Links #9

Solar in Africa, the solar-industrial revolution, and much more

Mar 29, 2024

1.

Why Isn’t Solar Scaling in Africa? The article echoes a lot of the problems that other big projects face in Sub-Saharan countries, high-risk investment, political risks, legal problems, high interest rates, high cost of all goods due to it being in Africa, corruption, and the general difficulties that western agencies have operating in other countries.

I’m excited by solar in Africa, it feels like a perfect fit, there’s plenty of sun, demand for energy, and a lack of electrical infrastructure. But stories like this make me wary of large-scale solar installations.

One thing that does seem to be working is to give individuals solar panels. Individual electricity needs are quite small, most people just need to charge a few phones and run a few lights. That makes me wonder if an individualized rollout of solar and batteries might work better.

Solar and batteries are fundamental building blocks. People working with them directly can find solutions to their own problems. Could you bolt a solar panel to an electric scooter and let it charge during the day? Could a battery pack and a fan constitute a personal cooling device to combat the heat in Lagos? Can you use electricity to cheaply make fertilizer from the air (see point 3)?

If solar and batteries make people more productive, they could be a profitable personal investment, spurring grassroots adoption. This has the advantage that it’s already happening, it just needs more of a push.

As a side note, if large solar farms ever get off the ground near the Sahara desert, they may be able to increase precipitation there:

Mega solar farms could make it rain in deserts

2.

I’m going to quote this tweet by Jesse Peltan in full:

The turning point in the Industrial Revolution was not when we started mining coal. It's when we started mining coal - with coal.
Carting coal with horses wasn't a huge leap forward from carting wood. Coal-powered railways and coal-powered mining created the feedback loop that transformed the nature of industry.
Similarly, solar's true potential is not unlocked manufacturing solar panels with fossil fuels.
It's unlocked manufacturing solar - with solar.
Fossil fuels are a step in the boot sequence - as human labor and wood were for coal. They are neither necessary nor optimal for large scale manufacturing.
The materials for solar panels are ubiquitous. Silicon makes up 28% of Earth's crust. Solar really gets cheap when we minimize energy and logistics costs. That means producing panels with solar energy, where solar is strongest & installing them locally.
Electrification of the solar supply chain creates the feedback loop that leads to Type 1 Civilization.

We’re in the middle of a solar-industrial revolution that’s being overshadowed by AI and other advancements. Notice the resonance with points I made about protein engineering in the conclusion here.

A paper estimates that solar will be the cheapest form of energy virtually everywhere by 2027:

I think it could be even sooner as I doubt the paper takes into account recent advances. For example, Jesse Peltan notes that we’ve been increasing conversion efficiency and lowering material use of solar panels for decades. “Longi now has cells that use just half a gram per watt and are 26% efficient.”

Related: the scaling of solar costs with global PV production:

3.

With the solar revolution, every industrial process will have to be reformatted to take full advantage of cheap but intermittent energy, as I discuss here.

One process that will take advantage of this is nitrogen fixation, used to make fertilizer from air. There are two common forms of fixed nitrogen: ammonia and nitrates. Ammonia is made using hydrogen, so low hydrogen costs are important. Fortunately, there’s plenty of work that goes in to cheap hydrogen production and synthesis of ammonia using electricity.

Nitrates are made by oxidizing ammonia using what’s called the Ostwald process, this wastes all of the hydrogen we used to make ammonia by turning it back into water. But there’s an old process that makes nitrates directly from air using electricity! Could we revive it in the wake of cheap solar energy?

The authors of the paper “From the Birkeland–Eyde process towards energy-efficient plasma-based NO_X synthesis: a techno-economic analysis” issued a correction, which makes the process of NOx synthesis look even more cost-effective. A couple of quotes:

“… the cost of HNO₃ depends on the geographic location. While the market value is as low as 250–350 € t-HNO₃⁻¹ in some locations where the cost of transportation is minimal, the cost at remote locations (e.g., the interior of sub-Saharan Africa) can be multiple times that of the production cost^108,109 so that electricity driven processes may become favorable at higher electricity cost.

“… there is potential for decentralized HNO₃ synthesis, instead of importing HNO₃ to remote locations.¹⁰⁹ While the Haber–Bosch process suffers from a high CapEx upon scale-down to capacities below 50 t-HNO₃ day⁻¹, the plasma-based NO_X synthesis process may be scaled down more effectively (see Fig. 10). Hence, plasma-based NO_X synthesis may be used for decentralized nitrogen fixation”

More important than energy efficiency or conversion efficiency, the new process has lower CapEx relative to Haber-Bosch and the costs depend more heavily on energy costs. The cost per watt of on-site solar will continue to fall, and this might be a good solution for rural farmers. I know companies like Nitricity are already working on this, but I wonder if an open source effort focused on low capital costs, small scale, and actual engagement with rural farmers would have more impact. You can generate NOx by blowing air over the electric discharge across two wires, perhaps some hobbyists could make a dent in the problem?

Everything else

Very cool: trees stalling effects of global heating in eastern US, study finds. Considered in the context of the cooling effect of large scale irrigation, the temperature-moderating effect of bodies of water, and the cooling effects of solar panels it suggests that using more water and solar panels to support local ecosystems could help combat heatwaves.

The Plan to Build an Island Using Only Electricity details a failed plan to make seasteads by using solar power to produce a sort of concrete at sea. It would be very hard for a plan like this to be profitable and doing construction at sea is super hard, but it’s not infeasible to make concrete and structural metals like magnesium from seawater and solar power. Maybe one day hobbyists will self-fund a more serious attempt without the pretense of making a profit.

Hydrogen Might Self-Renew, Reservoir found in France. There might be geological hydrogen that can be extracted from the ground. We’ll see in the coming years how accessible these resources are, but this could be very valuable for the energy transition. Depending on where the deposits are, techniques from fracking and enhanced geothermal could help extraction efforts.

The Burrows–Wheeler transform is a reversible way to reorder a text string to compress it.

OmniPred: Language Models as Universal Regressors

3D Chip Tech Is Key to Meta’s AR Goals with wafer-to-wafer bonding discussed in this article and chiplets that allow manufacturers to assemble different hardware in a modular fashion, what’s stopping us from making huge 3D blocks of compute with built-in cooling systems?

A thread on Astro Mechanica’s plans to build a new type of jet engine that can improve space launch and revolutionize air travel. Their system pairs particularly well with eVTOL and space tethers.

Sparks of function by de novo protein design

Personality changes following heart transplantation: The role of cellular memory

Blocked Transmission. A nice piece on surface sanitation, an understudied component of biosecurity. It seems like spray-coating high-touch surfaces with copper would be a valuable, if expensive, step towards eliminating surface transmission.

Genetic insurance is close to what I imagine as an ideal for health insurance, something like how people would construct publicly-funded insurance from behind a veil of ignorance (people would still have to buy private insurance, but pre-existing conditions would be covered). These proposals involve buying insurance before taking a genetic test, but some conditions are obvious before the test and some individuals have observable characteristics that predispose them to certain conditions. Instead, we would want a state-run system that insures people at birth across as many people as possible, with payouts coming from taxes.

“First country with de facto 100% electronic money system is...Somalia?”. The developing world continues to be early adopters of new financial technologies.

Meshtastic is pretty cypherpunk. An encrypted (though not very secure) radio mesh network using open-source hardware and algorithms.

12 New Immigration Ideas for the 21st Century

Why you, personally, should want a larger human population

Braised Pilchard

Contra Australia

Mar 30Liked by Sam Harsimony

Coming from, Australia, the home of rooftop solar, I think it would be great for Africa. The homeowner owns the panels outright, is not dependent on infrastructure, and can expand the number of panels or add a battery as resources allow.

Expand full comment