Crossed Wires: Solar energy in the age of Musk and Trump

‘Solar is within a hair’s breadth of eclipsing most, if not all, other energy sources in terms of cost per kWh and, if innovation continues at the same pace, possibly only a decade away from surpassing them across every metric in the energy value chain,’ writes Steven Sidley in this week's Energy Report.

This thought led me to take another look at the solar industry, to get a sense of how much wind will be sucked out of its sails (to wildly mix metaphors), or the opposite, given the energy demands of new technologies like AI.

My, my, how things have changed since I last looked.

But first, I had an argument. A friend of mine, a well-known curmudgeon, got wind of my preparing to write this article. He protested vociferously, claiming that solar power is the greatest energy boondoggle ever foisted on a gullible public, a massive waste of money inspired by misinformed optimism. He besieged me with papers and articles all making the same argument. I read them all and then did a bunch of research. A greater contradiction between his research and mine you could not imagine, although I suspect my friend’s objections to solar energy have more to do with the taint of liberalism than the pure science and economics of the matter.

What I found was this: solar is within a hair’s breadth of eclipsing most, if not all, other energy sources in terms of cost per kWh and, if innovation continues at the same pace, possibly only a decade away from eclipsing all other energy sources across every metric in the energy value chain.

As a side note, I used OpenAI Deep Research, Gemini Deep Research and Perplexity Deep Research to crawl through mounds of research on the subject and it was an education in and of itself. I wanted to steer clear of any research that might be coloured by politics or commissioned by ideological paymasters.

This turned out to be easier than I thought. I specifically asked the AIs to concentrate their efforts on neutral research — to exclude industry-funded efforts (renewable or fossil) or policy papers. The AIs helpfully provided their sources so it was easy to check the authorship and provenance of the research. The result was numbers from generally reliable sources like the IEA (International Energy Agency), BloombergNEF and Lazard’s Levelized Cost of Energy. The AIs correctly ignored sources like the Intergovernmental Panel on Climate Change which is, after all, a policy-making institution and not a scientific body. Quibblers will still quibble but at least some filters are better than none.

Most of the charts and stats were startling. From a purely economic perspective — ignoring externalities like the availability of the raw materials to make solar cells and issues such as grid connection and deep engineering electrotech arcana —we find that the cost per kWh has fallen vertiginously over the last few decades, and continues to do so. Here are some highlights:

• Since 2010, the cost of utility-scale solar has gone from $0.40 per kWh to less than $0.04. That is more than 90%. That is also in line with Moore’s Law, a prediction of exponential accelerating efficiency originally applied to the number of transistors on a chip. Moore’s prediction was accurate and we ended up with a computer in every home and a phone in every pocket in just a few decades.
• The cost of utility-grade solar energy is now less than half that of the coal equivalent. Solar costs continue to fall, driven by tech and manufacturing innovations. The production costs of coal, oil and gas have not fallen at all in the last 10 years and are unlikely to do so now.
• BloombergNEF analysts report that in 2016 it took a week to install 1GW of solar. In 2025 there will be 600GW installed, which is 10x the 2016 rate.
• According to The Economist last year, the International Solar Energy Society reported, “Solar power is on track to generate more electricity than all the world’s nuclear power plants in 2026, than its wind turbines in 2027, than its dams in 2028, its gas-fired power plants in 2030 and its coal-fired ones in 2032.” (Yes, I know, this one violated my neutral source principle but even so…)

The numbers and charts are endless. They include manufacturing efficiencies, battery efficiencies, investment and customer demand. Plot a graph showing increasing demand, increasing product efficacy, increasing output and falling costs at this scale and it provides a mouthwatering proposition for investment, which is the fuel for innovation, in the US as much as in China.

The battery world?

Solar needs its sibling (energy storage) to see it through dark times. So what is happening in the battery world? The numbers look like solar. Over the past 15 years, lithium-ion battery costs have seen one of the most dramatic declines in energy history, falling by more than 90%.

The IEA notes that solar PV coupled with batteries is now cost-competitive with new coal plants in India. Add to this the application of AI to materials science in general and battery science in particular. Already, a novel electrolyte that uses 70% less lithium has been discovered by a Microsoft-funded AI project (gnarly details here). There is much more good news, but you get the gist.

How close does all this come to being a real on-the-ground benefit for the rest of us? Last year I installed a solar system at my home. The monthly cost of my electricity, even taking into account the capital cost of installation, is less than half the price charged by our local utility. I have not bought any electricity (beyond a small prepaid purchase) for six months. And installation costs continue to fall, projected to be 15% lower in 2025.

Are there headwinds coming? Of course. Everything from manufacturing economics to the longevity of photovoltaic cells, to the technical vagaries of solid state energy vs things that spin (turbines), to the deflation of subsidies (which will knock some smaller players out of the game), to the rarity of some of the base mineral ingredients needed for the kit. But these hurdles looked insurmountable just a few years ago. Not any more.

Many detractors of solar energy point to subsidies as the reason for the rosy projections, claiming that the industry would collapse without them. Certainly, US subsidies are not going to continue. But you only need to point to the Clayhill Solar Farm in the UK, a privately funded company which is already profitable without any subsidy of any kind. It was first over the line and has solidly proved the commercial case.

What about AI and its robust energy demands? This too was a surprise. The big energy hogs, the AI behemoths like OpenAI and Anthropic which require massive “utility-scale compute” for training and inference, are fairly small in number — perhaps 10-20 globally. They are heading over to nuclear where they can lock up reliable 24-hour, predictable, uninterrupted baseload.

But while the future compute needs of more general-use data centres are robust, they are not overwhelming. They currently demand about 5% of the global energy supply and are growing steadily, but not unduly so. There are already 400 hybrid solar-powered data centres globally, and 100 that use solar exclusively. Fifteen percent reductions in operational costs (compared with the cost of utility power) are common. These data centres will be part of the solar demand equation.

In 2023 I outlined some of the hidden troubles facing the renewables industry (you can read my article here). In solar at least, things have changed, even considering some of the anti-renewables research my curmudgeonly friend sent me (the most complete of which is here).

Notwithstanding Musk and Trump and their contempt for climate change, the solar industry is now breaking free of the tether of politics and culture and arguments about the climate. It is on its own trajectory, and the financiers and entrepreneurs who will reap the bounties of solar in the future no longer have to concern themselves with any of that noisy debate.

Solar is now simply a better mousetrap. Everyone will benefit, even the Earth. DM

This article first appeared in Daily Maverick here. Steven Boykey Sidley is a professor of practice at JBS, University of Johannesburg, columnist-at-large for Daily Maverick and a partner at Bridge Capital. His new book “It’s Mine: How the Crypto Industry is Redefining Ownership” is published by Maverick451 in SA and Legend Times Group in the UK/EU, available now.

Recommended newsletters:

• First Thing: Your daily update on the news that broke while you were sleeping. Sign up
• Business Maverick: Daily business news and analysis from the finest minds at Daily Maverick. Sign up