China's Energy Transition at Odds with Solar Glut, Cheap Power

While yes, too much surplus capacity would eventually indicate overinvestment, for now it seems like great news for the compute and software industry in China. If you reliably have times when energy is better-than-free, that's the ideal time to scale up and generate embeddings from your stored data, or train LLMs, or do all sorts of other batch-able but intensive tasks.

  > The deluge of solar power has thrust electricity prices into negative territory in regions with liberalized energy markets, which bodes ill for the economics of further expansion.

The market is sending California the same signal now, and they are responding accordingly.

There is no boding ill here. This is everything working as it should be.

1) You can just install solar capacity and even on the best of days that power will be used and simply offset CO2 production.

2) Reach a point where the peak is a exceeding the transmission lines and the mid day power curve is showing Solar meeting all demand and maybe exceeding it a little. Prices now drop significant on these rare occasions.

3) There is now enough Solar that with the right balance of storage its enough even in winter when aggregated across a large area and have effectively decarbonised completely.

During phase 1 its easy you just install panels.

Phase 2 however is a painful phase involving upgrading of transmission lines as power now comes from different places and in peaks and also the addition of storage to move power across the day and smooth out delivery. There are moments of free or negative price power and these are signals to the market that more storage is required.

Phase 3 involves overprovision of solar capacity in the summer but enough in winter combined with the storage to aggregate across the day. Its also going to need efficient long range transmission upgrades as sunshine has tendency to appear in big blocks.

People expect that the storage needs and overprovision will be a lot bigger than they actually need to be. There is an experiment for this in Australia which shows that about 120% solar and 1 days worth of storage is all it would take, which as a combination is vastly cheaper than their current Coal plants. The mix is going to be different for different countries dependent on the weather and irradiance but there is always going to be an optimal mix based on the relative prices.

What I think is interesting is what do we do with that 20% of "free" energy through the summer? I tend to hope we put it towards CO2 capture but I suspect its a also a good time to training AI, electrolyse for hydrogen and filter salt water. This business opporrtunity is going to increasingly appear as a natural consequence of green energy solutions, wind and solar both produce the same energy market effect.

  Chinese spending on renewable energy dwarfs its rivals, with over $800 billion — or 4.5% of its GDP — invested last year, according to BloombergNEF. 

That said - cheap electricity is a huge boon to any country - so while they may have some challenges it is far better than the alternative of too much power. Solar manufacturers are in for a rough couple years.