain reward), and 2) it's expensive to comply with safety regulations. How does fusion solve these problems?

Much, much less long-term radioactive waste.

It's a lot more energy, for less fuel, but that may be mitigated by the cost of the reactor.

There really isn't an overall practical advantage, at least for DT fusion, and this has been known for decades.

http://orcutt.net/weblog/wp-content/uploads/2015/08/The-Trou...

> Fusion will almost certainly have a lower power density than fission and therefore will require a larger plant to produce the same output. Suppose a fusion plant had to be ten times as big and therefore likely ten times as costly — as a present-day fission plant to produce the same amount of power.

I don't see how that follows. Different forms of power generation have orders of magnitude differences in costs on a per volume basis. If you're comparing within a type it might make sense, a coal plant 10 times larger probably does cost between 5 and 20 times more. But I don't see any reason to expect a fusion plant to have comparable per volume costs to a light water fission plant. And why is he assuming the cost is per volume rather than per mass?

It's because aside from the "nuclear steam supply system", a DT fusion power plant and a fission power plant would be the same. The non-nuclear parts will be similar (or cheaper for fission; a fusion plant needs a cryo plant and a tritium recovery system.)

A fusion reactor will likely be MORE expensive per unit mass (or volume) than a fission reactor. A fission reactor is a rather simple thing, operated at relatively low thermal and neutron loading.

Fusion's only real chance would be if some of the non-nuclear parts could be avoided, for example by direct conversion of plasma energy to electric power (so no or smaller turbines/generators). This would require advanced fuels (like DD or D3He).

There is no practical advantage of fusion power over anything, because it has no practicality of any kind.

We don't see more fission power because whenever a new fission project is started, the majority of the money is siphoned off for political patronage. Construction stretches out over a decade, because once it is finished, the money stops flowing; nobody actually involved wants that ever to happen. If it must happen, it is somewhat better to have a working power plant at the end, but with enough money involved, that is dispensable, as we see in South Carolina.

This contrasts with solar and wind projects. They can start producing almost immediately. It is easy to multiply the cost of a unit of generating capacity (panel, wind turbine) times the number of units, and see what it should cost. There is just too little scope for corruption. (That might be solved, in time.) So, costs keep going down, year over year, driven by healthy competition.

> We don't see more fission power because whenever a new fission project is started, the majority of the money is siphoned off for political patronage.

However bad an over budget nuclear plant is, an under budget one would be worse.

When the extra money is not actually spent on the reactor, but is just handed around in exchange for political favors, does that really make the plant seem safer to you? (My mother used to say getting pawed at the airport made her feel safer, so I guess almost anything is plausible.)

Safest are the ones that get cancelled without being finished. Which is the norm, lately. Nobody seems to ask for the money back when that happens; it must have been money well spent.

Nobody have present any evidence of any significant amount of nuclear power plant funding "being exchanged for political favors".

Note, we'd need both evidence of it happening at all, then evidence that it is endemic in nuclear power plant commissioning to explain that extremely poor economic performance of nuclear power plants.

> Safest are the ones that get cancelled without being finished.

We are in agreement here!

It should suffice to note that they invariably cost at least 3x the price quoted (if not 6x) and take 2x as long to complete (if not 3x).

Extremely complex and potentially catastrophically dangerous projects are prone to overruns, you don't say.

It's strange that solar plants in general don't seem to have this issue, I guess all the people involved in solar plants are just honest people.

Projects that seem extremely comples and potentially catastrophic are easy to arrange overruns in, if that is what you're after.

It is hard to invent bogus costs in solar installations, or excuses to stall delivery. They will probably solve that eventually. But solar projects do also seem more likely to have idealists involved.

> Projects that seem extremely comples and potentially catastrophic are easy to arrange overruns in, if that is what you're after.

No, that's what you are after. And after asserting that there is some sort of worldwide conspiracy against nuclear power plants in the form of financial sabotage and graft, you have provided exactly zero evidence for your claim.

There is no "worldwide conspiracy against nuclear power plants".

(1) It does not violate laws--it is all wholly legal, from beginning to end--so, no "conspiracy".

(2) It is mainly in the US, so not "worldwide". And,

(3) It is not "against nuclear power": nukes are the gravy train, how could they be against nukes? They love nukes, and want us to love them; just not enough so to complete them.

Safety is a lot easier to achieve with fusion because it stops producing power instantly when containment fails and the amount of energy in the plasma isn't that high. It's hard to make a fission reactor both large enough for commercial scale and safe without cooling.

The second benefit is not producing high level radioactive waste. With fission you can't really get away from the decay products of your fissile materials, they're intrinsic to the process. With DT fusion you get He4 which is stable and neutrons. The neutrons are dangerous but will either quickly decay or be absorbed by something. By controlling what gets exposed to the neutron flux you can in principle achieve arbitrarily low levels of radioactive waste. That's a serious engineering challenge to get as low as possible but my understanding is that it's not hard to produce much less than with fission. You also want to produce Tritium which will later get consumed in the fusion process leaving only a moderate amount stored at any time so if things go well its not a long term problem. If the storage ruptures in a gentle then it escapes high up into the atmosphere. In the event of a containment breach you're probably going to get super heavy water steam which is an issue.

The regulatory framework is a real concern, depending on how many rules carry over from fission reactors. Back in the 1970s the term "fallout" carried connotations of people dying painfully in hours after exposure which is a different association than people born after the Cold War have. Also electricity generation back then was done on a cost plus basis so the people running the fission plants had every incentive that their costs were increased as much as possible. That isn't how we run our electricity grids any more so there aren't strong forces on the industry side fighting for tighter regulations.

> It's hard to make a fission reactor both large enough for commercial scale and safe without cooling.

Gen IV reactors all fit that bill. They all have passive safety properties (so, cut the power and the plant shuts down, no nuclear meltdowns) and they have very large outputs (2GW in the largest plants).

> The second benefit is not producing high level radioactive waste. With fission you can't really get away from the decay products of your fissile materials, they're intrinsic to the process.

Again, Gen IV reactors are much better than previous reactors in this aspect. While they do produce radio active waste, it's generally far more used (producing less waste) and radioactive for a far shorter time (hundreds of years vs 1000s).

> The regulatory framework is a real concern, depending on how many rules carry over from fission reactors.

This is the big sticking point that hurts the bottom line of fission reactors and will hurt fusion reactors. Without changes to the laws, nuclear power plants are going to have a rough time turning a profit. They mostly aren't unprofitable because of construction cost, but rather regulatory overhead.

It takes over 10 years to construct and turn on new reactors in the US, which is a major problem.

Only 10? Do you have a source for that? I seriously thought it’s more like 20.

You are probably right. I'm relying on my fuzzy memory for that number (I've looked it up in the past).

This article [1] says 5 years to license, 5->7 years to build, and then the verification time with the electric company before it can be switched on.

[1] http://large.stanford.edu/courses/2017/ph241/park-k2/

The fuel production does not deal with heavy metal mining and then processing. The reactor itself is save in the sense that the worst thing that can happen is, that you scorch the reactor wall, because there is just not enough energy in the reactor at any given time. And if there is something going wrong then you're dealing with a nuclear inventory of a few grams hydrogen that will quickly oxidize into heavy water, instead of a thousand tonnes of fuel that will happily melt under its own few hundred megawatts of decay heat. And the end product of the fusion reaction is helium, which is chemically inert and not radioactive. Instead most of the radioactive waste are the reactor walls, which are activated iron and therefore have a half live of 2.5 yrs, so you have to keep the radioactive waste for something like 250 years save instead of for 250 000 years, and it is iron, not a complex mix of heavy metals and gases that over time changes it's composition.

[PS:] Forgot the proliferation aspect. There is actually not that much difference between a civilian and military nuclear program. By contrast the main proliferation problem of fusion is, that it deals with a lot of tritium and it is straightforward to build a tritium booster for a bomb, however for that you need a bomb first.

If the half life is 2.5y, you probably don't even have to keep it 25y. 100 half lives is completely ridiculous.

I’ve read this book and it’s really good - interesting as well as entertaining. I usually don’t like popular science but the authors here do a great job of presenting complex ideas clearly along with their opinions.

Reply on news.ycombinator.com