Fusion Foolery (NIF conversion facts)
(...)
In the end, the NIF fusion accomplishment might be called a stunt. Stunts explore what we can do (often after an insane amount of preparation, practice, and failure), rather than what’s practical. Stunts hide the pains and present an appearance of ease and grace, but it’s a show.
(snip)
A very few articles mentioned the energetic price of generating the laser pulse. In particular, I found one in The Atlantic by Charles Seife:
The “more energy out than laser energy in” equation masks several fundamental problems. NIF’s doped glass lasers have an efficiency of about 0.5 percent, meaning that they would have sucked in roughly 400 megajoules of energy from the grid in order to produce the 2.1 megajoules of light energy…
The second was a Big Think article by Tom Hartsfield.
The laser energy delivered to the target was 2.05 MJ, and the fusion output was likely about 3.15 MJ. According to multiple sources on NIF’s website, the input energy to the laser system is somewhere between 384 and 400 MJ.
And that’s just the laser energetics. The whole facility consumes scads more for countless other purposes. According to the LLNL NIF FAQs (are you letting me get away with a triple acronym?),
NIF’s 192 powerful laser beams, housed in a 10-story building the size of 3 football fields, can deliver more than 2 million joules of ultraviolet laser energy in billionth-of-a-second pulses onto a target about the size of a pencil eraser.
The emphasis is mine, to highlight the point that this is a massive laser and facility. It’s like ten Walmart superstores stacked on top of each other. The lighting alone is likely taking tens of kilowatts, which could hypothetically be run for less than a minute on the energy gain from the fusion pop. It would be fun to count all the megajoules that went into press coverage of the event!
Power Plant Energetics
Let’s connect the 3 MJ output to that of actual power plants, forgetting for a moment the tremendous energy loss represented in getting 3 MJ out from a 400 MJ input. A typical electrical power plant (nuclear, coal, etc.) delivers about 1 GW of electrical power. But it’s a heat engine operating at 30–40% thermodynamic efficiency. So it takes roughly 3 GW of thermal energy to export 1 GW as electricity. 3 GW is 3 GJ per second, or 3,000 MJ per second.
The same efficiency factor would apply to a putative fusion plant. The concept behind fusion power is that it’s just another thermal source—an excruciatingly elaborate way to boil water to make steam to drive a turbine to run a generator. So our 3 MJ would need to be replicated 1,000 times per second to amount to 3 GW.
Laser repetition rates can be all over the map. 1,000 Hz is not in itself unusually fast by any stretch. What is the repetition rate of the NIF laser? Handily, LLNL provides these statistics. The average since 2015 is 377 shots per year, with a high of 417 and a low of 327. That’s about a shot per day—or two on a good day. It’s only 100 million times shy of 1 kHz. Oh dear.
Economics
An interview of physicist Bob Rosner in the Bulletin of Atomic Scientists helpfully puts the NIF in context (it’s not about societal energy). In it, he reinforces some of what we’ve covered, and adds some financial detail.
This facility can do one shot a day; this is at slightly more than two megajoules (of output). For an energy source, it would have to do the same thing at least 10 times a second. If you ask, “Do the lasers exist that can do this?” Not in your dream. The pellet cost a bit over $100,000 to manufacture.
The 10 shots per second, I gather, is if the fusion yield could be improved by a couple orders of magnitude—approaching actual break-even. At $100,000 per (literal) pop, and even just ten shots per second, we’re talking a cool million dollars per second!
Let’s wave a magic wand for a minute and say that the 400 MJ input produced a 700 MJ output for a net of 300 MJ: 100 times the recent breakthrough. This accords with the ten shots per second mentioned above. What is the price of the delivered electricity? After thermodynamic inefficiency is accounted, we get 100 MJ out for $100,000 cost, or $1,000 per megajoule. We are accustomed to using the kilowatt-hour (kWh) as a measure of delivered energy, which is 3.6 MJ. The cost becomes, then, $3,600 per kWh. Typical electricity costs are in the neighborhood of $0.15–0.20 per kWh, so we’re dealing with a cost that is 20,000 times higher than nominal. And don’t forget, we used a magic wand to even get there. It’s closer to 2 million times more expensive currently, and as a net energy loser to boot.
This massive reduction, incidentally, translates to a cost of $5 per pellet. I don’t care what mass-production slave labor you might dream of employing. A cryogenic hydrogen-ice target made to demanding precision specifications, containing deuterium and transmuted lithium (to make tritium) is not going to cost $5. You lost me at cryogenic. Also, they would have made many pellets by now and I’m sure don’t relish spending $100,000 each. If they’re clever enough to accomplish fusion, they would be clever enough to have already reduced costs dramatically if it were straightforward.
(more)
The problem is that such imaginings are not tethered to physical reality. They are driven by ideology, or I would say mythology. The physical reality is that we are living in an ecologically, evolutionarily untested paradigm that is very recent (on relevant timescales) and powered by patently unsustainable practices and resource use. The cost is rapid ecological degradation and global disruption to the biosphere. It seems quite clear that the track we are on does not lead to the stars, but to ignominious self-termination of this whacky mode called modernity. It simply does not add up, once the mythology is stripped away. The venture capitalist of nature is about to slam the door on our faces."
https://dothemath.ucsd.edu/2023/08/fusion-foolery/(ooh, author has a pdf book for UCSD avail)
Energy and Human Ambitions on a Finite Planet
https://escholarship.org/uc/energy_ambitions