Hacker News Reader

Comment by ncmncm | Hacker News Reader

There will be no earthbound, commercial Tokamak fusion power in the lifetime of anyone reading this. For reasons.

The energy density of the plasma is necessarily so low that a useful power plant would be absolutely monstrous. Then, it would destroy itself with neutron flux in short order, leaving thousands of tons of radioactive scrap. There is no possibility of such a plant being competitive with solar and wind at current prices, and they are still on their way down.

The prospect of commercial power generation is not the driver of fusion research. Rather, it is a jobs program for high neutron flux physicists, to maintain a population available to draw upon for weapons work. So, expect it to continue burning $billions indefinitely, with no deliverable.

View on HN

The numbers I've seen thrown around for the flux expected out of ITER are containable with existing material science, and it's a very conservative design which is modeled to generate 10x the input power. I'd call that a deliverable.

Note that ITER's output is heat, rather than electricity. Maybe the heat will be useful, but it's an important distinction.

ITER's volumetric thermal power density (volume of reactor + plasma) will be 50 kW/m^3, some 400x worse than the volumetric power density of a commercial PWR (reactor vessel + core).

ITER also use a CuZr alloy backing up its first wall materials. This material is unsuitable for long neutron exposure. DEMO would have to use RAFM steel instead there, with an order of magnitude lower thermal conductivity. As a result, the power/area at the surface of the first wall in DEMO may be even lower than in ITER.

The first wall has an 8 mm layer of Be facing the plasma.

Yes, and it is backed up by CuZr. Replacing that CuZr with steel will significantly impede cooling.

Only replying to this low-effort response because it is currently the top response.

> The energy density of the plasma is necessarily so low that a useful power plant would be absolutely monstrous.

The size of e.g. the SPARC proposal is so low that you could fit on the order of 10 units of 50 MW within the tower of one 10 MW wind turbine [0].

> ... Rather, it is a jobs program for high neutron flux physicists ...

You do realize that many current effort in fusion power are privately funded, right?

[0]: https://www.cambridge.org/core/journals/journal-of-plasma-ph...

your link has 404 status

The power density of the ARC proposal (volume of the magnets and everything inboard of that) is some 40x worse than the volumetric power density of a commercial PWR (volume of the primary reactor vessel + core). This is much better than ITER, but still much worse than existing fission reactors.

I'd characterize as "low effort" the shallow dismissal of very real issues facing fusion.

PWRs are some of the highest power-density energy sources that exist, though (which is why they're so popular on submarines). 40x less than a PWR is not low.

I agree that ITER is an overwrought boondoggle, however.

Fusion and fission plants will be similar outside the nuclear island. So if you take a small fission reactor, and replace it with a much larger, much more complex, fusion reactor, you're going to increase the cost of your power plant.

How’s it fare against a gas turbine power plant? Because that’s what it would actually replace, not a PWR.

Since a fusion or fission reactor is a source of thermal energy, and the power density there does not include the volume of external heat exchangers or turbines, the proper comparison is to the COMBUSTOR of a gas turbine. Gas turbine combustors have thermal power densities of 100 MW/m^3 or higher, some five times the power density of a PWR primary reactor vessel and 2000x the power density of ITER.

Gas turbines also have large advantages in the rest of the plant, though. Fusion plants or PWRs will require at least two sets of heat exchangers (one for transfering heat to the external cooling water, and one for isolating the nuclear coolant from the fluid going through the turbines). Simple cycle gas turbines (without regeneration) don't require any heat exchangers, and combined cycle plants get most of their power from the topping cycle, so the heat exchangers on the bottoming cycle are smaller.

The cost of a simple cycle turbine power plant is around $400/W, maybe 5% of the capital cost of a fission power plant.

$400/kW, I meant

Not an expert at all but this contradicts everything else I've seen on the subject, could you support your claims? I thought there recently were advances in superconductors that would enable way smaller reactor designs?[0]

[0]: Dr. Dennis Whyte. - Small, modular & economically attractive fusion enabled by high-field superconductors - https://www.youtube.com/watch?v=rY6U4wB-oYM

Better superconductors allow reactors to be made smaller, up to a point. But they still would have lousy power density. ARC's power density (total reactor volume, not just plasma volume) is 40x worse than a commercial PWR.

Fusion reactor power is ultimately limited by the need to radiate energy through the wall of the reactor, and the limit from that would apply even if plasma confinement were a nonissue.

Turns out we just needed stronger magnetic fields: https://cfs.energy/

Materials science delivered an order of magnitude more powerful magnetic field in the form of better superconducting ribbons. Add in molten salt blanket for neutron recovery and heat working fluid, design for maintenance: and voìla!

You have a garage sized fusion reactor which outputs net energy.

Imagine you have installed one of these things in your garage, or maybe better 10 or 20 ft below, to keep from sterilizing your offspring. You got it for a fire-sale bargain at only $50M, even including the turbine and cooling tower. It produces, what, 100kW? (Plus 100kW of waste heat you might be able find a use for.) Or is that too optimistic?

How many MW of solar panels can you buy for that $50M? Let's even figure in plenty of battery backup. (Hint: lots.) How many MW of wind turbines? (Hint: lots.)

Where would such a garage-sized fusion reactor be useful? And how long before it is a radioactive slag heap? (It was good planning to bury it up-front.) How many GW-years does the fleet need to produce before it pays off the tens of $billions already sunk, to date?

The only guaranteed product of all nuclear power schemes is monstrous exaggeration of cost effectiveness.

"Outputs net energy" is very, very far short of useful.

You seem to be comparing construction costs of an experimental project to the LCOE (basically, utility rates) for a massively scaled and mature industry. Let's ignore the R&D costs it took us to get solar to its current state. According to https://www.eia.gov/todayinenergy/detail.php?id=36813, the construction costs in 2016 for utility-scale solar plants were ~$2.5k per kilowatt, or ~$2.5M per megawatt.

> How many MW of solar panels can you buy for that $50M?

About 20MW. Definitely not 1000GW, which is what you would get by dividing $50, today's approx. solar power MWh LCOE, into $50M. From the tone of your reply it seems like you were assuming the latter.

Ignoring large jumps in the price of land, and ignoring potential taxes for their environmental impact (deserts and plains have ecosystems, too), solar and wind will likely be cheaper than fusion far into the future, even by optimistic fusion predictions. But it's still debatable whether it'll be so expensive that we shouldn't even bother making the investment. The LCOE of natural gas is 3x that of solar power today, and yet we're still building natural gas plants like crazy.

(FWIW, the LCOE for new fission plants as expensive as Vogtle Units 3 & 4, the former coming online next year, is comparable to natural gas. https://www.lazard.com/media/450784/lazards-levelized-cost-o... In China the LCOE seems to be closer to that of contemporary solar LCOEs. https://www.mdpi.com/2071-1050/10/6/2086/pdf Which explains why China has many more reactors under construction.)

I will only note that 20 megawatts is always "lots" more than 100 kilowatts. Even 10MW is. And if you think you will be able to buy a garage-sized fusion reactor for $50M in our grandchildrens' lifetime... good luck.

One might be useful onboard a spacecraft, at the end of a long, long tether. But a pB reactor would still be a better choice.

China is also building new coal plants. The actions of a command economy are not a sound indicator of relative cost and value.

Reply on news.ycombinator.com