Comment by citrin_ru
2 days ago
> Even the solar panel market is self defeating. Once there is enough installed power the demand will drop off sharply as the refresh cycle is too long.
It's not going to happen soon - solar is still just 8% of world energy production. Even if solar will cover 100% of consumption on a sunny day it still would make sense to buy more panels to have enough output on a cloudy day or in the morning/evening. It's likely production of solar panels will be a good business till at least 2050 and oil business will start to decline before that unless will be propped by corrupt politicians.
8% of electricity, not total world energy.
But the growth rate has been huge for as long as records have been kept, and was a factor of just over 10x between 2014 and 2024, speeding up more recently.
PV and wind together are likely to start breaking the electricity market severely in the first half of the 2030s; I hope, but it's not certain yet, that ongoing battery expansion will allow the demand for electricity to increase and this can continue to the end of the 2030s, because at the current pace of development those scale up to all our energy needs, not merely our present electrical needs, in a bit less than 20 years from now. (PV alone would do all of it in 20 years at present rate of change).
Energy use goes up as civilization advances, and Jevon’s paradox suggests that we’ll use more energy as its cost goes down. Couple that with the need to replace some portion of the installed base of solar capacity over time and I think solar will be a growth industry for the foreseeable future.
I can't believe it's taken this long for someone to mention this. Even just phasing out fossil fuels (if we're still serious about that) plus ordinary growth means today's demand is a fraction of what could potentially be fulfilled by additional solar buildout.
It also assumes that there will never be demand for improved solar generation orthogonal to currently-prioritized metrics. As an example, a nice park near my house was clear-cut to install a solar farm a few years ago. I used to enjoy walks under the trees in that park, and seeing the animals that lived there. Perhaps as solar infrastructure becomes more stable and secure, concerns will turn towards the ecological ramifications of covering so much of the Earth's surface with ecological deserts, and there will be a desire to replace older generations of solar panels with ones that somehow can support or integrate more elegantly with nature. And then the next thing. And then the next thing.
Assuming we consume ~20 TW on average, a metre-squared panel kicks out ~40 W on average, and we halve that to account for batteries and other infra... I reckon we're talking about 1 million square kilometres (people will be along in a sec to check my working, but it's just a Fermi estimate).
Call it 10% of the Sahara.
Bear in mind that if we go all-electric, raw energy consumption falls significantly, many panels will be sited on buildings, solar isn't the only renewable, and solar farms aren't ecological deserts - you can graze animals below them.
Honestly, seems like a good trade to me.
I'm not saying that that magnitude of solar generation isn't a good thing. I'm saying that the solar farms of 2050 don't necessarily need to be arrays of panels on top of clear-cut land.
Grazing land is often essentially an ecological desert when compared to previous uses. Farms in general, honestly. Actually, this is a good forward example as agricultural expansion goes hand-in-hand with the Anthropocene die-off, but late advances in land use efficiency via fertilizer and other technologies means that even though these lands are super dead we also require less of them per person. What I'm proposing is analogous to even further development, where you're still somehow able to produce the same volume of food while reintroducing ecological diversity to the same land; moving away from traditional monoculture farms to ultra-efficient food forests. I don't know how you'd do it in farming, but it energy generation, it would probably involve engineering equipment to some level of symbiosis with the preexisting environment. Could we someday build literal forests of photovoltaics that support energy generation as well as a diverse natural ecosystem? Maybe. I'm sure we'll try. And that's why, ultimately, my point is that the idea that solar is an economic dead end is incorrect. This is just one potential branch on a tech tree (heh) that isn't anywhere near done growing.
> concerns will turn towards the ecological ramifications of covering so much of the Earth's surface with ecological deserts
About that "solar panels cause ecological deserts" trope:
https://www.sciencedirect.com/science/article/abs/pii/S03014...
https://glassalmanac.com/china-confirms-that-installing-sola...
https://en.wikipedia.org/wiki/Agrivoltaics
>Traditionally, deserts have been seen as harsh, lifeless landscapes
This is incorrect, depending on the geographic location. Many "deserts" are actually ecologically vibrant, and "greening" them (especially for farming) threatens to destroy a measure of natural diversity.
That said, I think you and the other poster placed emphasis on the wrong part of my post, as my point was less about solar land area coverage as some sort of singular evil, and more about the *opportunity* present in continuing to develop solar technologies so that they impact the environments they're placed in less and less over time. This would mean that efficiency is not the be-all-end-all of development, and that further improvements are possible even after reaching a satisfactory level of efficient generation. The energy economy would not fall off a cliff, as some predict. It would simply shift to solving other problems.
You can see an example of this in computer engineering, with Moore's Law's fall-off and the rise of GPU-based innovation.