As useful context, world marketed energy consumption is about 20 terawatts, of which about 5 terawatts is China. Ground-level terrestrial insolation is about 128000 terawatts.

In this press release, the Global Solar Council is estimating an average 20% capacity factor, which means 2 terawatts of installed capacity amounts to 0.4 terawatts of average power. 20% seems like a broadly reasonable capacity factor; https://en.wikipedia.org/wiki/Capacity_factor#Capacity_facto... shows that from 02013 to 02018 the PV capacity factor in the US varied between 25.1% and 26.1%.

However, yesterday I was looking at CREA's statistics on China https://energyandcleanair.org/wp/wp-content/uploads/2024/08/.... Their plot seems to show 40 TWh generated from solar in June. 40 TWh per month works out to 55 GW. If we assume 670 GW of solar generation capacity was installed as of June (of the current 790 GW in https://www.nea.gov.cn/2024-11/22/c_1310787342.htm, which has grown 48% over the last 12 months) that's an 8% capacity factor, which is implausibly low—when I've done the stats for previous years in China, the capacity factor was about 10%, which was already very low.

Am I calculating this wrong? Are there better numbers on China's solar capacity factor? Why would it be so low? Will it improve if, e.g., more HVDC transmission lines or battery storage are built?

I note that some recent Chinese-built solar farms abroad like the Gobustan farm in Azerbaijan (60km from CoP29) https://www.youtube.com/watch?v=qIJA5hHHawA are using single-axis sun trackers, which you'd expect to raise the capacity factor significantly over the 29% achieved by utility-scale solar in California, which is almost entirely fixed in place and thus unable to track the sun. [Edit: pfdietz says this isn't true! https://news.ycombinator.com/item?id=42266903 https://emp.lbl.gov/sites/default/files/2024-10/Utility%20Sc...]

Module prices have dropped about 70% (more than a factor of 3) in the last two years, to €0.10 per peak watt https://www.solarserver.de/photovoltaik-preis-pv-modul-preis... since the apparent failure of the price-fixing cartel announced at Davos five years ago https://www.reuters.com/article/us-davos-meeting-solar-gcl-i...

It's understandable that financing is tricky in the face of such rapid change. It seems likely that most solar farms built two years ago will owe too much money to ever be profitable due to competition from much cheaper PV modules—either in newly-built solar farms or in difficult-to-stop distributed generation. As toomuchtodo points out in https://news.ycombinator.com/item?id=42267156, this is already a significant factor in Pakistan.

> utility-scale solar in California, which is almost entirely fixed in place and thus unable to track the sun.

This is not true at all. Most utility scale PV in California uses trackers, I believe single axis.

https://emp.lbl.gov/sites/default/files/2024-10/Utility%20Sc...

https://emp.lbl.gov/technology-trends

As I remember from the times we did solar installs in Tanzania, 8-10% is about right for PV in real life. It's the 20+% figures look like creative accounting.

>If we assume 670 GW of solar generation capacity was installed in June

Was installed /as of/ June. (sorry I was double checking your workings and got confused)

Assuming the figures aren't typos.

1) What is counted as generation capacity? a 'finished' solar farm that is waiting for paper work / connection to the grid could be included in the figures.

2) What about solar that doesn't see the grid? the panels on my roof might be included as installed capacity, but not show in generation figures.

edit: 3) another commenter mentioned solar with pumped hydro. Is that classed as pumped hydro in the generation figures?