The Star Builders

by Arthur Turrell

  When the costs of existing experimental reactors are spread over the world’s population, they’re minor. The going rate of around $10 billion is about the same as a US aircraft carrier. The first commercial reactor would be in the same ballpark, with subsequent reactors much cheaper. Compared to the USA’s total spend on energy of more than $1,000 billion each year, the cost of fusion reactors is small.31 It’s positively tiny compared to the cost of America’s war in Iraq, which is in the multiple thousands of billions.

  But these are somewhat silly comparisons. It’s more sensible to ask whether funding for fusion is enough for it to happen on a reasonable timescale. It hasn’t been, at least historically—and we’ve known that for a long time.

  In the 1970s, the US Energy Research and Development Administration put together estimates of how quickly fusion might be achieved under different levels of annual funding. They estimated that the minimum level of funding that would see commercialization of fusion ever happen was $2 billion per year. From 1976 to 2012, the average US level of funding for fusion energy sciences was roughly $0.6 billion per year for magnetic and laser fusion combined. That’s well under the minimum that it was thought might be required to ever reach a prototype fusion reactor, whether via tokamaks or giant lasers.32

  Funding of around $0.6 billion per annum (which has also been the allocation in more recent years) may sound like a lot, but in 2018 it was only 0.1 percent of the US’s research and development budget and less than what big oil and gas firms spend individually per year on research and development.33

  Worldwide, by far the largest fusion expenditure is on ITER, which has totaled around $1.7 billion per year over its long gestation period. Again, that sounds like a lot, but it’s only around a tenth of 1 percent of the global spend on research and development.

  Moreover, it’s good to remember that 9 million people a year are dying from air pollution and 150,000 or so are already dying from climate change–related phenomena such as drought, floods, and fires. Fusion comes with risks because—despite what some star builders say—it’s impossible to know for certain whether it will ever be commercially viable. But the potential upsides of eliminating those unnecessary deaths seem worth it.

  For context, consider the fact that the world spends roughly $20 billion per year on research into cancer, which is thought to be responsible for 10 million deaths a year.34 Cancer research today can probably translate into lives saved or improved much more quickly than fusion research (because the latter will only be helpful if it results in technology that is commercially viable), so it’s certainly not a perfect comparison. What the numbers do highlight, though, is that fusion is currently less well funded than you might expect given its potential.

  People who complain that fusion is taking a long time should be aware that money is a large part of the reason. Investment is key to speeding up or slowing down progress. And we’ve known since the 1970s that investment levels in fusion are too low to make rapid progress. It’s not that net energy gain isn’t possible—it’s just that we haven’t wanted it badly enough. If we really want scientists to save the planet by building a star, someone’s going to have to pay for it.

  The funding situation is changing. “Five years ago no one in government spoke about fusion,” Ian Chapman told me. “Now, when I go into the Treasury, people want to have discussions about it.” As he also noted, China has a much bigger appetite for risk and funding when it comes to fusion, and their plans suggest they may set the future pace of global research. As for the money from private investors that is now pouring into fusion, some of it is being directed toward projects with very long odds, but as star builder David Kingham put it, private sector fusion grows the pie. To make fusion deliver in time, it’s all going to be needed.

  At the risk of getting ahead of ourselves, it’s interesting to speculate on how expensive fusion energy will be for consumers if and when it’s commercialized. To spotlight again Ian Chapman’s fifth challenge, an energy source that is perfect in every other way but is one hundred times more expensive than anything else isn’t going to achieve much in the way of market penetration. To compare the costs of different types of power, the LCOE (levelized cost of electricity) is often used. It’s calculated by comparing the costs of a power plant with the entire amount of energy that the plant will produce over its lifetime. It’s usually measured in dollars per amount of energy produced. For instance, based on a number of countries, the International Energy Agency estimates that a conventional coal power plant has an estimated LCOE of $24 per gigajoule (a gigajoule is a thousand megajoules). It’s a useful way to think about how expensive energy sources are.

  It’s hard to know how expensive fusion energy will be because there are so many unanswered questions. Will fusion plants have different regulatory requirements from fission plants? How big will a plant be? How long will a reactor chamber last? How much downtime will there be? Any estimate of the cost of fusion has to make heroic assumptions to get to a number. A recent US Department of Energy report on the likely costs of nuclear technologies refrained from calculating an LCOE precisely because of this uncertainty—and even the LCOE values for established technologies are significantly different depending on the assumptions used. That’s not to say people haven’t tried, but any estimates should be taken with a grain of salt.

  One of the best guides to the likely cost of fusion power is the cost of fission power. Fission power plants are a good comparison because they’re also a nuclear technology and would probably need a similar scale of facility—though the regulations for fusion are likely to be significantly less stringent than they are for fission. The cost of fission power is an upper bound for fusion given that the overall costs of fission include dealing with long-lived radioactive waste and risks like meltdown. What is similar between fission and fusion is that much of the expense is likely to be in the form of capital costs covering construction, commissioning, and financing (the fuel only makes up a small part of the LCOE for nuclear technologies). The International Energy Agency put the price of advanced nuclear fission reactors at $19 per gigajoule, cheaper than conventional coal.

  As part of the designs for the prototype fusion power plants, DEMO for magnetic fusion and LIFE for inertial fusion, cost-of-electricity estimates were made. For a successor to DEMO, cost is reckoned at $21 to $45 per gigajoule. LIFE came in a bit lower, on the high end at $21 to $30 per gigajoule. Nick Hawker has published detailed estimates of cost ranges for his more modular machine that suggest $30 per gigajoule is likely but that as low as $7 per gigajoule could be possible.35 Twenty to thirty dollars per gigajoule isn’t too far away from the estimates for fission, so seems plausible. But, honestly, no one really knows. What we can say is that it’s unlikely that fusion will have an LCOE that is dramatically less than the LCOE of nuclear fission, unless star builders find a way to substantially shrink the size of the plant required, via new technologies such as smaller lasers or better superconducting magnets.36

  What about other power sources?I The cheapest fossil fuel is the combined gas cycle turbine at $18 per gigajoule, but that cost doesn’t take into account the negative externalities of air pollution and carbon dioxide. The other fossil fuels are relatively expensive. The really interesting comparisons are with renewables; concentrating solar thermal (using mirrors to heat a liquid) is $34 per gigajoule, industrial-scale solar photovoltaic (panels) is $16, offshore wind is $24, and onshore wind is $14. Solar cells and onshore wind are already the most competitive forms of power generation, and fusion is unlikely to beat them on price.37

  Fusion isn’t magic and it certainly won’t be free or too cheap to meter.38 It’s like Nick Hawker said: the problem with power generation isn’t cost, because wind and solar are already the cheapest power sources, and are likely to remain so. If we take the cost estimates seriously, then the price of fusion energy will be competitive, perhaps one of the lowest, but it’s unlikely to be the cheapest form of power.

  What fusio
n can buy the world is carbon-free energy on the scale that we need, and at the rate of deployment that we need, for the period of time that we need, in order to save the planet.

  I. The quoted figures are medians for a range of countries as estimated by the International Energy Agency with a discount rate of 7 percent. The order of prices given is similar to the order of prices for electricity given by the US Energy Information Administration.

  EPILOGUE CAN WE AFFORD NOT TO DO FUSION?

  “Thermonuclear energy will be ready when it becomes necessary for humanity.”

  —Lev Artsimovich, head of the Soviet Fusion Program, 1951–19731

  Life on Earth is robust but fragile. It’s robust because life has been going on for at least 3 billion years. It’s fragile because most species that have ever existed have gone extinct. When I think about the planet’s most successful species, I don’t think about humans. I think about crocodiles and coelacanths.

  Crocodiles have racked up 85 million years as a species. But even hardy crocodiles are novices compared to another species that has persisted for 400 million years. The coelacanth is a lobe-finned fish species that was thought to have died out 66 million years ago. Apparently, no one told the coelacanths this because, in the 1930s, one was caught by a fishing boat. Far from going extinct, they’ve survived four of Earth’s five mass extinctions over their incredible run (you can see the 66-million-year-old fossil of a coelacanth next to a twentieth-century specimen preserved in formaldehyde in London’s Natural History Museum). In our current form as Homo sapiens, we’ve only been around for a few hundred thousand years. We need to start playing the long game when it comes to our own survival. We need to be more coelacanth.2

  Mass extinctions and global catastrophes do happen. All too recently, we’ve seen how a pandemic can leave death and economic ruin in its wake, and how unpreparedness can amplify its worst effects. But the Earth as a whole isn’t safe from other terrors. Three of the biggest existential threats we face are asteroid or comet impact; massive volcanic eruptions, called super-eruptions; and—unique to us—runaway human-caused climate change.

  An asteroid about eight miles wide is probably responsible for wiping out the dinosaurs and 75 percent of all other species at the time. The dinosaurs reigned for more than 100 million years, so they weren’t too shoddy in the survival stakes either. In 1908, we got a tiny, frightening taste of this kind of event when an asteroid that was at most a few hundred meters wide disintegrated above Siberia and destroyed more than 2,000 square kilometers (approximately 772 square miles) of forest. The chances of an asteroid or comet big enough to kill most of the world’s population smashing into Earth are around one in ten thousand per century. Super-eruptions are more likely and would affect at least half the planet and probably all of it, raining down ash over entire continents.3

  The aftermath of these extinction-scale events would be enormous climate change. Dust and earth from asteroid impacts or volcanic eruptions could block out part of the Sun’s light, producing a cooling effect and making it harder to grow food or use solar energy.

  What can we do to prepare for these rare but world-changing disasters? Having a source of energy that can keep going despite sweeping and adverse changes in climate seems like a good precaution. As we know, fossil fuels will run out before too long. And renewables that rely on large areas of land are susceptible to environmental changes. Fission could be one solution. Star power is another: the fuels are (relatively) common—deuterium is found in all the world’s oceans while lithium is found on all the world’s inhabited continents—and in any case not that much of either is needed.

  This may all sound scary. It is. But it’s also prudent to think about in the long run. We know that these events can happen. On long enough timescales, they’re almost certain to. Personally, I’d like humanity to thrive far into the future. The choices we make today have enormous consequences for future generations. Star builders say that using a fraction of world research budgets to perfect fusion energy is a small price to pay for a disaster-resilient power source.

  Of course, there are reasons beyond just saving our skin to want to see fusion achieved.

  The pursuit of fusion has led to scientific discoveries that are among the most extreme and surprising of any field. Think for a second about plasmas. Understanding them is key for fusion, yes, but every plasma discovery also gives us a better understanding of 99 percent of the visible universe. Plasmas are one of the most dramatic examples of how “more is different”: while we might understand the behaviors of the individual components—nuclei and electrons—something changes when they’re combined in large numbers and complexity emerges out of simplicity.4 Understanding these rich phenomena can be its own reward, as with so many other topics in science. Even if the study of plasmas wasn’t worth it for the joy of discovery alone, their emergent complexity holds practical lessons for other subjects, like economics, where people’s interactions are also different from the sum of their parts.I5

  A growing understanding of plasmas has led to more practical applications too, like cleaning surgical equipment or, quite literally, growing diamonds.6 Also, using lasers and plasmas together has resulted in new and better ways to fight cancer: lasers can be used to accelerate protons in a plasma to very high energies, and those protons can more accurately target cancerous cells than, for example, X-rays.7

  Machines like NIF are doing incredible science in addition to the experiments on inertial fusion energy and stockpile stewardship. NIF has been used to re-create the conditions in the core of stars that have ten times the mass of the Sun, leading to better estimates of their rate of fusion reactions.8 NIF experiments have taken us elsewhere in space too. Shortly before his death at the age of ninety-five, Livermore founder Edward Teller told scientists there that what he wanted for his one hundredth birthday was to get “excellent predictions—calculations and experiments—about the interiors of the planets.”9 The strides forward came too late for Teller, who passed away in 2003, but since NIF opened, Livermore’s scientists have managed to re-create the huge pressures of the interiors of gas giants like Jupiter and Saturn, albeit on a tiny scale. In the experiments, NIF’s laser beams were used to compress liquid deuterium to 6 million times Earth pressure and to a temperature of a few thousand degrees. As the pressure increased, usually transparent deuterium liquid first became opaque and then, most remarkably and bizarrely of all, turned into a shiny metal. It is like squeezing your coffee cup and finding that it has turned into a plate.10

  The engineering challenges that star builders like Professor Ian Chapman are solving en route to commercializing fusion have industrial spin-offs. Culham’s remote handling robots can be used in many situations that require dexterity but aren’t safe for humans to enter. Also, developing resilient fusion reactors is pushing engineers to create new materials suitable for extremes.11 Lawrence Livermore has filed a host of patents based on what they’ve had to invent to make NIF work. As happened with research into crewed space flight, fusion research is driving innovation beyond its own needs.

  The scientific and industrial reasons to pursue fusion, good as they are, aren’t the most bold or ambitious arguments to perfect the power source of stars though. There’s a reason to achieve fusion that speaks even more loudly to our existence as a species—which is that we could spread our wings and explore the universe.

  Venturing farther into space sounds like a wild dream, and it is, for now. But we’ve been to the Moon. We’ve landed an un-crewed spacecraft on an asteroid. We’ve sent probes outside of the solar system. Before too long, we may send a crewed mission to Mars. What person doesn’t want to open the next door and see what’s waiting for us in the rest of the universe?

  The only way we’ll travel to the universe beyond our celestial backyard is with plasma physics and nuclear fusion. Fusion rockets are humanity’s best hope for traveling across the vast distances of space.

  Rocket science has a reputation for being complicat
ed, but it all boils down to two simple ideas. The first is this: if you expel stuff in one direction, you’ll travel in the other direction. The second relates to how much force can be created. That’s determined by how much mass is being expelled, and how quickly it’s being expelled. Rockets expel a lot of mass quickly to get into orbit. But there’s a problem with expelling lots of mass; you have to carry the mass with you until it’s expelled. The more force you need, the more mass you have to carry, which increases the force you need, and so on. To avoid this problem, propulsion methods that work in space won’t be able to rely on expelling lots of mass; instead they’ll need to maximize the speed of the mass being expelled (and to expel only a little mass).

  You can probably guess why fusion is a good choice for the hasty space traveler. Fusion can create enormous exhaust speeds with only small amounts of mass because of its high energy density. While the best chemical rockets can achieve exhaust speeds of 4.5 kilometers (about 2.8 miles) per second, and nuclear fission might reach 8.5 kilometers per second, a working nuclear fusion reactor could potentially produce exhaust speeds of hundreds to thousands of kilometers per second.12

  Sticking a fusion reactor on a spacecraft is, surprisingly, not the only fusion-spacecraft option out there. Project Orion was part of Edward Teller’s “Plowshare” program to turn nuclear weapons to peaceful purposesII and was co-led by physicist Freeman Dyson.13 It looked at chucking exploding hydrogen bombs out of the back end of a spacecraft to cause it to accelerate in the other direction. The scheme isn’t quite as insane as it may seem, and Dyson himself estimated that it could produce exhaust speeds of one thousand to ten thousand kilometers (approximately six hundred to six thousand miles) per second. Apart from this approach to fusion-powered space travel posing significant proliferation and safety risks, tests of pulsed nuclear explosion rockets are effectively banned by international treaties—so it seems much more sensible to use controlled fusion reactors to achieve similar ends. However fusion-powered rockets are ultimately achieved, research into fusion for energy will aid their development.14