Chile

Intro/Abstract

Why are today’s cities where they are? If the Solar Industrial Revolution (SIR) plays out, where will the great cities of 2100 be? I propose the single most likely place for a new megacity is on the Mejillones Peninsula in Chile. This takes into account the factors that have influenced the great cities of the present and incorporates highly likely future trends. Privileged access to the sunniest region on earth has non-obvious and substantial implications.

Why are cities where they are today?

Geography

Logistics

Most of the biggest cities of today were the biggest cities of their respective regions 100-200 years ago, just with more people because of population growth and urbanization. Strength leads to strength. Most commonly, a city’s reason for existence is some sort of logistics advantage. A city can be seen as a node in a larger trading network and the city itself can be seen as a travel time minimization network. As a node in a larger trade network, a city’s logistics advantage is everything. Each successive wave of infrastructure buildout further locks in the that day’s winners. For example, railroads locked in Atlanta, Dallas, and Chicago as major cities. Interstates locked in the previous three further as well as Phoenix and Las Vegas. Most cities are old enough that their access to the original transportation layer, water,¹ determines their locations.

Common water features might be the confluence of two rivers (Portland, Pittsburgh, St. Louis, Wuhan, Belgrade), or a coastal gap in the mountains (Tokyo, LA, Mumbai, Cape Town, Vancouver) or where a river meets the sea/lake (Shanghai, Guangzhou, Buenos Aires, Chicago, Hamburg, Kolkata, New Orleans) or at the limits of a river’s navigability (US Fall Line cities, Ottawa, Minneapolis) or a spectacular natural harbor (New York, SF, Hong Kong, Seattle, Sydney, Rio de Janiero, Halifax, Honolulu) or a trade chokepoint (Singapore, Istanbul, Panama City). Some cities emerge as the largest in a fertile agricultural area (usually on a river) (Cairo, Sacramento, Chengdu, Bangkok, Beijing, Paris, London).

Resource Access

Materials

Some were started for their proximity to natural resources (Denver, Midland, Houston, Perth), though more commonly what caused them to grow is their logistics advantage in moving or processing those natural resources. The cities actually extracting the resources are fairly small, especially since mining became mechanized.

Energy

There is substantial evidence that cities will spring up around access to the cheapest power in a region. This has historically almost always been hydroelectric dams, but this has happened around the Persian Gulf and US Gulf Coast as well.

The original Niagara Falls power plant saw new energy intensive industries springing up around it like wildflowers. Nowhere else could this kind of cheap electricity be achieved at the time, and nowhere else did the economics of many novel electric processes work. This is where Hall–Héroult aluminum and silicon carbide were first successfully commercialized.

“In [Acheson’s] case, electricity would fire new arc furnaces capable of reaching unheard-of temperatures. He, too, sought electricity at Niagara. […] Hall with his aluminum and Acheson with Carborundum would soon be followed by many other entrepreneurs starting or expanding electrochemical or electrometallurgical firms producing “acetylene, alkalis, sodium, bleaches, caustic soda, chlorine.””From Empires of Light

Eventually the economic advantages of being colocated by Niagara were minimized by advances in transmission, and much of this industry moved to Buffalo or elsewhere in the region. Despite Tesla’s pronouncement that “the result of this great development of electric power will be that the falls and Buffalo will reach out their arms and will join each other and […] form the greatest city in the world”, there just wasn’t all that much energy to be extracted from Niagara Falls compared to demands of industrial civilization.

Other

Some other cities were settled for their natural defenses at the time (Athens, Lisbon, Venice, Edinburgh). Particularly during the colonial era, some were started for their mild climate and lack of disease in tropical regions — usually due to altitude (Sao Paulo, Baguio, Nairobi, Quito, Addis Adaba).

Cultural/Political/Tourism

The only semi-common exception to the economic/logistical/geographic raison d’être is cities started for political/cultural reasons. Some cities are started/chosen purely for being in the middle of an administrative area; many capitals are like this (DC, Brasilia, Madrid, Canberra). Some have unique regulatory freedom (Shenzhen, Dubai, Las Vegas, Macau). Some are cultural/religious sites (Mecca, Medina, Kyoto, Lhasa). Some are tourist sites (Orlando, Cancun, Aspen).

What are the future trends that influence where cities will be?

Future Logistics

Logistics advantages will continue to play the dominant role in city location. It’s worth considering what the future transportation modes might be. The scale transportation advantages of waterways is unlikely to disappear at a dominant factor, but some new technologies may influence future cities.

Future Transport Tech

Many are bullish on cargo airships. These will favor mountainous and island regions far more than previous technologies have, since there is no cost to “changing mediums”. The densely populated mountainous and forested archipelagos of SE Asia (Indonesia, Philippines) seem like they would be by far the main beneficiaries of this technology. These geographies are the final boss of an ocean shipping/railroad/auto dominated transport system; the immense cost of great infrastructure in these areas is a constant tax on economic growth. It’s possible that the cost of airships here is lower than other technologies; my intuition says this is where the costs breakeven first if they ever do. The large cities that spring up are where cargo has to be transferred from to and from the airship to the more distributed local logistics networks.

The worst possible geography for land transport.

I’m bullish on flying cars (mass-produced, 300 mph, self-flying, MOSAIC compliant, Joby with a turbogenerator. (full post eventually)). These likely won’t start many new cities but they will dramatically reshape the urban fabric of existing ones. In fact, the idea of cities as distinct entities will start to blur. Just as Dallas and Ft. Worth merged over time (30 min drive vs a 12 hour walk), so too will Seattle and Portland, San Antonio and Houston, Boston and NYC (30 min flight vs multihour drive). There will likely be a barbell outcome in land use. Revealed preferences of wealthy people (people who can live wherever they want) seem to optimize for a lovely urban place (West Village) or a rural estate outside of the city (Hudson River Valley). It seems reasonable that many more will chose similar arrangements once they have the freedom of flight. The reduction of busy roads as people transition to flying cars allows for Euro/Asia style walkable/bikable areas to flourish. Many will choose to live in these areas. On the opposite end of the barbell, many will choose to commute 100s of miles from their rural property. At 300 mph, that is a 150-mile 30-minute commute radius. That’s like living in the Catskills and commuting to Manhattan. Sequoia to LA. Yosemite to SF. Olympic NP to Seattle. In the US, I expect the West to be the primary beneficiary of this because of the absurd amount of natural amenities you can now access in a reasonable time. Land between cities will become more valuable, land with access to natural amenities (waterfront (islands especially) and mountain properties) will become more valuable. It also allows for the closest thing to teleportation; falling asleep in your self flying car and waking up 2400 miles away 8 hours later.

Super/Hypersonic flight. Presumably fewer, bigger airports. Large regions become more tightly knit. Stronger forcing function on reducing time spent in an airport as flight time decreases. The more airports feel like metro stations (full post eventually), the more closely knit the regions get.

Space travel. Spaceports will almost certainly become hubs of high value advanced industry over the long term. Maximally equatorial sites for a given region with oceans to their east will become spaceports. We have yet to see how hard it is to build a high throughput spaceport. Will there be on the order of 10s, 100s, 1000s of spaceports? Will there be some economies of scale where it makes sense to have relatively few ports? It’s expensive to move the first stage of rockets, so maybe it makes the most sense to colocate the relatively few rocket factories and launch facilities. This appears to be SpaceX’s plan in Texas and Florida. In the limit though, I’m skeptical that ocean transport to any coast wouldn’t be affordable even if land transport is hard. Still, land around Brownsville and Cape Canaveral is almost certainly undervalued. More important is going to be access to cheap natural gas and cheap LOX as the reusability of rockets increases and marginal costs dominate. In the short term this heavily favors Texas, in the long term solar synthetic methane and O2 wherever it’s sunny. Cities will not cluster immediately around spaceports because of how loud a high utilization port would be (15 mile distance from city to launchpad is common with today’s low utilization pads).

I don’t think drone delivery will change much about the urban form and I don’t think hyperloop will happen.

Future Resources

Materials

The majority of future materials for the SIR will be the same as the current materials, just in larger volumes. Unique access to materials will spawn mid-sized cities, but historically large cities don’t grow purely because of this. A substantial materials/manufacturing hinterland is necessary for port cities to grow however.

Energy

Just as coal and oil transitions sprouted new cities, it seems likely that the SIR will as well. As the cost of solar falls, transmission makes up an ever larger portion of the cost, putting a floor on the final cost of energy. It seems not unreasonable that solar is so cheap that it once again makes sense to colocate the load with the supply. We’ll return to the trend started at Niagara: cities and industry springing up around cheap energy.

Geographically, where will the cheapest energy be? The obvious places to look are deserts. They’re definitionally sunnier than anywhere else (and more importantly less cloudy than anywhere else). They’re largely empty, meaning the land is cheap (important for solar’s massive footprint) and the difficulty of building huge industrial agglomerations is minimized.

Freedom to choose where to live

In addition to logistics considerations for city siting, people have increasing freedom to choose where they want to live. Beautiful places with great amenities will do well in the long term. As robots take over undesirable work in undesirable locations, the successful cities will increasingly be the ones that are great places to live in. There’s a strong argument to be made that Silicon Valley is where it is just because Shockley’s mom was sick and liked the weather. Intel would fly its engineers out to an industry conference on the East Coast in January to remind them of how good they have it.² And the US has seen the revealed preference of living in sunbelt cities ever since air conditioning was widely deployed.

Source

Where will the great cities of 2100 be?

Under the assumption of the SIR, the highest potential future cities should be great places to live, have a unique logistics advantage, and have access to cheap solar energy.

The industries in which energy is a large portion of the product cost often have low value per unit mass. The unit economics of these products is hugely influenced by transport costs. High transport costs will almost certainly cancel out the advantages of cheap energy.

Most deserts have the energy, but are not great places to live and have no particular logistics advantage (No waterways obviously, and roads/railroads can be built on any reasonably flat land). Also, cities and industry consume tons of water for their operations and these areas definitionally do not have much water. There are some exceptions however.

Temperate Highland Deserts

Most deserts are miserable temperature-wise, but not all of them. There are some high altitude deserts in tropical latitude bands, like the Mexican Plateau or Ethiopian Highlands or the Altiplano or Arabia Felix. In the tropics, temperature is generally stable year round, but hotter than is comfortable. Higher elevation means both a stable and mild range of temperatures. As well as having cheap energy, these regions would be great places to live for their temperature and natural amenities. Unfortunately they fail massively on the most important criteria for what makes a city: logistics. Mountainous regions are at a huge disadvantage from a logistics standpoint; mountain roads/railroads are very expensive to build and operate, and inland waterways through mountains are almost never navigable.

Temperate highland deserts check 2/3 boxes, but fail at the most important one, logistics

Coastal Deserts

More interesting to me are a rare phenomenon called Eastern Boundary Upwelling Systems, which occur only 4 places on earth. Usually associated with some interesting marine biodiversity, these coastal regions have the side effect of producing near perfect year-round temperature ranges. A cool ocean current from the poles stabilizes the temperature of coastal areas. This is the mechanism that creates the famous California climate, and it happens in 4³ other places on earth.

This phenomenon is present at every blue arrow less that 45 degrees latitude

Prevailing winds blow east to west from 0-30 degrees latitude, and west to east from 30-60 degrees. These winds, among other things, create 5 major ocean gyres. The current that pulls cold water towards the equator is always on the east side of the gyre and on a continent’s west coast. Generally, these coastal regions are deserts below 30 degrees in latitude and lush north of 30 degrees due to the above prevailing winds (eg NorCal vs Baja).

Prevailing wind transition zone

The most comfortable temperature bands are usually in the desert portions of these west coast regions.

Source

Chicago for scale. Endless summer

These 5 coastal deserts scattered across the globe have almost identical—and idyllic—temperatures. These regions are utterly bizarre; there is nothing like them anywhere else on earth. They’re touching the ocean but are bone-dry and get almost no rainfall.

Chilean Coast

Being on the coast fixes both the water problem and logistics problems of the temperate mountainous deserts. Ocean shipping become attractive, dramatically lowering transport costs for industry. And desalination/coastal fog unlocks unlimited water availability.

The Options

Coastal deserts have 3/3 of the important things for a future SIR city. Let’s introduce a 4th condition, materials access, to narrow down the remaining 5 locations. The SIR will first affect energy intensive materials processing, so this seems like a sensible condition to determine high-potential locations.

I believe any of these 5 regions have the potential to host megacities, and likely all of them will given time, but it’s worth narrowing down to the best option.

Namibia is extremely sunny, but is a poor, corrupt, low population African country with minimal economic activity and weak institutions. There is next to no infrastructure on its coast; it’s almost entirely barren sand dunes. There is a small mining sector, but materials opportunity is limited.

Western Sahara is also basically completely empty and isn’t even a country. It’s kinda sorta run as a part of Morocco. It again has weak institutions, minimal infrastructure, and a tiny economy with negligible materials flux across its coast. Doing business there is difficult.

California’s coast is either too mountainous or built out already. There is not a huge industrial opportunity here. Despite being inland and not under the influence of the California Current, I’m bullish on the Mojave and Colorado river region. It’s the sunniest part of the world’s largest economy, its hinterland has significant materials extraction, and the Colorado river can be made navigable up to Nevada. Baja has some potential, but it is mostly mountainous, lacks strong institutions, and the climate is mostly not stabilized by ocean currents.

The above options all have 3/4 of the our imposed conditions. The following two have 4/4.

Unfortunately, the farther north you go up Australia’s coast, the less stable (and hotter) the temperature gets. The sunniest parts of Western Australia are the inverse of the parts with great weather. But it’s a country with strong institutions and it’s fairly close to the world’s population/economic center of gravity and fastest growing region (Asia). There is immense amounts of sunny land. There is a gigantic mining industry and great infrastructure. Western Australia is a great option.

My personal favorite however is Chile. It checks all 4 boxes: great logistics, great weather and natural amenities, the highest solar irradiance on earth, and a massive mining industry. Additionally, it is politically stable and economically developed, has relatively low crime and corruption, and has healthy trade relations with China.

Chile Deep Dive

Most of the above regions could host a city anywhere along the coast, but I can tell you exactly where Chile’s solar city is going to be. This is useful from a land value appreciation perspective; if you could buy Manhattan for $1000 in 1626 you absolutely would. We know exactly where to buy low in Chile, whereas there’s a good chance you get it wrong in the other regions.

The entire coastline of the country looks like a desert version of California’s Big Sur. It’s nearly impossible to build any substantial industrial agglomeration on, despite the flatness of the interior plateau.⁴

Except for a single break in the mountains.

This is the Mejillones peninsula. 400 square miles of flat, undeveloped land astride a three thousand mile largely unbroken mountainous coastline. This is the gateway to the Atacama and the Altiplano, the sunniest region on earth. It’s already home to the largest copper⁵ and lithium port⁶ in Chile, a major railroad junction, and a major airport. It has access to Antofagasta, the richest city in Chile with 400k people.⁷

If the solar industrial revolution ends up favoring the sunniest places on earth, this is the only realistic option for a sizeable city on the ocean with access to the interior. The only other option is Caldera, 250 miles south, with significantly less infrastructure and a much more mountainous interior not immediately suited for solar arrays.

This sunniness and lack of rainfall have caused multiple resource booms in the region. The first was the guano fertilizer boom, where the lack of rain causes immense islands of the stuff to build up, with the birds attracted to food brought on by the upwelling for thousands of years. Once this resource was depleted, the solution happened to be just inland, where this region’s immense nitrate reserves were found. These again built up because of the lack of rainfall eroding these salts away. Chile went to war and beat both Bolivia and Peru to claim this land, 10x-ing its national treasury over 20 years from the earnings. A century later, it was discovered that lithium salts were found in immense quantities in the same region. They’re the result of volcanic ash buildup and evaporation over thousands of years, leaving massive flats of concentrated lithium; essential for creating high performance and cheap batteries.

War of the Pacific

This region has well-developed transportation and industrial infrastructure because of 150 years of mining history. It is a region built for resource exports; solar industrial exports would slot in easily and shipping costs wouldn’t be a huge burden.⁸

The peninsula provides an ideal terraforming opportunity. As already established, the weather is comically perfect, with highs ranging between 61F and 74F.⁹ There is a complete lack of rain, but modifying the amount of water in a region is far easier than modifying the temperature. The standard answer is to desalinate immense amounts of water with cheap solar electricity, irrigate the peninsula, and turn it into a garden city. Alternatively, unique to this peninsula is a dense coastal fog with droplets too fine to coalesce into raindrops called the Camanchaca. The air will approach 100% humidity in the early morning, yet it never rains. From a non-expert’s perspective, this seems like an extremely promising target for cloud seeding. Alternatively, a simplified electrostatic precipitator — no more than a high voltage wire — shows promise in getting the suspended droplets to snap to it and fall as raindrops; a maximally low capex solution. With the temperature bands of this city, spectacular Mediterranean gardens where you can grow just about any plant seem inevitable. As the delta between indoor and outdoor temperature shrinks, a plethora of interesting architectural opportunities emerge; the dividing line between inside and outside evaporates. Courtyards, loggias, terraces, atriums, arcades are all wildly underrated. This will be by far the most dramatic example yet of humanity transforming its environment, and it isn’t even that hard. The driest place on earth will become Eden. Terraform can live up to its name here.

The peninsula today

Lake Como

The Sunniest Places on Earth are Underrated

The sunniness of the Atacama/Altiplano is ridiculous. You basically could not make a sunnier place on the surface of the earth. It is high altitude and the sky is blindingly clear year-round; locals wear extensive protection at all times to prevent inevitable sun burns. There are multiple weather stations here that have never recorded rain in the past half century. This paper estimates the LCOE of solar in the Atacama to be less than 1.8c per kWh. This can almost certainly be lowered significantly. Some back of the napkin math¹⁰ puts that as nearly identical in cost to US natural gas and Chinese coal on a thermal energy basis, which are about 1.65c per kWh, and Icelandic low-grade geothermal at 1.5c per kWh. The average global horizontal irradiation (GHI) (the amount of energy that hits a horizontal surface)¹¹ in the Atacama is about 7.2 kWh/m^2. That’s 10% sunnier than Namibia (the 2nd sunniest place on earth), 22% sunnier than the Mojave, 51% sunnier than the Texas Triangle, 77% sunnier than the Northeast Corridor, and 178% sunnier than the UK.

The advantage compounds with trackers. What stands out on this map?

Unlike Niagara Falls, there’s huge overhead on the amount of energy that can be supplied. The larger combined Atacama/Altiplano region has an average GHI of 6.67 kwh/m^2. At 800k km^2, every day this region is hit with 5.3E12 kWh of energy, or more than the US’s annual electricity consumption, which is converted to electricity at 20-30% efficiency.

More important than how sunny it is is how not cloudy it is. This region renders solar’s greatest weakness irrelevant. Backups for unpredictable long stretches of cloudy days in the winter? Those stretches simply don’t happen.

(GHI data from NREL, Extended America GOES East dataset 10 minute intervals, for the year 2023)

Taken from near Calama

The hours are in either UTC or GMT and are a bit odd b/c of that.

Taken near Edwards AFB. Note the killer cloudy stretch in early January.

The initial ramp is the clipped the end of day being brought to the beginning due to the weird time zone. The cumulative value at the end of the day should be unchanged however.

Taken from College Station, in the Texas Triangle

In addition to it being sunnier in absolute terms, the number of days where generation is significantly obscured by clouds is stunningly low in the Atacama. Literally thrice in all of 2023 did daily generation dip below 4kWh/m^2/day. Whereas nearly a third of days in the Mojave and half of days in College Station dipped below 4kWh. Say you can turn off your load no more than three times per year; what is the minimum kwh you can expect to collect daily under this assumption? It’s around 4kWh in the Atacama and 0.3kWh in the Mojave! These examples are not very serious real world considerations, but it should give some intuition of how much better the sunniest areas are than those just 22% less sunny.

It’s obvious why this is important. For nearly 24/7 uptime loads, backups of batteries and solar overbuild are necessary. Even the sunniest place in the US requires significant amounts of this. But the sunniest place on earth basically does not have this problem. A load that needs 4kWh/m^2 of energy to fill up its batteries will hit that goal 99% of days instead of something like 70% or 50%. Many fewer solar modules are necessary to fill the batteries to get through the night. For intermittent loads, the advantage of the Atacama is only linearly better as the average irradiance increases. For high uptime loads however, there is a superlinear relationship between irradiance and energy cost, with surprisingly large differences between places with only a 22% different in average GHI. Given time, as solar/battery costs continue to fall, high uptime loads will make up the majority of SIR loads. While Terraform has a unique competency of making low capex, throttling processes, the industrial economy as a whole does not do this. This strategy is great for Terraform from a market power perspective, but for maximally fast adoption of SIR power sources, near 24/7 loads are much easier for industry as a whole to build.

Intermittent or high uptime, the most favorable economics are always in the Atacama.This is wildly underrated. Unit economics for industrial processes selling into a commodities market are extremely tight and competition is cutthroat. Any marginal reduction in cost means a corresponding increase in profit, which means increased IRR, higher capital efficiency, faster growth, and a higher stock price. The siren call of the Atacama grows stronger as the ratio of energy cost to transport cost favors energy in a product’s unit economics.¹²

Additionally, for each new SIR technology, economic breakeven always happens first in the Atacama.As solar and batteries come down the cost curve, new processes will continuously become profitable. Being a first mover matters a ton in business; there are huge gains to being the first to commercialize a process. A staggering number of the largest industrial companies today were simply the first to do a thing in their region.¹³ It’s not unreasonable to think that all of tomorrows solar industrial giants will first commercialize their tech in the Atacama.

Furthermore, Mejillones will always be the node through which materials flow. It’s obvious that a city (or garden-variety land speculation) to capture the upside of this Cambrian Explosion of technologies is a good bet. As demand for the limited land in this area increases, so does the price. Assuming perfect market power on the land supply side (not really a reasonable assumption imo, but shows what happens in the limit), nearly all surplus profit from cheap energy cost-savings can go to the landowners. That’s us, if we play our cards right. In fact, it’s probably a better business than commercializing the technologies themselves.¹⁴ This city run like a business has economics similar to a venture capital fund, all you need is for a few novel SIR technologies out of 100s to hit it big and dramatically increase land value. Use these windfall profits and plow them into R&D to further speed up the SIR.

A major message of Terraform is that the SIR will be a global thing and a solution to climate change. And it will be, eventually. More interesting to me is the possible geographic concentrations of its effects. The idea of a single geographic location being somehow much better than anywhere else is very exciting. Definitional cornered resource. There’s way more alpha and potential for profit if geographic exceptionalism turns out to be true for solar economics.

The SIR will not happen first in the US

The transition to a solar powered industrial civilization will happen first where solar energy is maximally cheap compared to the alternatives. That is not the US.

The case for the Solar Industrial Revolution is very strong. Over the long term, our industrial stack will use the cheapest source of power available to it. The marginal unit of hydrocarbons is more expensive than the previous unit, but the marginal cost of solar is less expensive than the previous unit. It doesn’t take much imagination to see where this goes! However, it’s pretty obvious that the place where this will first happen is not in America. Solar power is maximally cheap in the sunniest places with access to cheap solar and batteries. US solar modules and batteries are notoriously expensive. There is significant domestic manufacturing (aside from cells), but the scale is not close to China’s and the prices aren’t either. US modules (domestically manufactured or imported with tariffs) are 3-4x the price of free trade Chinese modules. It’s unlikely that this multiple comes down given the current trade war. US batteries are less dramatically expensive, something like 30-100% more than Chinese ones. US natural gas, however, is about half the cost of Chinese natural gas, and a third the cost of the world LNG price. The cost calculus in favor of solar over natural gas as a power source in the US will happen much later than in other places. The US has uniquely some of the most expensive solar power in the world and the least expensive natural gas in the world. Methane is too cheap, tariffs too high. It is probably the worst place on earth to run a SIR business. Even you believe we’re already at price parity in the US, the math is even more in favor of the SIR approach outside of the US, where you can presumably make massive profit margins and scale much faster.

I love the US and believe in time we can build up our domestic manufacturing base, particularly in solar and batteries. Eventually we can reach price parity with China. But that is not yet true. For a company hoping to be on the leading edge of the SIR, deployment first in the US is obviously nonsensical (particularly if your flagship product is natural gas, one of the few things we’re better at making than literally anyone else on earth). It makes sense to look elsewhere. Sunny places with cheap imports and expensive energy alternatives. Chile is ideal. It has a free trade agreement with China that gives them access to the cheapest modules and batteries, it’s the sunniest place on earth, and it has extensive energy-intensive industrial activity. Australia, Saudi Arabia, and the UAE are other good options, but all are net natural gas exporters already; competing with them on energy cost will be hard. Chile on the other hand already imports LNG and Argentinian gas, and their pipeline network already runs through the Atacama.

Summary

The single highest potential location for the Solar Industrial Revolution’s City of Tomorrow is the Mejillones Peninsula in Chile. This location has the cheapest energy on earth, perfect natural amenities, excellent logistics infrastructure, and a substantial materials economy. Nowhere else on earth has this ideal a combination of traits.

Privileged access to the sunniest region on earth has non-obvious and substantial implications:

For either intermittent or high uptime applications, the most favorable economics are always in the Atacama.Profit maximizing companies will line up at the door once they realize this.

Additionally, for each new SIR technology, economic breakeven always happens first in the Atacama.Like Niagara Falls, there will be a proliferation of novel energy intensive technologies in this one location.

Furthermore, Mejillones will always be the node through which materials flow. As the SIR advances, so too will the city built on this peninsula.

The economic implications of this region’s characteristics are clear. Terraform should strongly consider pursuing this opportunity.

Water is still vitally important as infrastructure, but today it heavily favors the largest boats. This means that ocean shipping and deep inland waterways remain important, but the rivers that originally spawned many cities are no longer used.

“Intel used to send employees to an annual semiconductor conference that took place on the East Coast during the winter, in part to keep them up to speed on industry developments and in part to remind them that if they were tempted to leave for IBM or another East Coast competitor, they could say goodbye to sunny Santa Clara.”

Byrne Hobart. Byrne Hobart_ Tobias Huber - Boom _ Bubbles and the End of Stagnation (2024, Stripe Press) (p. 117). Kindle Edition.

This also happens in Western Australia but there is no significant associated upwelling

Pumped hydro here could be interesting.

I noticed that you didn’t list copper on your board of primary industries to conquer, but my understanding is it’s worth looking into. It has fairly high value per unit mass, and is fairly energy intensive to process. (Though not necessarily low capex and throttleable. Still Chile a unreasonably great option for 24/7 loads)

Chile produces 5.3m metric tons and China produces 1.8m metric tons of copper, but Chile refines 1.2m metric tons and China refines 7.9m metric tons. This is presumably because of China’s cheap coal. Because of the weight-loosing properties of the copper supply chain, there is alread an economic forcing function to refine it as close to the source as possible. Once China’s energy advantage flips to Chile, I can think of little reason for this not to be economically favorable. This seems like low hanging fruit.

This is also perfectly positioned for Chile’s extensive nodule mining prospects

Enough people nearby that starting something is doable, unlike Africa (no demand, no infrastructure, no workers, no urban amenities when starting), but few enough that nobody is all that upset by what you’re building, like California

A place like Iceland (abundant, cheap thermal and hydro energy) struggled with shipping costs because there were few existing trade routes it could piggyback on when starting.

(Climate analysis)

Thermal cost comparison:

Assume $5 per thousand cubic feet US natural gas (lower than the average Henry Hub spot price, but transport cost will increase the delivered price)

1036 btu per cubic ft

207200 btu/$

3412 btu/kwh

60.73 kwh/$

or $0.0165/kwh

Assume $110/ton chinese thermal coal

1 ton = ~23m btu.

3412 btu/kwh

$0.0164/kwh

Coincidently nearly the exact same

https://geocom.geonardo.com/assets/elearning/5.16.1999-Geothermal-District-Heating.pdf

This puts icelandic district geothermal at $0.015/kWh

GHI was chosen for these analyses because it seems obvious that in the limit as costs decline, solar will look like more like paving the desert with solar arrays than farms of tilted and/or tracking panels.

Something like the Siemens process (aside from it’s deep integration into local semiconductor supply chains) seems maximally attracted to the Atacama. Energy is a large portion of the unit costs and the sale price per kg (specific value) is quite high, meaning shipping costs are relatively low.

Linde, Shell, Samsung, Siemens, GE, Saudi Aramco, Tata, Reliance Industries and many more.

On priors, real estate has been a much better business than industry over the past 50 years. I’m not convinced this dynamic isn’t just downstream of the Great Stagnation though, and could cease to be true as the stagnation ends. Likely housing values will fall, but land is fundamentally scarce and will probably increase in value in line with the size of the economy at least.