Nuclear bombs are the most economical form of power production that mankind has ever developed, bar none. This is an uncomfortable reality that society is content to ignore because nuclear bombs have become less relevant in recent decades while resistance to expansion of dualistic (having both civilian and military uses) nuclear technology has strengthened. At the current point in time, it seems like the idea of using nuclear explosions to produce commercial energy has never been less relevant, and there’s no sign of this changing.
I argue a contrarian thesis here — that we are nearing an absolute minimum throughout all of history for the relevance of nuclear explosions. Their relevance has declined from introduction up to the 21st century, but, by the 22nd century (at a bare minimum), thermonuclear explosives will be immensely important, seeing regular industrial use. This prediction hinges on the reasonable assumption that a great deal of human activities will be happening in space. The PACER concept (detonating nuclear explosions inside a sealed cavity to use as a heat source) would have good economics on Earth, there is an upper limit to how far you can scale it up. Economies of scale would taper off because of complex management challenges of the blast chamber for high-yield explosions. In space, this restriction is outright lifted because the size limits of blast chamber construction are lifted. Economics of PACER in space won’t just be good, it will be unfathomably off-the-charts kind of good.
Modern Disciples of PACER
One of my favorite thought leaders on this subject is Ron Maimon. His arguments are made in piecemeal Q&A discussion flareups. In the first exhibit here, he made glancing references to factor of 100 efficiency in scale-ups, although that’s not anywhere near to the extreme version that I’m intending to argue. A small comment here was more forthcoming with the vision, noting that megaton devices can be 99.9% fusion, and that is central to the efficiency scale-up. In that discussion, a subterranean cavity to contain a 1 MT device was seriously considered, although thought to be difficult. A minor point on Quora discussed the state of engineering, going out of the way to note that it is a “solved” problem.
On to the core thrust of his writings. This appeared in a question about fusion that was on-topic for this matter. I have to quote a good deal of this answer, because it is all quite important.
But there is a dirt-cheap way to make fusion power immediately, using proven technology. You just blow up hydrogen bombs. This project, the PACER, was proposed in Los Alamos in the 1970s, and it was immediately calculated to be cost-effective the day it was proposed, it has only gotten cheaper since. A nuclear warhead costs about $300,000 in mass production, and easily delivers a megaton of energy. You can’t buy a million tons of carbon fuels for anywhere near $300,000 dollars, and carbon fuels are less efficient kilogram per kilogram than TNT anyway. The costs don’t scale linearly, so that a 10 kT warhead also costs about the same, but try to by 10,000 tons of carbon fuels for $300,000. You can buy 1000 tons of coal for $40,000, but coal is not very energy intensive compared to TNT, so the break-even point for a PACER is around 1kT bombs, any smaller, and it is not going to be fuel efficient. This is also around the size of the smallest devices you can make.
The fuel costs for a reasonable PACER, using 10kT, 100kT, or even megaton devices, are orders of magnitude cheaper than any other fuel, even plain old Uranium for fission. Further, the fusion process in the bombs produces neutrons, which are a useful breeding resource, because they can be used to make plutonium from uranium, fissile uranium 233 from Thorium, tritium from deuterium, and many other elements, since there is an excess of neutrons, even after replenishing all the fuel that is used in the reaction. The PACER system will not run out of fuel in any forseable timespan, millions and millions of years, even with growing energy usage, and it can breed materials and reprocess its own waste. It’s really a fantastic proposal.
A real working PACER would probably use 1kT or 10kT bombs, not megaton bombs, at least at first. The 1kT bombs are more fission, they aren’t much more attractive than usual nuclear power, but already at 10kT, you can make devices that are 95% fusion, although I don’t know how the heck they did that, it’s classified neutron bomb work. For really small explosions, you can set the explosion in an artificial steel lined cavity.
This excerpt captures the essence of the concept extremely well. The idea is all about scaling, and it only has genuine merit if it can get up to a size where fusion can dominate the yield. As a deep future concept, this argument becomes vastly more powerful, and if you throw in micro-gravity, it becomes an unstoppable economic force. But let’s consider the rest of the intellectual background first, and why people are not yet thinking in this direction.
Another promoter (or at least “raiser of the question”) is Ralph Moir, who is the author of almost all the modern genuinely academic literature on the subject.
His modifications on the early concept were to make it more viable to actually build in the modern world. This included lowering the desired yield (while counting on newer technology to eek out a high fraction of fusion yield), and other things involving showers of molten salt. Seeking a realistic design, his plant’s output unsurprisingly came out to 1,000 MWe, which is right in-line with the outputs of other nuclear power plants being built at the time the papers were published.
In certain places, I believe that Maimon’s tone changed to be a little more generous to smaller yields because of Moir’s writings, although not backing away from the scaling argument. For Moir’s part, he advocated dialing down the yield, because a higher yield would have no chance of being built in the economic conditions of the 1990s.
In my view, Moir’s point is valid, but misses the forest because of the trees. No version of PACER ever had a chance on Earth of being built. As I see it, the only problem is a failure of imagination to go beyond that constraint.
PACER in Space
Blast chambers are limited by mechanical engineering and seismic constraints. The surrounding rock provides back-pressure, which is a benefit, meaning that the walls are not pressure vessel walls holding tensile strength, but just a barrier.
Physically, you can build a pressure vessel of any given size in an orbital micro-gravity environment. A slight drawback of orbital constructions is that there is no source for back-pressure, like in subterranean cavities. If you consult the mathematics of pressure vessel, you find that the structural material (steel) needed for their construction scales with the volume and pressure. The blast wave attenuation also scales with the volume and pressure, so we see that these environments lose economies of scale in terms of structural material inputs versus yield. However, this constraint isn’t highly prohibitive. Since there are few social and environmental problems with leaking radiation into the already-highly-radioactive abyss, this becomes simply a matter of steel versus output. My prediction is that the ratio is vastly superior to alternatives, such as sunlight concentration into a boiler, solar photo-voltaic, or even traditional nuclear power plants. One obvious drawback is that the scale has a minimum size needed for viability, and this size would be quite large. Instead of imagining 1 GWe plants, we might as well start imagining at 10 GWe and go up from there. This means that only mammoth industrial facilities could make use of this. That is nothing new to history. Energy transitions have always pushed the upper-limit of project scale that social institutions could sustain at first. Then, eventually, that scale becomes the norm.
At this point, I should discuss the specifics of “high” yield nuclear explosive options. Here is a great article on this subject. The idea of massive scaling-up of nuclear explosions had been seriously considered by some of the greatest minds of the original atomic age, and Teller was widely criticized for going off the deep-end. A thermonuclear explosion starts with a first “stage” that is a fission explosion, and radiation pressure from that triggers subsequent fusion stages. Exactly how many stages of fusion can be lined up, and how much yield can be gotten out of it? The answer was given by scientists in no uncertain terms — infinity. Fuel resources also don’t seem to be a problem, as Deuterium can suffice as a fuel, and there is no practical limit to its abundance.
Considering the potential of very large yields, one is tempted to lift even the requirement of the structural strength of the pressure vessel. Even if you’re off-world, you can build an underground blast chamber on the moon, as an example. However, doing this reduces to the same limitations that PACER had on Earth. At large enough sizes, the roof wants to cave in on you. It is still true that you can build a much larger blast chamber in the moon due to its 1/6th gravity and more stable geological conditions. Assuming identical conditions, you may be able to build a chamber 6³ = 216 times as large as a comparable plant on Earth, and probably get that many times the yield. Clearly this would be an economical plant, but it’s still limited at “only” 2 orders of magnitude scale-up. There’s no reason to stop here.
Problems with underground blast chamber come from the gravity gradient… so just build it inside of a large body in a place where there is no gravity gradient. To do this, you could consider large asteroids, like Sylvia, Vesta, or even Ceres, building the blast chamber in their centers. Those have gravitational binding energies and sufficient differentiation to fully contain the blast wave pressures from the rock back-pressure. This lifts the structural constraints entirely. Not only could you build a blast-chamber miles in diameter, but you could do that with absurdly high pressures. There’s no reason to atomize a flow of molten salt into the chamber (like in Moir’s design) because the gas pressure itself will be of considerably high density, and considerably good at absorbing the blast wave from the explosion. In Teller’s most absurd musings, he dreamed of 10 Gigaton TNT-equivalent bomb designs. In terms of typical energy units, that would be 12,000,000,000,000 kW-h thermal, 40 exajoules, or on the order of a trillion dollars of energy value from a single explosion in today’s accounting. What’s more, this isn’t close, not even remotely close, to the fundamental physical limits of the size of detonation that the center of large asteroids and dwarf plants could contain based on their gravitational binding energies.
Just some extremely back of the envelope calculations here, I can speculate a blast chamber filled with super-heated steam inside of Vesta or Ceres with a radius of 2 miles without must self-doubt. The 10 gigaton explosion could raise the temperature of the gas in this volume by on the order of 25 degrees Celsius. This is probably too much, but it’s within shooting range, and the general order-of-magnitude picture here stands.
Of course, as I say this, I only mean to present it in-principle. Building a blast chamber larger than necessary would not be prudent, and the ultimate constraint just becomes the construction of the radiator. We might as well approximate the fusion energy power from these explosions as being free, and there are unique optimization mathematics that arise from that situation. Radiators are kept at reasonable temperatures so as to efficiently use the energy being produced, but in this case, you use special high-temperature radiators to take advantage of the T⁴ term in the blackbody heat emission equation. What temperature would be the optimal radiator temperature in this case? That would be 3/4th of the hottest temperature that your turbines can withstand.
This allows us to compute a maximum heat rejection capability for objects in the solar system under the extreme version of this energy production method. Assume a reasonable modern value for the maximum energy a turbine can handle, use the Stefan-Boltzmann law, and the above math, and you find that a Vesta civilization could sustain a Petawatt level of useful energy production. Our doomsday-level example, Teller’s brainchild gigaton bomb, would suffice to power this society for about half of a day (11 hours in absolute terms, not a “day” of the asteroid itself). Practical considerations would seem to favor smaller detonations maybe every hour or so, simply due to complications of moving the explosive in place in the middle of the blast chamber. For reference, humans today are about a 18 TW society. I’m speaking of a single plant with a 1,000 TW output.
What about utility of a mega power plant?
Because ideal sites for these plants would be somewhat limited, we could predict an extremely uneven distribution of energy prices throughout the solar system. Thus, a valid objection is what good a Petawatt is if there’s one one there to consume it.
For starters, economic activity is specialized from region to region in the modern world anyway. Some nation’s export industries benefit from lower energy prices than other nations, and that’s not fundamentally a bad thing. An asteroid outfitted with a PACER mega-plant could be an industrial center. Even with a low domestic population, it could have an out-sized impact on the rest of the solar sytem..
On a per-manufactured-material basis, these envisioned power plants have mind-melting performance. But if such a thing were ever built, the cost of manufactured materials in its vicinity itself would drop like a rock due to the abundant energy. This would open up the dream of major industrial gains from micro-gravity manufacturing.
Another wacky use is that energy, itself, could be transported due to the numerous orders-of-magnitude economic gain compared to smaller power plants. It’s easily reasonable to envision a chemical processing plant, which loads the chemicals onto spaceships which travel to other nearby asteroids, where the chemicals are consumed in local power plants. The cheaper energy offsets the transportation costs here. The mass-economy calculations still become unreasonable for objects in far-off orbital locations.
Political Components
These technical arguments are presented with an air of inevitability. If you assume that the human demand for energy will grow exponentially, then the mega-sized PACER plants in microgravity will at some point, sooner or later, become necessary because no other energy production technology can compete with it. However, this says nothing of the “how”.
Lacking a such a concrete answer to the political and social side of the transformation, I asked about it here:
I was surprised by the lack of what seemed like the obvious (to me) answers. There was a mention of the outer solar system. In terms of how I see the top-priority points:
- A interplanetary society is less connected due to communication lag, causing more independent political institutions to form
- The outer solar system has both a greater need for nuclear power and is distanced from Earth’s influence
- Conflict between superpowers on Earth is still mediated by M.A.D., and larger bombs would be necessary if we tried to emulate this in space
- Living in space desensitizes us to the destructive power of nukes because orbital velocities allow for kinetic weapons of similar scale anyway
Without understanding the specifics of how this will happen, I want to re-emphasize the point that this is a political question. I think this quote summed it up quite good.
From Ron Maimon:
Part of the problem was the fact that these power plants are simultaneously weapons-testing facilities, and the idea of powering the world with them would be a proliferation nightmare — — every nation on Earth would be clamoring for hydrogen bombs to light their cities.
Modern light-water commercial nuclear reactors have been able to firewall themselves from weapons programs. This will not be the case for PACER by any stretch of the imagination. The ability to produce peaceful energy at these scales is technically indistinguishable from the ownership of weapons of mass destruction.
State of Advocacy… or Alarm?
The Project Orion (using nuclear bombs to propel a spacecraft) is comparatively well known. People who are into the idea aren’t necessarily “advocating” it. Part of the interest is purely retro-futurism curiosity, and maybe part of it is morbid curiosity. Another valid reason to talk about it, as noted in this TED Talk, is that we might need the capability if something sufficiently urgent threatens life on Earth.
The PACER project is like a younger half-brother to that concept, but vastly more reasonable. While Orion would necessarily detonate nuclear explosions in open-atmosphere, there’s a fairly valid claim that a PACER plant could contain the radiation it produces. What’s more, it could make a valid claim to being able to minimize the quantity of radiation produced in the first place because of a looser requirements on the design of the explosive and the more highly-controlled conditions they explode in. PACER can also keep a low inventory of actively usable explosives with just-in-time fabrication.
In order to make the argument for PACER in space, I see two possible spins. One, is that we need economic growth in order to maintain stability of our industry, and by extension, of society. Two, is that a responsible program of civilian nuclear explosions might actually help prevent nuclear war.
On the first point, we need to be cognizant that there are serious time-sensitive problems that we need greater capability to solve. Even if we were able to eliminate carbon-based emissions within the 21st century, there is a very good argument that geoengineering will be necessary in order to constrain the warming caused by the lingering CO2 gas already in the atmosphere. If we don’t, we will lose parts of Earth’s geological, ecological, and biological record which can never be recovered. Also, inequality is widely thought to be exasperated by low-growth conditions and lessened by high-growth conditions. If economic egalitarianism will help us get through our technological adolescence alive (a reasonable claim), then it follows that nuclear explosions may be a “good” technology, given a different historical context.
On the second point, there are factors that argue in favor of a re-emergence of nuclear weapons that have nothing to do with PACER. Even if we are able to ignore the tremendous economic benefit in the coming centuries (which I recognize may be possible), there may be an unrelated re-emergence of nuclear explosives in a purely military sense. In that scenario, the interplanetary international community would be woefully unequipped to institute appropriate controls.
Both of these strike close to the heart of the Doomsday Argument. The argument states that the absence of other intelligent life (as per the Fermi Paradox) likely indicates the advanced civilizations have a tendency to kill themselves. Humans now find themselves in a lull of a sort, where both technology and frontier expansion has leveled off — but perhaps not for long. If the darkest terrors of the cold war are going to return to haunt us, then we best dedicate our best brainpower to solving the moral challenges of that technology transition.
I feel a certain chill writing about this topic, because it’s not as irrelevant or as playful as the Orion nuclear rocket idea. It is central to our economic backbone, and to our destiny as a species.
