Solar power has begun to transform the world’s energy system

Do you have any source for this extraordinary claim? It's practically a claim of perpetual motion.

Carbon dioxide a tiny fraction of the atmosphere, even in concentrations which are immediately harmful to human life.

At the moment it's 400 parts per million. So in order to extract 1kg of Carbon Dioxide from the atmosphere you have to pump 2500kg of air through the system. This alone makes it unlikely we can do this profitability.

You then need to extract the carbon dioxide using some technique which will probably involve cooling or pressuring that volume of air. Before finally transforming carbon dioxide, a very stable chemical compound, into a reagent which is actually useful (probably carbon monoxide).

Difficult engineering problem but working from first principles suggests that the energy requirememts are not insurmountable. The roundtrip efficiency is worse than batteries but much better than photosynthesis.

Terraform Industries (and others, like Synhelion) has a plausible if slightly optimistic target to be price competitive with fossil fuels for methane in the early 2030s.

Some places with very cheap to extract hydrocarbons like Saudi Arabia may be able to compete for a very long time, but there are many futures where most of humanity's hydrocarbon consumption (including the ones used for the chemical industry, plastics, etc) derives from atmospheric carbon.

And this can happen fast, the world (mostly China) has developed a truly massive manufacturing capacity for PV.

They said "If electricity is sufficiently cheap", which is less a claim and more a tautology.

Will it be that cheap? I think so, given that trees and grass etc. exist and get their carbon from the air.

Even with free electricity, the capital (and maintenance, consumables etc) costs of the process could easily be too high.

I suppose that in principle that is indeed possible; in practice, trees exist and self-seed, so the limit is our own ignorance.

We are also limited the incentives to that make us cut tree for money, and not develop technologies if they are not profitable within a short time-window. We have the technology to plant more trees right now, but we aren't.

Terraform Industries (and others); I'd seriously consider taking a long bet that these companies turn out to be better at converting investor capital into employee salaries, for a finite period of time, than they are at converting atmospheric CO2 into natural gas.

If such a technology was possible then it would be far better to start with carbon capture from existing emitters. The concentration of CO2 being easily 3 orders of magnitude higher.

You could make the same argument about AGI. Just because nature does it doesn't mean it's easy for us to replicate in an industrial setting.

People plant loads of trees for lumber, but you're right, it's an economics question in the end.

This actually means I'm also worried about something currently impossible: that when we do develop the tech sufficiently to be useful, if it's cheap enough to be profitable, nothing would seem to stop extraction. So CO2 goes down to, what, 300ppm? Pre-industrial? Ice age? Same coin, other side. We want to flip a coin and have it land on the edge.

A single world government could organise to fix this either way, but as all leadership roles come with the risk of the leader being fundamentally bad, this isn't something I'd advocate for either.

Sure, but you said "It's practically a claim of perpetual motion." which is overstating the challenge to a much greater degree than this understates it.

For hydrocarbon synthesis, hydrogen production from electrolysis dominates the energy usage, along with driving the Sabatier process. DAC might be like 5-10%.

Higher CO2 concentration is better but certainly not needed, it doesn't make or break the economics.

I'm not going to argue over the numbers but any business which ignores such an obvious upside / upside scenario is not really serious about achieving economic criticality. It would allow a power plant, iron ore plant, cement producer, what have you, to make claims about their environmental credentials while simultaneously improving the efficiency of the process.

> that when we do develop the tech sufficiently to be useful, if it's cheap enough to be profitable, nothing would seem to stop extraction. So CO2 goes down to, what, 300ppm?

This is an extremely improbable scenario, for several reasons:

1) If you actually use the extracted CO2, then it gets re-emitted on use, and the atmospheric concentration is virtually unaffected.

2) Concentration difference alone makes it very unlikely that we'll ever extract CO2 as cheaply as O2 from ambient air (or carbon from a mine), and CO2 is not really an appealing ressource compared to its components, either (so demand would presumable be pretty low for centuries, even if the price comes down a lot).

> 1) If you actually use the extracted CO2, then it gets re-emitted on use, and the atmospheric concentration is virtually unaffected.

Depends what you use it for, e.g. synthetic diamond windows won't re-emit unless they catch fire.

> 2) Concentration difference alone makes it very unlikely that we'll ever extract CO2 as cheaply as O2 from ambient air (or carbon from a mine), and CO2 is not really an appealing ressource compared to its components, either (so demand would presumable be pretty low for centuries, even if the price comes down a lot).

Underestimating how big an industry would get is the mistake Svante Arrhenius initially made, thinking it would take millennia to emit enough CO2 to cause noticeable global warming.

And remember, with this concern I'm inherently presuming tech (mainly energy) that makes it sufficiently cheap that business and/or governments are willing and able to remove in the order of at least one teratonne of the stuff (but hopefully not two or more teratonnes) — because less than that, it's not solving global warming.

Consider a chemical synthesis that needs carbon. Right now it uses oil. But is has to be extracted and transported. With carbon capture from the air that no longer required. And maintaining the extra facility at the chemical factory can be cheaper than maintaining the extraction and supply chain for oil or coal.

It may never be “worth” it in economic sense, but it offers a way to separate the time of “energy is used” and “energy is available”. Assuming sufficient captureable volume you could capture the emissions of a fossil power plant during the ~two weeks per year where weather is sufficiently bad. And then take the other 50 weeks to capture that carbon again. It can be completely inefficient (like sub 5% round trip efficiency) if a) we pay for it via a capacity market and b) have sufficient excess (clean) energy to run it

Yes. But it still has to compete with all the various types of biomass, for example.

The idea that we'll have huge excesses of clean energy seems like wishful thinking. We may have issues with excess energy at certain times of day for sure. But intermittent excesses like that are difficult to make use of economically because of capital costs and low utilisation. A general excess would be countered by falling energy prices to the point that it's difficult to make a business case for new installations.

I don't see a future where technologies which are massively inefficient reach their break even cost before other energy intensive activities or more efficient grid scale storage soak up the excess.

Of course it seems like wishful thinking! Because that’s the historic norm. Energy was (in some way) always the limiting factor. And every time energy access got meaningfully cheaper society massively reorganised around it.

And yes “energy” in general won’t be free. We still need to build the generation and distribution systems. But we reached a point where just dumping solar on all _new_ roofs rounds to essentially free (the costs are the labor and the access to qualified personnel). The exact same is currently happening to batteries. Any transformer project will be able to just integrate 4-12 hours of batteries without getting meaningfully more expensive. The same for every domestic or industry service upgrade

We are not there yet. But give it another 5 years and we will. And then we are only talking about financing what little distribution system we will need (basically you only need average-sized cables not peak-sized ones) and a capacity market for backup power systems (also only for average residual demand). And those we simply cannot (efficiently) finance by a per-kWh-used charge