Radical Energy Abundance

Originally published at Roots of Progress, based on this thread.

Everyone knows the industrial revolution occurred centuries ago and involved steam engines. That was just the warm up. We are now a decade into the ~sixth and final industrial revolution. We are in the midst of a fundamental transformation in our economic civilization-scale industrial energy metabolism. 

While aspects of this post are necessarily speculative, its predictions are inferences falsifiably derived from easily verifiable axioms and public information. That the inevitability of this outcome is not yet more widely accepted is a side effect of anchoring bias and decades of industrial stagnation. 

The key unlock for the radical energy transformation is cheap solar power. Solar electricity is now in its fifth decade of steadily declining cost. Per doubling of production, costs have fallen a consistent 30-40% since 1970, but it’s only in the last few years that this price has become competitive with legacy electricity generation, as illustrated in this chart, which also shows how little further improvement is needed for solar to displace oil and gas. 

A number of commentators have consistently predicted that solar cost improvements will level out “this year” since about 1990, and yet if anything deployment, cost decline, and the learning rate have only improved. The graph below shows that if we weight the learning rate fit by production volume, thereby de-emphasizing largely irrelevant data from the 1970s and 1980s, we’ve actually averaged a 44% learning rate since 2009.

Is it reasonable to expect that this trend will continue, just as Moore’s Law has described computing advances for 60 years? Yes, and for the same reason.

First, it’s not adequate to draw a line through a series of points and extrapolate 10 or 100 or 1000 years in the future. We need to understand causation.

Let’s examine the demand side and understand the market conditions. Solar deployment results in increased production and revenue. Increased production results in process improvements and manufacturing efficiencies, lowering cost. Lowered cost results in increased competitiveness within existing energy markets, inducing additional demand. Is this induced demand enough to continue driving expansion of production? 

This is not guaranteed if the market were to saturate, but if anything increased solar deployment is inducing super-linear increases in demand, along with virtuous synergy 

from the parallel ramp up of grid-scale batteries for electricity storage. A full exploration of this fascinating dynamic is beyond the scope of this post but here it suffices to remark that the total addressable market (TAM) of solar electricity is roughly 30 TW, solar e-fuels is 400 TW, and current (2023) annual production is ~450 GW, growing 40-50% per year. Even at this breakneck growth rate, developers simply cannot get enough solar PV panels and at least 15-20 years of additional unserviced demand exists. 

In other words, capitalism has an insatiable demand for cheap solar power. Will capitalism get what capitalism wants? Let’s take a look at the supply side.

First, in terms of incremental improvements all data is pointing in the direction of another 2x reduction in cost over the next 3-5 years without any miracles required. Until recently, panels were the dominant cost of solar installations and the “balance of plant” was not under significant cost pressure. But that is no longer the case. Even as panel price competition gets ever more cutthroat, cost pressure is being applied to racks, labor, fencing, permitting, and land costs, all of which will result in substantial near term cost reductions. I’m not the only person saying this – in the last 18 months the industry itself has raised and invested more than a trillion dollars, in the US alone, on the basis of thorough analysis supporting these observations.

Second, let’s do a first principles-based bottoms up cost estimate. What is the Platonic ideal of a solar array? An array needs a 50 um thick layer of silicon to be fully opaque, and perhaps 100 um of necessarily flexible plastic “backing” material to provide mechanical support. Throwing in power cabling and installation rigs, I expect the installed cost of solar arrays to fall to $30,000/MW within 15 years, again with no miracles required. This is roughly 10x cheaper than the current cheapest costs. 

If we’re prepared to consider the implications of materials science wizardry – essentially expanding the class of known manufacturing techniques to include arbitrary configurations of known elements, a solar array could be made that’s even thinner, lighter, and cheaper, or even self-assembling. But even without such science fiction, existing manufacturing techniques will be extended to give us at least another decade of steeply falling costs, along with commensurate additional installations. The market will demand it and industry will provide.

Against this backdrop of ever-cheapening solar, it’s a fun exercise to consider historical, contemporary, and near future changes in the way we get our energy. 

Before the industrial revolution, if we wanted energy we needed to eat plants, animals who ate plants, or burn wood, which is made by plants. Plants are solar powered, but as we will see about 1000x less effective at capturing energy for us than silicon solar panels. That is, in terms of per-area carbon fixation with the same solar resource, plants are 1000x less productive than solar PV+chemistry.

With the industrial revolution, we found we could dig up and burn rocks (coal, oil, gas) to make heat. This heat was cheaper and often more useful than burning wood. To make electricity, we boiled water and spun a turbine. Later, we figured out how to make heat by burning oil and gas, or with nuclear fission power, or maybe soon even with fusion – but ultimately these sources all come down to the cheapest source of heat.

More recently, solar photovoltaic (PV) power got cheaper than coal, and in advanced markets it is already cheaper to build and operate solar than to continue operating a legacy thermal power plant. This seems crazy but it’s true – the operating costs of even a brand new coal or gas power plant are higher than the construction and operating costs of a new solar plant. 

As time goes on and solar gets cheaper, the markets where it enjoys a strong competitive advantage continue to grow. As a rough rule of thumb, the solar supremacy zone grows outwards from the equator at about 1 mile per day, or nearly 4000 miles per decade. If we consider that most humans live in places with intermediate solar resources neither too hot nor too cold to grow the crops their grandparents probably farmed, it becomes clear that this market expansion will occur very quickly compared to the timescale of a human lifetime!

Similarly, in advanced markets characterized by abundant solar resources, the cheapest heat is no longer obtained by burning rocks, it’s actually achieved by routing solar electricity through a resistive electric heater. So not only is it no longer optimal to burn rocks to make electricity, it’s no longer optimal to burn rocks even to make hot rocks. It seems crazy because intuitively electricity is a higher grade form of energy than heat, but burning rocks is not getting cheaper while solar is. This transition is only a few years behind direct electrical competition, occurring first in California and similar markets, then more generally across the southern US, then large swaths of the rest of the world, then Europe, then Scandinavia, and finally (maybe) Finland. 

If it’s cheaper to make heat electrically than burn rocks, what is the next transition after that? The mission of my company Terraform Industries, which is making synthetic carbon neutral natural gas from sunlight and air. Very soon it will be cheaper to make rocks to burn from solar energy than to find rocks to burn in the Earth’s crust.

Finally, early studies today are finding ways to electrically synthesize starches, proteins, and fats from water and air. By 2050 it may be the case that solar panels and chemistry give us the ability to make most of our food with 0.1% of the land, reducing the enormous and catastrophic environmental impact of our civilization farming literally 40% of the Earth’s entire land surface area, while reducing food costs and improving food security.

Now we’re done with the broad intro, let’s dive into the specifics of how humanity will productively dispose of this imminent cornucopia of cheap, high grade energy? What is going to change?

Since 1973 oil has been scarce. Until 1973 energy consumption grew and long term investments in energy-intensive industries were easy to price. Since 1973, energy scarcity has driven a general stagnation on many key axes of progress, while our civilization found growth in the less energy-intensive industries of computation and services. In 2023, the exponentially expanding growth of solar is putting our civilization permanently back on track to increased productivity, longevity, prosperity, and happiness. 

Unconditional open-ended energy abundance changes everything. It’s not called an industrial revolution for no reason. Let’s get specific about what we expect to occur.

We will use more electricity for more things. Electricity and electrical machines will both displace existing fuel applications, such as with electric cars and residential heat pumps, and lead to new applications. 

On the generation side, we expect to see solar, batteries and, in certain markets, wind continue to aggressively displace existing production while increasing total supply and reducing cost. Why? Solar and batteries can dispatch power in real-time markets with zero moving parts. Any power generation system that runs steam through a turbine is fighting an uphill battle, given that costs already favor solar approximately 5 to 1 and are trending further in this direction at an accelerating rate. The logical end state of this displacement is not some hybrid grid with around 20% solar backed up by spinning generation, but a much more dynamic, versatile grid with around 200% solar generation and hours to days or weeks of battery storage, fully decoupling time of generation and consumption.

As battery cost falls for the same reason as solar (exponentially increased induced demand and a lot of low hanging fruit on production improvements), we will see batteries aggressively displace transmission as the most cost-effective way to ensure continuity of electricity supply, leading to an eventual pruning of the grid and drastic reduction in the average distance that power travels between production and consumption, even as the average time that power is stored increases.

As solar cost at the array falls below $10/MWh in the next few years, it will be cheaper to synthesize hydrocarbons from CO2 and water vapor than to drill them from the ground and refine and transport them. Synthetic fuels are cheap, local, carbon neutral, and above all scalable solutions to both fuel scarcity and CO2 emissions. Developing this tech is the primary focus of my professional effort at my start up, Terraform Industries. 

Obviously unlimited local production of any desired hydrocarbon fraction (methane, propane, octane, kerosene, etc) changes a lot, but there’s more. What of the second order effects?

CO2 per se has no intrinsic moral valence. NASA CO2 emissions data correlates spectacularly well with indices of development and wealth creation. Our entire civilization is built on the unimaginable wealth of coal, oil, and gas sitting below our feet, free for the taking. The problem is that CO2 emissions result in the net transport of carbon from the crust into the atmosphere with no corresponding reverse transport, leading to a buildup of CO2 and its deleterious climate consequences. 

But cheap fuel is equivalent to reduced poverty. The default climate discourse in recent years has been congruent to an acceptance that curtailing CO2 emissions would require the regrettable but ultimately impersonal de-industrialization and re-impoverishment of most of the world, which is both morally repugnant and politically impossible to achieve without first use of nuclear weapons. My controversial position is: We shouldn’t do this. 

Instead, we should find a way to capture atmospheric CO2 and turn it back into fuel, closing the carbon cycle in the atmosphere, self-funding the scaled deployment of 50 GT/year of direct air capture (DAC) capacity, and aligning aggressive poverty reduction with aggressive industrial expansion. Cheap solar is the unlock that brings human progress and climate action from the realm of fundamental incompatibility to complimentary inevitability. 

Cheap energy drives economic growth. For 50 years we’ve gotten by with an average of 2-3% per year, hardly enough to make up for increases in lifespan, reduction in birth rate, and economic opportunity for young people. Cheap solar and unconditionally abundant electrical and chemical energy will grow the economy at the rate that humans can convert the crust into solar panels. Battery investments were already growing 250% year over year before the Bipartisan Infrastructure Law (BIL), the Inflation Reduction Act (IRA), and the CHIPS act. Solar investment and manufacturing R&D is screaming along at record rates, with US re-shoring and re-industrialization exceeding even the build up to WW2. 

I believe there is a good chance we will never again experience a major recession. We will put essentially all the physical needs of all the people below the API and direct our collective cognitive effort (helped by trillions of increasingly sentient AIs and AI-human cybernetic hybrids) towards the larger projects – terraforming and interstellar travel. 

The oil and gas industry turned over $6.4T/year in 2022. This is roughly $1b/hour, but it will expand several times over as it is freed from the constraints of geology and scarcity. Following current trends to second order, >95% of humanity’s energy needs will be downstream of solar panels by 2042, just 19 years away, and alongside enormous expansion in direct use of electricity, fuel consumption will grow to more than $25T/year in equivalent dollars. 

Half the world’s population is currently experiencing oil scarcity due to supply chain complexity, weak buying power, and political instability. Solar synthetic fuel is relatively simple and can be produced anywhere the sun shines, bypassing this Gordian knot and bringing energy abundance to every corner of the Earth. 

For much of the other half, energy use is constrained by cost. Solar cost reductions will lower fuel prices globally to the range of $1/gallon, inducing additional consumption and increasing revenue for additional construction in a positive feedback loop.

Current high consumers of energy will expand in number and energy intensity, such as by a re-introduction of supersonic commercial aviation and its extension to mass market use.

Finally, more people consuming more energy will lead to expanded economic growth across all sectors, not just energy. We expect growth to at least double to 6%/year, for a cumulative tripling of the global economy by 2042. This is a conservative estimate, it could easily be much more.

End uses for fuel, whether fossil or synthetic, will continue to evolve. Fossil supply will continue to trend towards sour crude as we tap the remaining dregs, while synthetic fuel starts at methane (CH4) and can trend towards heavier molecules (C3, C8, C10 etc) via Fischer-Tropsch synthesis. As a result, we expect  the relative cheapness of natural gas to drive end uses in that direction, as is already occurring, in turn pushing the fuel mix towards lighter hydrocarbon fractions. 

Ground transportation and residential heat will trend towards electric, though the process of rolling over the existing fleet will take at least 20 years for vehicles and probably more than 50 years for buildings in the absence of extremely ambitious and ultimately economically inefficient policy on a global scale.

Cheaper fuel that happens to be synthetic and carbon neutral will also drive expanded production of heat-intensive products, such as metals, cement, silicon, and ceramics.

Probably the single sector most poised to benefit from cheaper synthetic fuels is aviation. As I write this on a 737 over the Gulf of Mexico, I reflect on the fact that flying planes is a huge privilege that saves our most irreplaceable resource – time. Yet the legacy zeitgeist for aviation and climate action is driven by a scarcity mindset, and this is not a hypothetical. There is legislation under consideration in the EU which would mandate carbon offsets in aviation, further driving up the cost of flying and restricting its benefits only to the very wealthy. Quite apart from the fact that carbon offsets are an ineffective and counterproductive waste of money, is enshrining the exclusivity of jet travel in law a positive outcome for humanity? No!

Instead, consider an aviation industry that adopts, out of economic necessity, cheaper and better synthetic sustainable aviation fuel (SAF) or even liquefied natural gas for next generation high performance jet engines, such as those being pioneered by AstroMechanica. SAF can be designed to have higher density (for longer range), lower emissions, and is naturally carbon neutral. Globally driven solar cost reductions will drive a steady trend in SAF cost reduction until (very soon) it is cheaper than fossil jet fuel and ultimately cheap enough that jets can be enjoyed by hundreds of millions of people, even billions. This is in contrast to the default current case where perhaps only 10 million people can afford to fly jets semi frequently, and who in the future will drive adoption and development of high performance supersonic transport, further saving millions of person-years from waste in transit.

Reliably cheap carbon-neutral hydrocarbons are also a huge help for the chemical industry, forming the basis of plastics, paints, medicine, dyes, resins, epoxies, fertilizers, explosives, and a million other things.

Let us consider other forms of material abundance enabled by cheap solar power.  Electricity can be readily converted into heat, radio waves, or mechanical force, all of which are transformational for mining. Hitherto mining technology has been barely more sophisticated than Flintstones-level “man dig hole in ground”. But there’s no reason that our pursuit of valuable minerals in the crust needs to start by destroying large swaths of the upper few feet that are so critical to life. Instead, we will see the commercialization of integrated tunnel boring machines of enormous size that will extract valuable minerals at depth before extruding a narrow, fast-moving wire of any desired alloy through a small, surface-mounted “well”. Ordinarily, these machines will back-fill their tunnels with the dross but could also carve out enormous underground caverns for any desired use, be it high speed trains, vertical farming, nuclear reactors, underground rivers, or secret realms. 

The other major growth area for power consumption is in computation. Unless room temperature superconductors enable heat-free reversible computation in the super near future, we’re going to need to use a lot of silicon out in the desert to power the silicon in the data center that forms the first non-human consciousness we have encountered since the extinction of the Neanderthals. 

In addition to reducing the cost of hydrocarbons for thermal processing of metals and cement, and the direct synthesis of plastics, copious cheap electricity also enables direct electrocatalytic production of a range of materials, including light metals like aluminum and magnesium. There are even electrocatalytic methods for producing steel but I am unsure if they will prove competitive. 

To take one example, current steel:aluminum production is 20:1. Much cheaper electricity will see the relative cost shift further in aluminum’s favor, perhaps to the point of 1:1 production, achieved through a 20x expansion of aluminum production. Aluminum is mostly used for weight-critical applications, such as aircraft, but in the near future we may see it used for bridges and rebar!

Many other materials can also be refined electrically. Titanium, magnesium, fresh water, copper, lithium, potassium, rare earths, all the good stuff.

To linger on water for a moment, solar powered reverse osmosis already appears to be cost-competitive with other water sources for the desert irrigation of high value crops. Billions of people worldwide depend on rivers whose flows are threatened by climate change. Their economic prosperity will ultimately depend on their ability to safeguard their water sources against disruption, through desalination and perhaps cloud seeding. For example, California’s total extraction from the Colorado river, 5 million acre-feet per year, could be supplied via solar desalination with capex of just $20b over 10 years, less than half the price of Twitter. This is very achievable. Half of the US is essentially desert suffering acute water scarcity, and this could be economically addressed with relatively trivial investments on top of solar’s ongoing cost reductions.

What is the last thing that cheap power might enable, if solar gets cheap enough? Generic recycling of waste not by streaming similar materials such as plastic, metal, and paper and feeding them into secondary products, but by converting the entire waste stream into plasma and sorting it atom by atom with a gigantic mass spectrometer. This is a level of materials capability far beyond our current one, at least at scale, but it is not that different to how biology generates self-improving self-assembling organic robots (plants, animals, us) and it is permitted by the laws of physics. 

What other unexpected things will occur when electricity falls below $10/MWh?

The value of currency is backstopped by energy price, so once solar energy crosses some productivity threshold (~10 TW) to become the dominant source of energy in the world (around 2035), further cost decreases will be expressed not as a reduction in $/MWh but in partial deflation for some goods/services and inflation for others.

Finally, despite essentially unlimited gains in energy production and cost-effectiveness, some basic commodities and factors of production will remain scarce without subject-specific innovation. These include time, human lifespan and youth, attention, access to more capital than we’ve yet built, some illegal configurations of matter and information, scientific and astronomical knowledge, algorithms, organizational capacity, cognition, delta-V, usable antimatter, and (though this one is completely self-inflicted) housing.

Let’s build!