At the end of October 2023, China’s NEA (National Energy Administration) reported China’s installed renewable energy capacity had exceeded 1.4 Terawatts. This capacity includes 420 GW of hydropower, 404 GW of wind power, 536 GW of solar, and 44 GW of biomass. China added 216.9 GW of solar capacity in 2023, marking a 148 % YoY increase compared to 87.4 GW in 2022. Major power generation enterprises invested CNY967.5 billion (~$151.17 billion) in power projects, representing a 30.1% YoY increase.
In 2022, China’s solar power generation reached 418 terawatt hours (TWh), a 20.9% increase from 2021. China had 392 GW of installed solar at the end of 2022. In 2022, the US had 110 Gigawatts of installed solar and it generated 204 TWh. China is averaging a little 1 MWh/year (2.8kWh/day) per kilowatt of installed solar. The US is getting 1.85 MWh/year (4.8 kWh/day) per kilowatt of installed solar.
It is a common mistake for people to assume that the kilowatts of installed energy are all equal. US nuclear power can get 8 MWh/year (22 kWh/day) per installed kilowatt. Other countries get 6-7 MWh/year per installed kilowatt of nuclear. Coal power can get about 4 MWh/year per kilowatt of installed power. Hydro can get about 3 MWh/year per kilowatt of installed power. There is also a lot of variability by country, locations and projects even for the same energy type.
The fast build out of solar power in China could take time to fully connect to the grid. China had much more installed solar power in 2017 at 130 GW than the US in 2022 but it took until 2019 for China to generate more electricity than the US using 110 GW.
Mongolia is a good location for solar power generation, with 270–300 sunny days per year. This is equivalent to 2,250–3,300 hours of sunshine. Mongolia has cold winters with snow and ice. The number of sunny days should not be the problem. Too much heat can also be bad for solar panels, reducing their efficiency by 10%-25%. China’s solar power only generates about 55-65% of the electricity for the same installed capacity as the USA. The fast buildout could have a lag factor before everything is connected. The China solar locations are probably 20-40% worse than the best places in the USA. China has dust and air pollution that is reducing the effectiveness of the solar power. China is choosing to mass install in north and northwest where the Gobi desert is located. China is also building solar in many other locations by the super-scale projects are in the Gobi desert.
China had 3.56x as much installed solar capacity but just over 2X in electrical generation.
The Kubuqi Base Project is a 16-gigawat (GW) solar, wind, and coal project in China’s Inner Mongolia Autonomous Region. It’s the world’s largest wind and photovoltaic power project developed and built in a desert. The Kubuqi base project is roughly the size of 20 Central Parks. China will build 450 Gigawatts of solar and wind power in the desert.
China’s solar power reached 610 GW at the end of 2023. 74 GW was activated in November and December. China’s company and government have put $130 billion into solar cell and energy production. This will capture about 80% of the global market. China will have losses and low margins while they generate a solar power glut while capturing market share. Solar module prices dropped 42% in the last year. They are now at $0.15 per watt. This is down to $150 per kilowatt. The US has prices for home installation of $3 per watt which is $3000 per kilowatt. The $0.15 per watt is 60% below the $0.40 per watt for wholesale utility scale US prices.
China will reach over 1 terawatt of solar installed capacity by the end of 2025. This will generate about 1100 TWh/year. China’s total power generation volume was about 9,360 TWh in 2023. China’s electrical power generation and demand will be about 9600-9700 TWh in 2024 and about 10000 TWh in 2025.
An Australian energy project, Uaroo, combined a 3.33GW solar park with a 2.04GW wind farm and a battery with a storage capacity of 9.1GWh. Swedish state-owned utility Vattenfall also aims to address the grid congestion in the Netherlands with its integrated wind, solar, and BESS project. Its €61 million Energypark Haringvliet, which opened in March, has 38MW of solar capacity, 22MW of wind and a 12MWh BESS.
The BESS (Battery electric storage system) sizing for these renewable projects is roughly 2 hours of storage for each megawatt of solar or wind generation.
The Tesla Megapack provides about 3.9 MWh of storage which would match up with 2 MW of solar or wind generation. 1 Terawatt of solar in China would match up to 2 TWh of battery storage for a fully stored energy solution. This would move the generation from say 9AM-11AM to 6pm -8pm. 2 TWH would be 500,000 Tesla Megapacks. It would take 10 Tesla factories at full 40 GWh per production to build this number of megapacks in 5 years. Global Solar would need to continue scaling for 20 years to meet current world electricity demand. World electricity demand would increase 30% by electrifying all cars and trucks. World electricity demand could start doubling every 5-10 years with growing energy demand from AI data centers. The existing solar and wind energy (2TWh) already need about 1,000,000 megapacks of storage. There is some pumped hydro and other storage but almost all of the solar and wind energy has no battery storage to make the energy created more efficient, productive and valuable.
China’s Energy Storage increased from 8.7 GW at the end of 2022 to 31.4 GW at the end of 2023. China has invested $14 billion into mainly lithium ion battery energy storage in 2022 and 2023.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
Brian, it would have been nice to cover what’s happening on the BESS side of the technology equation there as well. There seems to be a lot of investment in multiple storage technologies not typically found in the west, such as redox flow batteries and thermal batteries.
Strikes me the analysis is somewhat of a classroom exercise well removed from practical reality. Solar is not available most of the time. Nuclear is. The location of solar resources is critical, but not much of a factor for nuclear energy. Also, the greater the distance from load centers, the greater the energy losses. Mongolia is pretty far away from coastal China where a lot of the population lives.
Solar energy shows up more-or-less when it feels like it. Also requires huge amounts of territory if meant to be a mainstay of energy production. Seems like a poor choice for heavy reliance by a modern civilization.
… so why the emphasis for large scale deployment of solar energy? Follow the money.
So “nuclear power can get 8 kwh per installed kilowatt” should have been typed “nuclear power can get 8 Mwh per year per installed kilowatt”
I think I fixed it.
In 2022, China’s solar power generation reached 418 terawatt hours (TWh), a 20.9% increase from 2021.
China had 392 GW of installed solar at the end of 2022. In 2022, the US had 110 Gigawatts of installed solar
and it generated 204 TWh.
China is averaging a little 1 MWh/year (2.8kWh/day) per kilowatt of installed solar.
The US is getting 1.85 MWh/year (4.8 kWh/day) per kilowatt of installed solar.
It is a common mistake for people to assume that the kilowatts of installed energy are all equal.
US nuclear power can get 8 MWh/year (22 kWh/day) per installed kilowatt.
Other countries get 6-7 MWh/year per installed kilowatt of nuclear.
Coal power can get about 4 MWh/year per kilowatt of installed power.
Hydro can get about 3 MWh/year per kilowatt of installed power.
There is also a lot of variability by country, locations and projects even for the same energy type.
That makes sense. 24×365=8760 so about 8000 operating most of the time, so 1 kW = about 8 Mwh over a year operating most of the time.
Other sorts of power operate less consistently leading to a reduced number of Mwh over a year per kW. Solar would vary a lot by area.
Hi Brian.
Have a look at CO2 energy storage, as Energy Dome are building this year at full scale for completion by year end.
None of the temperature hassles of compressed air, industry standard components, closed loop system for round trip efficiency of 75-80% without substantial degradation over 30 years.
Overnight storage practical with the build going for a unit of 20MW/200MWh
Demonstrator up and running, and should come in at half the levelised cost of batteries, with no use of rare materials.
The calculation of needed battery storage is interesting but too simplified.
China is geographically big enough to have many options other than storing everything in batteries. We have a bit of the same situation here in Sweden, which is an elongated country where a lot of the power generation is up north and needs to be transported 1500 km south.
Factors affecting the storage need:
Not all solar and wind is generated in the same weather system, which means output is very different for different regions.
Hydro is normally used to balance out the intermittent solar/wind production and thus replaces most of the storage need. In Sweden, we have no storage and we just up-regulate hydro when the wind is not blowing. Wind power varies between 10 – 50% of national production.
Transmission lines between different regions and even other countries are an effective way to reduce storage need. Consumption and production evens out a lot over a big geographic region. In EU, the energy system is now integrated and power flows between countries to even out everything. This is what has allowed the massive buildout of wind and the closedown of nuclear.
The real time energy flow with hourly prices in northern Europe can be followed here: https://www.svk.se/om-kraftsystemet/kontrollrummet/
Stabilizing the grid frequency is a big use case for battery in a system with a lot of intermittent wind and solar. It doesn’t require gigantic capacity though.
So 110GW of US installed solar making 204TWh is 21% capacity factor. You report about 11% capacity factor in China. Well, so long as the return on investment happens within 10 years energetically AND financially, solar could be considered a good investment, from the perspective of the Chinese salesperson and their captured US representatives. I’m sure the batteries look similarly appealing to those in the battery business.
This KWh/KW metric obfuscates the basic capacity factor information. A low estimate of US nuke capacity factor is like 90%, and somehow that can gymnastically be converted to 8KWh/KW per your claim, but it makes my head hurt.
I’m still having a lot of pain understanding this KW-hr/hr figure of merit.. I can duplicate your numbers but what KWhr/hr is supposed to represent still baffles me.
1.10E+11 Watt Solar Installed
2.04E+14 Watt-hr Solar Generated
9.64E+14 Watt-hr Max Possible Generated
21% Whr/Whr Capacity Factor
1855 Watt-hr/hr
1.85 KW-hr/hr figure of merit discussed in article (US solar)
1.17E+09 Watt Single Nuke Installed
9.23E+12 Watt-hr Single Nuke Generated
1.03E+13 Watt-hr Max Possible Generated
90% Whr/Whr Capacity factor
7889 Watt-hr/hr
7.89 KW-hr/hr figure of merit discussed in article (US nuke)
It is pretty interesting that the 110 GW of installed solar in USA pumps the same energy as 22 1.2GWe PWRs, except the solar does it with 21% availability (unreliability). I’m sure the solar works well in San Diego, but we haven’t had a sunny day here in the armpit of the mid-Atlantic in about 10 days.
I get it though. There is a heap of commercial-grade silicon sitting around and might as well put it to use… I get that. We can’t put them everywhere though… and at a certain penetration many argue they will create stability problems. Who cares, I just don’t understand what we get out of the KWhr/hr metric.
[ “understanding this KW-hr/hr figure”
this relates to the amount of resources being invested for a projected power&energy supply (&backup) and also related to EROEI for primary energy consumption for construction&maintenance&materials upcycling (while financially it could be comparable or mostly cheaper for renewables wind&solar, it’s requiring relatively bigger invest on (energy&)materials&work force&area for plant’s generator parts)
Directly utilized electricity and distributed between ‘dense’ population areas by power grid is (mostly) preferable compared to storage options.
It would be interesting, what’s the CO2 balance of transferred electricity (import/export over state’s/countries borders), if ‘significant’ amounts? Statistics?
A global supply grid would profit from room temperature superconductors (reduced distribution losses by 10-20%). ]
The downside of relying on long range distribution of electricity, though, is that, even if it works, it enables large scale outages if something goes wrong. Depending on geopolitical considerations, it’s really easy to find that somebody has actually arranged for something to go wrong… The US electrical grid is terrifyingly vulnerable, we hardly need to make it more so.
Also, while weather isn’t strongly correlated over really long distances, it’s hardly anti-correlated; You’re going to get bad weather at the same time at both ends of a country on a fairly frequent basis, even if it’s not as frequent as getting it at one end or the other.
I’ve seen at least one study using actual weather records that indicated that, even with a coast to coast integrated grid, here in the US we’d be having occasional brownouts. Here’s such an analysis: https://www.nature.com/articles/s41467-021-26355-z
Notice that their definition of “success” still involved brownouts, and required both overbuilding and significant storage, in addition to a nation-wide lossless grid. Nation-wide brownouts, mind you.
When a source of power is only available on a probalistic basis, and correlated to boot, no reasonable amount of overbuilding, grid size, and storage can change the fact that sometimes you’ll get a bad roll of the dice and the lights will go out.
[ thanks, absolutely (me with emphasizing more on technological implications),
it’s a percentage between modeling a global super grid and a global super storage conception, but i don’t know this useful percentage number (A(G)I super computers or robot networks?), not for now or in a decade nor for local optimization nor with global share for renewables (on projected cost) combined with backup plants/storage/efficiency options, not for technological&science discovery singularities (financing numbers can be the playing area for the dice)
Yes, who would vote (on a global democracy) for connecting China, Asia, India, Africa, Australia, Europe and US, America into one power grid? Including Russian Federation, North Korea, Taiwan, Syria, Palestine – means shares of multi GW connections(?) and maybe a 22. century necessary for humans to develop(?)
Some is about trust, also. ]
Mega-scale grids are, unfortunately, a solution fit for a world we don’t find ourselves in. They’re a high trust solution in a low trust world. I wish we were in a world where they were a safe solution to anything, I really do, but we’re not.
We should really, at this point, be trying to localize the grids, not globalize them. Interconnects to help out when there are problems is a great idea, but routinely relying on them is a very bad idea indeed. The capacity to move power long distances is great, the necessity to move it long distances is equally bad.
The advantage of power sources like nuclear is that they aren’t just reliable and dispatchable, (Rather than producing power stochastically without regard to when it’s needed.) but to the extent they do have outages, they are completely uncorrelated between plants.
The ideal grid isn’t a continent wide web of low capacity factor and randomly available sources. It’s a quilt of localized high capacity factor sources, with sharing between local areas available as an emergency option, not a normal mode of operation.
You have too much common sense.
Unfortunately, politics hates common sense.
I am all for renewables and batteries and steam fracking and CO2 compression. But natural gas and nuclear are affordable, and reliable, unless people are convinced by special interest groups they are bad. Most of the nuclear costs are due to regulations and bureaucrats who have an agenda against nuclear waste.
8760 hours in a year. 91% capacity factor is 7971 kWh per year for 1 kilowatt
On average, Brian, there’s 8,766 hours per Julian Year. Almost the same per Gregorian Year.
So “nuclear power can get 8 kwh per installed kilowatt” should have been typed “nuclear power can get 8 Mwh per year per installed kilowatt”