A Tale of Two Elements
Hype and Hope in Imagined Energy Futures
This post is adapted from a keynote address I gave at the Centre for Independent Studies Consilium conference in Australia, October 24, 2024
Today, I want to tell you a tale of two elements: Silicon and Uranium.
Silicon is about how far I can get when reciting the Periodic Table from memory, and Uranium — well, we can all dream can’t we. Left to themselves, these elements are nothing special, but with the input of human ingenuity, they can acquire magical properties.
In the form of photovoltaics and fission reactors, vast quantities of energy can be unleashed. The fact that humanity has harnessed almost every source of energy in the universe—from combustion to photovoltaics to harnessing the strong nuclear force, is remarkable. Unleashing these energies to transform the world and solve our most dire problems is one of the most inspiring stories humanity has to tell.
And yet, as with so many human endeavours, our stories can become embellished, occasionally to the point of delusion.
Part 1: Uranium
In the late 1940s and 1950s, the world's best physicists, chemists, and engineers were leaving the Manhattan Project. Some were known as "Atomic Malthusians."
These scientists were obsessed with the idea of a population bomb and the depletion of natural resources, especially fossil fuels—which, much like climate change today, was perceived as the great crisis of their time.
They understood that energy is the ultimate determinant of human carrying capacity and worried that as we ran out of conventional energy sources famine would ensue.
The most prominent of these thinkers was Marion King Hubbert, better known as the godfather of peak oil theory. In his role as a consultant for the US Atomic Energy Commission, he championed nuclear energy in the context of fears over oil depletion.
What was needed was an energy technology breakthrough. Enter the breeder reactor.
Conceived and developed by a cohort of Manhattan Project physicists at the Chicago metallurgical lab, this reactor could, in theory, provide an inexhaustible source of energy, as it created more fuel than it consumed.
Here was a technology that could prevent the collapse of modern industrial civilization under the crushing burden of uncontrolled population growth. Perhaps they felt that by defusing the population bomb, the breeder reactor could absolve them from the original sin of unleashing the world-ending power of the atomic bomb.
Harrison Brown, one of the most prominent Atomic Malthusians, carried out some striking calculations.
He determined that 7 billion people, a staggering number by 1950’s standards, could live American-style lives for hundreds of thousands of years with 17,000 breeder reactors burning the uranium contained in seawater—an inexhaustible source of the fuel.
Lewis Strauss, the chairman of the US Atomic Energy Commission, famously claimed that nuclear energy could become "too cheap to meter." With no limits to energy abundance, humanity would desalinate ocean water to make the deserts bloom, transforming vast tracts of barren land into farmland to feed the billions of hungry mouths on the horizon.
We could manufacture fertilizers, extract and refine declining ore grades, and recycle metals ad infinitum, all with a minuscule environmental footprint—a utopian vision indeed.
Atomic Malthusians imagined this energy transition would occur over hundreds of years, a relatively slow pace alongside the gradual decline in fossil fuel resources. It was an ambitious timeline but an eternity compared to the lightning speed called for in our current net-zero energy transition plans.
Sadly, this world has not materialized, and its partial manifestations in the golf courses of the Gulf States are largely fossil-fueled.
Humanity has failed to create economically viable breeder reactors despite 70 years of effort. Instead, conventional light and heavy water reactors have been deployed at a more moderate pace in 32 countries around the world, accelerating particularly in the aftermath of the OPEC crisis.
Peak nuclear deployment occurred in 1984, when 33 reactors were simultaneously grid-connected around the world in a single year. Nuclear power reached its apex of 15% of global electricity in 2006—a respectable accomplishment achieved in only 50 years, given that only 32 countries had pursued nuclear energy.
But this was still a far cry from the science fiction utopia imagined by Harrison Brown.
My home, the province of Ontario, did its bit. Between 1966 and 1993, we constructed 20 large reactors in just 27 years, ultimately phasing out our large coal fleet and providing 60% of our current electricity needs. The result: an accidental decarbonization—to my mind, the best kind.
Since its heyday in the 1980s, deployment has slowed. Only 7 new reactors were commissioned worldwide in 2023.
It is alarming that we are setting consumption records every year for every other source of energy: biomass, coal, oil, gas, wind and solar, while nuclear has receded in absolute and relative terms. Its 15% share of electricity generation has dropped to just 10% today. Biomass, our most ancient fuel, supplies more energy than our most advanced source, nuclear.
Many technologies follow an “S curve” in their adoption. Diffusion of the technology begins slowly, accelerates and eventually plateaus. Nuclear has followed this pattern. Knowing its long term value in a carbon constrained world, I sincerely hope that with innovation in project management, not necessarily in exotic reactor designs, we can add another “S curve” on top of this historic nuclear deployment, a phenomenon known as “S curve stacking.”
But enough of Uranium, It is time for Silicon’s time in the sun!
Part 2: Silicon
E=mc², the world's most famous equation, explains the incredible energy density and promise of nuclear power.
Ironically, Einstein won fame, but not the Nobel Prize, for this stunning discovery. Instead, he won this most prestigious award for his description of the photoelectric effect and its contributions to advancements in quantum theory.
His work demonstrated that light could behave like discrete packets (photons), thus challenging the wave theory of light. Although the photovoltaic effect was discovered almost 70 years earlier in 1839, Einstein's work provided the theoretical foundation for solar cells, explaining how light energy—photons—could excite electrons, causing them to move through a semiconductor material to create a current, and ultimately, electricity.
This application of human ingenuity to an otherwise unremarkable element is worthy of wonder.
More wondrous still is how photovoltaic cells have come down the cost curve and how solar deployment has blown past all expectations and IEA forecasts, continuing on a near-exponential rise.
For example, in 2022, China deployed 87 gigawatts of new solar to arrive at ~400 gigawatts of total solar capacity. In 2023, it added another 216 gigawatts—a staggering 150% year-over-year increase. By 2026, it looks set to arrive at 1 TW, equivalent to the capacity of 150 large nuclear plants when adjusted for average Chinese solar capacity factors.
Interestingly, in this solar cost reduction and scale of deployment, we have another emerging vision reminiscent of the Atomic Malthusians: Solar energy "too cheap to meter."
In this vision, despite capacity factors of 25 percent even in the sunniest places on earth, the sheer scale of envisioned solar parks can, in theory, produce power so cheap that it makes the vast amount of battery storage and green hydrogen necessary to firm them inconsequential.
Perhaps the utopian dreams that eluded the Atomic Malthusians can be fulfilled?
Today it is the "Solar Punks" who suggest we can make the deserts bloom and power a growing green economy with all of its industrial inputs essentially carbon-free due to the power of the photon and the miracle of polysilicon semiconductors.
In an interesting parallel to the psychological motivations and embrace of magical thinking of the Atomic Malthusians, the Solar Punks would atone for the original sin of carbon emissions and deliver us into another vision of utopia: Net Zero.
Part 3: "In Models We Trust"
Climate-concerned energy modelers have proceeded to brute-force this vision onto the impossibly tight timelines of Net Zero by 2050.
These modeling exercises, rather than being treated as entertaining curiosities, have begun to morph into policy prescriptions and even doctrines of faith.
A picture is worth a thousand words, but I may need a thousand words to describe the implications of this picture.
In the unconstrained model, 65,000 square kilometers of solar panels are built. If that sounds like a big number, it's because it is. Imagine 12 million football fields or the total land area of Tasmania blanketed in solar panels.
For reference, the largest solar farm in the world is the Gonghe Talatan Solar Park in China. It is an 8.4GW site sprawling over 420 square kilometers. The unconstrained scenario in the Net Zero Australia plan would require 154 such giga-scale solar farms.
The transformation of the Australian energy sector under this plan is estimated to require 7-9 trillion dollars of investment. The modelers suggest that through the magic of polysilicon, the desert itself will be transformed into a source of energy potent enough to attract the multinational investment required for its development—all because of the bounty of energy and products it will soon provide.
The underlying sentiment is somewhat akin to Donald Trump's insistence that he will build the wall but Mexico will pay for it.
In some ways, these solar parks—that I will now refer to as "Solar El Dorados"—are envisioned as analogous to and able to provide comparative advantage similar to the prodigious shale oil basins of the USA, which have made America the top oil producer in the world for the last 6 years.
Incidentally, those shale deposits are also the product of sunlight, albeit trickle-charged over fifty million years and concentrated into high-density, transportable, and versatile molecules that enable the very existence of our modern world.
The Permian basin does indeed spontaneously attract private investment. Well-capitalized frackers build their own roads, pipelines, and transmission to speed up their access to the tremendous wealth under their feet.
Why would Australia's current customers for its energy exports—coal and LNG—and its iron, copper and lithium help pick up the 7-9 trillion dollar bill?
According to the Net Zero modelers, this "Solar El Dorado" will power far more than Australia's existing electrical grid.
On the energy export front, coastal desalination plants will be built to provide fresh water through dedicated pipelines to inland solar-powered electrolyzers.
These in turn will create green hydrogen, which will then be put through the Haber-Bosch process to be transformed into green ammonia and exported by boat in sufficient quantities to replace all of Australia's coal and LNG exports.
But wait, there's more. The "Solar El Dorados" will also produce the green hydrogen to reduce iron ore and power the electric arc furnaces needed to turn it into steel.
They will also provide the electricity required to power the Bayer and Hall-Heroult Processes to produce green aluminum—an activity so electricity-intensive that aluminum is sometimes referred to as congealed electricity.
Perhaps this green industrial revolution will extend into polysilicon production, whose 3-4x greater electricity demand makes aluminum blush.
All of the aforementioned processes are not just electricity-intensive; they are incredibly sensitive to disruptions of electricity supply which are more likely if powered by intermittent renewable power sources.
One last miracle to note other than the scale of what is being proposed. The speed.
To implement this new reality by 2060 would require complete societal mobilization. Such a vision is reminiscent of the ruthless timeline of the Manhattan Project, which consumed an enormous level of American human and material capital, ultimately producing the bomb in just 2 years and 10 months.
It is hard to imagine that kind of mobilization in any context other than total war.
Part 4: "A Few Challenges to Overcome"
The vision of decarbonizing and re-industrializing Australia is a noble one. However, careful analysis of the available options make a few challenges with the plan leap out at me.
1. Reliance on China
Can we assume that China is capable—politically and economically, in the medium and long term—of continuing to flood the world with cheap solar panels and batteries?
China’s oversupply has been possible due to generous direct subsidies and tax incentives, as well as low cost financing and investment in innovation and enabling infrastructure, all paid for by the Chinese tax payer.
Human beings tend towards a cognitive distortion known as extrapolation bias. It reflects the tendency to assume that current trends, whether linear or exponential, will continue indefinitely into the future.
The dynamic “S curve” pattern discussed above is not intuitive. It is however conceivable in the long term that the “S curve” pattern of technological diffusion may also apply to solar power.
This may occur because of eventual limits to ultra-cheap Chinese production at scale. But more likely because of difficulty integrating the diurnal nature of photovoltaics into the enabling infrastructure necessary to provide the reliable power requirements of modern industrial civilization.
In any case, the party may not go on forever and Australia is in the short, medium and perhaps long term making itself vulnerable by buying so much of its kit, and putting so many of its energy eggs into Xi Jinping’s basket.
2. “Small is beautiful”
Australia's renewable energy transition to date has largely been a feel-good one. Subsidized rooftop solar panels, home batteries, and EVs make participants feel like they're doing their bit. It's a "small is beautiful" vision compatible with and not disruptive of life or their surroundings as they know them.
However, what is proposed in these modeling studies and championed by Australian foundations like the Superpower Institute is a dramatic transformation of the land and its people.
Massive solar farms, transmission lines, and water and hydrogen pipelines will crisscross the landscape and industrialize the continent. The resulting blowback will likely make protests over early transmission expansion currently underway look quaint.
3. Workforce demands
This vision is also a jobs program on steroids. The model involves increasing the energy workforce from 100,000 to up to 700,000.
Many of those energy workers will have to toil in the inhospitable solar fields of the Northern Territory, Western Australia, and central Queensland—places that have been historically lightly populated precisely because they are desolate.
Will the hundreds of thousands of workers constructing these projects at breakneck speed fly in, fly out for 14 days at a time and live in camps as approximately 37,000 Australian miners currently do?
Or do we envisage resettling them in new stable industrial cities required to house the workforces that will run, maintain, and operate the facilities of this green industrial revolution, complete with hospitals, daycares, schools, movie theaters, utilities, and of course, baristas?
Workforces hundreds of thousands strong will need to take to the inhospitable deserts
How do you convince hundreds of thousands of people, one-third of them skilled electricians, to live and work where few would choose to live and work?
Allow me, in the spirit of Jonathan Swift, to make a modest proposal, albeit with an Australian twist:
Perhaps convict colonies could be established, and early release could be negotiated in exchange for a certain number of years of labor in the solar fields.
I am here to criticize in a light-hearted way, the magical thinking of both the Atomic Malthusians and the Solar Punks.
Energy reality is likely to fall far short of our dreams, which have been made fantastical by our sincere desire to defuse existential threats, whether a population bomb or a carbon bomb.
I have been focusing my attention on the more contemporary example of the Solar Punk energy fantasy, since I think it demonstrates that there is no easy way out of our current predicament. Challenging as it will be to build civilian nuclear power in Australia, it may require fewer of the more far-fetched plans that characterize Net Zero Australia.
It is certainly possible, for instance, to reduce land impacts and energy workforce resettlement if you can build nuclear plants in the Latrobe and Hunter Valleys, where coal communities are hungry for new opportunities now that historic coal stations are being decommissioned.
It must be conceded that a 95% renewables system is a bold and unprecedented experiment. Nothing is certain; despite my doubts, it might just be possible to run a modern, industrialized economy primarily on wind, sun, and batteries. But if it can't, it would be a devastating blow to the Australia that I have come to know and love.
Given the limits of either strategy, I think we will need to deploy a mixture of resources harnessing the power of both uranium and silicon in rational amounts in rational places. Only then might we achieve a myriad of goals including climate stability, energy security, and affordability.
I believe that a critical reading of the flights of fancy of the Net Zero Australia model illustrates how unlikely it is that net zero can be achieved in decades, let alone centuries.
An obsession with net zero by the arbitrary date of 2050 produces poor decisions about long-term planning and a bias toward quick “cheap” fixes and against large-scale, long-lived infrastructure like nuclear. And we do know with certainty that nuclear can support heavy industry and manufacturing.
These industries, I believe, are necessary for Australia to reverse its rapidly declining economic complexity. Faced with a deglobalizing world and escalating great power competition, Australia urgently needs to grow its capacity to join its western allies in reshoring strategic industries.
The reality that the road ahead for nuclear power is not an easy one is no reason to avoid getting started. Rather, it is a call to meticulously study best practices, acknowledge sophisticated criticism and avoid being led astray by reactor vendors' rosy sales pitches.
But above all we must recognize the human propensity for magical thinking of which we are so capable when faced with existential crises, whether we are enamored with Uranium, Silicon or both.
Share this post, make a pledge and subscribe to stack another “S Curve” onto Decouples readership.
If you would like to book me as a speaker click on the button below
Decouple I love that you’re on Substack now
I found this article interesting but object to comparing early era nuclear engineers and their supporters as Atomic Malthusians. Nuclear energy pioneers did see nuclear energy as a way for humanity to have sufficient resources to realize our potential. That is much different than Malthusian theory that leans heavily on population control. In fact the term “Atomic Malthusian” first appeared in 1971 in a Bulletin of Atomic Scientist that feared abundant energy from nuclear energy could lead to growing populations and extravagant consumption. Atomic Scientist generally are avid anti-nuclear energy proponents. I am a great fan of Decouple and have learned a lot from the series of excellent podcast with fantastic guest and lively conversations. But be careful about the rose colored glass mocking. Any deep dive into the great inventors of history, and we can include Elon Musk and SpaceX in our time, demonstrate that it is indeed a much improved version of the world that leaves humanity in awe of what can be accomplished. The Experimental Breeder Reactor II was successful on almost every metric, and had that work been allowed to go on there is a possibility nuclear energy deployment would be decades ahead of where it is now. There is not some great central authority suited to dictate to humanity what shall be the technology to deploy. It’s a complicated, messy, and psychological process. But in the end free people working together to innovate find the right paths, and standardization is probably the last step. You mentioned, our early nuclear energy pioneers thought it might take hundreds of years to transition from fossil fuels to nuclear energy. We’re 80 years in now, I think they are still correct.