I think it helps to put things into context when comparing the growth of nuclear vs. that of solar and wind today. Necessity drove France's nuclear growth. France was motivated because it did not really have any other options during the Arab oil embargo, "No coal, no gas, no oil, no options." Nuclear was their only option if they were going to be energy independent.
The energy transition to low-carbon sources is much more complicated because it is global and not everyone is as motivated as France or Sweden.
I am from the “all-of-the-above” school of thinking when it comes to the energy transition in the U.S. Most net-zero energy-system studies include some fraction of nuclear in the U.S. in 2050. A good place to start when considering the future of nuclear in the U.S. is the DOE study completed last year, https://liftoff.energy.gov/advanced-nuclear/ .
If you are looking for a global perspective on the future of advanced nuclear, I recommend this article and study by the London-based New Nuclear Watch Institute,
The 444 GW solar deployment last year was GLOBAL. I think China deployed about 250 GW last year (which was more than the entire world in 2022 at 230 GW I believe).
Global installed production capacity for solar last year was over 800 GW but not fully ramped yet. So hopefully we see a 550-600 GW deployment this year!!!
Ultimately I think we will end up deploying over 2 TW of solar per year by 2030. With BESS coming online at the hundreds of GWh level within the next year or two it will turbo charge solar (and wind or any variable generation) deployments even more IMO.
It seems you are correct, and that would put China 2023 down to third in the list.
Of course, as pointed out in comments earlier, it's not really fair to compare China to Sweden, or even France. We can probably say that in 2023 China deployed more KWh per person of clean energy than Europe, or the USA, or India, have ever done in a year.
I was looking at what happens if the world can deploy 1 TWh of solar per year. In theory - depending on exactly how electricity displaces fossil fuels - it could "wipe out" global coal usage in a decade, global oil usage in another decade, and global gas usage in another decade.
It's not the rate of solar and wind which needs to be compared to nuclear, it's the rate of growth in storage technology particularily long duration energy strorage, to firm up solar and wind, so it is comparable in its ability to generate when needed. Without adequate storage we will still be reliant on generation driven by the weather and either too much and curtailment or not enough.
You don't need LT storage. The new system will be built with the same capacity requirements as the old system. You need to have sufficient capacity to produce enough on the lowest production days to meet the highest demand days. No more than 5-7 days of storage is required. If the lowest output day is only 20% capacity, you need to >5x capacity.
The US grid has capacity to produce over 11000 TWh per year but we only demand about 4000 TWh. Production ability has to be multiples of demand. Is so now, will be with the new system. You don't need seasonal storage.
5-7 days of storage isn't exactly peaking at circa 2-4hrs, what you have described is essentially long duration and we are no where near delivering 5-7 days of storage. As an example all the Pumped Storage in the UK would last 1hr at average grid demand. The UK also suffered 11 days of below 20% wind production so it also really depends on what the energy mix is providing the 20% and how firm it is.
Hanna, I love your work. But you tripped a pet peeve when you used the word "exponential" to describe solar progress when the more accurate term for describing technological progress is "sigmoidal." Sure, there are regions in a sigmoid where growth rates are the same as they would be for an exponential curve, but sigmoids illustrate that there are natural limiting factors that prevent infinite growth with continuing acceleration.
Technological growth is more accurately portrayed as a series of sigmoidal curves, but the start and stop are not predictable.
While it's certainly difficult to seen a resurgence in nuclear using the evidence accumulated during the past 20 years, that same statement could have been made in 1960 about nuclear and in 2005 about wind and solar.
Data scientists need accumulated data in order to make predictions. Innovators often recognize that they can create something new even there are no data providing that it can be done.
It's also a maritime axiom - Don't try to steer the ship by watching the wake. (Or maybe that is something I made up as a seamanship instructor.)
Well said. I have argued for a couple of decades against the SIngularitarian view that we should instead think of "surges" of progress that are not continuously exponential. Sometimes in some areas progress can slow -- as seen in US drug development despite the decoding of the human genome. Exponential views tend to ignore economic and regulatory factors.
The problem with variable renewables is that due to their intermittency, you have diminishing returns/escalating cost for the last 20-30% of coverage unless you have a lot of hydropower.
You don't have the problem as much with a nuclear buildout, since you can simply replace the fossile plants on a one-to-one basis and be done with it.
Especially poorer countries might stop at 70-80% of the way to avoid the "last meter" cost.
This being said, the clean transition needs maximum utilisation of all supply chains.
I am very optimistic about nuclear. For the first time since many years, it now has the full political support of most economic powerhouses with the exception of, drumroll, Germany.
Nuclear's problem first and foremost were political, not economical.
I agree/understand the reasoning re capacity factors and diminishing returns.
Combining wind and solar and having some hydro/pumped storage to start with and using modest amounts of batteries, I think you can delay the onset of significant diminishing returns to the 60-70% buildout in the German scenario. After that, the required effort escalates.
From my experience, the diminishing returns are not properly understood even by many interested people, and someone should make a good visualisation of this connnection.
In a few decades solar and wind plus pumped hydro batteries will make creating fusion power generation a waste of time. Fusion only heats water. Solar PV creates electricity instantly in a millimetre of material.
Solar PV does, indeed create electricity instantly. But you did not quantify your statement or acknowledge that many human energy needs are in the form of heat, not in the form of electricity. Though electricity can obviously produce heat, delivering the electricity from the point where it is created to the point where the heat is needed isn't trivial, especially when the amount of heat needed in certain factory processes is measured in the 10s of MWth in a piece of equipment the size of a shipping container.
Why pit PV against fission? Both have their uses in the enormous spectrum of energy needed for modern society.
If SMR fission can get its act together that's good. Terrapower shows promise, it's a pity Nuscale has problems. It's fusion which I think is a waste of time.
I tend to agree with you about fusion, though I am not confident enough in my understanding of the world to totally dismiss it.
But I am cautiously excited by the potential for new fission power in a wide variety of sizes, shapes and technologies. I spend my days reviewing technical documents, listening to presentations and talking with talented engineers/businesspeople. The coverage in the mainstream press is roughly equivalent to stories about the visible portion of icebergs that totally ignores the portion that is underwater and not visible unless actively searched.
TerraPower is a player that has made good progress, received major government support and has a media magnet founder in Bill Gates.
But there are many dozens of projects, some of which are moving at least as fast and with equivalently committed support.
For example - there is a company called newcleo founded within the past three years that raised >$300 M. A company called Naarea has 150 employees and is just one of more than a dozen French entities. X-Energy is working with Dow on a project in Texas and has received essentially the same level of federal support as TerraPower. But Kan Gaffarian is not as well known as Bill Gates.
Fission releases about 2 million times as much energy per unit mass as burning oil. There is a wealth of innovative ways to use that advantage. Watch the bow wave, not the wake.
I think Nuscale's problems dent some of the hopes around SMRs. Yes, factory built reactors can reduce cost ... but smaller reactors can increase increase costs.
At the end of the day, dealing with water at 160 bar is difficult and expensive. Hence all the hope in low pressure designs as from Terrpower, Moltex, Thorcon etc.
I tried to get more interest in the UK for Molten Salts a few years back, but none of the companies have the lobbying power of Rolls Royce, with their fairly conventional PWR design, and all the approvals processes are geared around PWRs.
France has woken up and realised that they need to replace a lot of nuclear power in 2 decades or so, and that the EPR is too expensive. They also have to cope with masses of cheap (or negatively priced) solar power from across Europe, most afternoons. That will favour high temperature reactors than can store heat for several hours.
We will still need fusion for of planet purposes. The issue w fusion might me the same as fission. It won't scale. Fission off course has a host of other issues. That being said I am pro-fission but just don't see it as a viable option. S+W+B will dominate by 2030 and it won't even be close.
Pumped hydro batteries require large and topographical correct land area and are not practical in most places. The whole growth of solar story ignores nighttime and intermittent and the large role of gas peaker plants and other necessary backup which nuclear doesn’t need.
In your last graph, you compare 4 year solar intervals with 10 year nuclear intervals. I understand why you did that, but would like to add that dividing a group that you average over into smaller sub-groups will often create sub-groups with an increased average. If you take 4 year intervals from Sweden and France with the most reactor starts, you will most likely increase the rate for nuclear by a lot too.
From a European perspective, there is masses of space in North Africa with great sunlight. But submarine HVDC is expensive, which suggests maybe we should be getting hydrogen from North Africa.
But what about the Adriatic? Sunny climate, very calm waters (windy, just like a lot of lakes), nearby steel making, and maybe 200 million people and much of Europe's industry within 400km of its shores. You could put a TW of solar in there and only notice it from space.
That ignores the fact that solar panels are not immune to corrosion from sea water, and the dead coral and fish from the blocked sunlight. Has sea based solar ever been tried anywhere, other than to power small buoys?
This is quite old now, but I thought I'd look at Chinese figures again, prompted by some news FUD. My notes:
Chinese energy production is hard to find (maybe not able to search in Chinese), but the amount of solar and wind energy produce is heavily understated.
The only figures I can find suggest solar produced about 880TWh in China last year, but this also seems to miss quite a bit - maybe unmetered solar. 1,000 TWh seems a more reasonable minimum given capacity and expected yields.
Estimates for wind seem to be based on capacity factor, 521GW x 23.4% X 8760 hours = c. 1000 TWh.
China is expected to treble solar capacity by end of 2029, which would raise solar output to 3000 TWh, and doubling wind to 2000 TWh, so these two would be providing more than electricity than coal.
Put another way, China's renewables over this decade will add about 300KWh per capita per year.
So only beaten by French nuclear (1979 - 89) and Swedish nuclear (1976 - 86).
A similar look suggests the Netherlands has been adding 400KWh per person of solar alone, but that was for a few years and thy may run out of space.
Wind and solar is not a reliable solution for the main energy grid although it does have some off grid applications, and certainly wont support industrial activity.
The only real alternative is Nuclear, and we must go back to that for all on grid requirements.
A quick study finds that in full operation nuclear will be 50% less expensive as compared to W&S…(even with costly subsidies) … just compare energy costs in France (70% nuclear) versus Germany (no nuclear) . Plus the reliability is just a joke with W&S
Also, everyone must realize that fossil fuels are needed for many other requirements outside of just electrical energy and without it we will have an economy more like the 1900s.
This is interesting! Not the main point, but I was surprised at how low China falls on the measure of new, carbon-free electricity generation per capita. I would have thought they'd be close to the lead given their big boom in capacity additions, and that they've presumably seen more load growth than the more mature western economies.
I imagine it's some mix of lower capacity factors in China, the remaining economic gap, or maybe China had particularly high growth in carbon-free generation in the uncovered years 2015-2018.
Great blog 👍 seems to suggest that nuclear only grew at a rapid rate (equivalent to current global renewables growth) in a select few countries for a limited period. Nuclear is a great low carbon fuel but it's being dramatically outcompeted by renewables and batteries. It's diseconomies of scale and complexity (not to mention risks) relative to clean-tech make it a much less attractive option now than it ever has been.
I think it helps to put things into context when comparing the growth of nuclear vs. that of solar and wind today. Necessity drove France's nuclear growth. France was motivated because it did not really have any other options during the Arab oil embargo, "No coal, no gas, no oil, no options." Nuclear was their only option if they were going to be energy independent.
The energy transition to low-carbon sources is much more complicated because it is global and not everyone is as motivated as France or Sweden.
I am from the “all-of-the-above” school of thinking when it comes to the energy transition in the U.S. Most net-zero energy-system studies include some fraction of nuclear in the U.S. in 2050. A good place to start when considering the future of nuclear in the U.S. is the DOE study completed last year, https://liftoff.energy.gov/advanced-nuclear/ .
If you are looking for a global perspective on the future of advanced nuclear, I recommend this article and study by the London-based New Nuclear Watch Institute,
https://www.powermag.com/2023-a-transformative-year-for-small-modular-nuclear-reactors/ and
https://www.newnuclearwatchinstitute.org/report/scaling-success-navigating-the-future-of-small-modular-reactors-in-competitive-global-low-carbon-energy-markets .
Other related nuclear news that gives me some optimism for nuclear energy in the U.S. and the West:
GE Hitachi and Ontario Power Generation have signed a contract to deploy a BWRX-300, https://www.utilitydive.com/news/SMRs-reactor-GE-Hitachi-Ontario-Public-Power-Aecon-Group-nuclear/641483/ at their Darlington project site. This is most likely to be the first SMR to be deployed in North America. They are optimistic that it will be operational in late 2028. The Tennessee Valley Authority has joined an international consortium to develop a standard design of the BWRX-300 for deployment in Canada, the U.S., and Poland, https://www.utilitydive.com/news/tva-ge-hitachi-small-modular-reactor-smr-nuclear/645861/ .
U.K. has shortlisted six designs in their SMR competition to encourage the development of SMRs,
https://www.powermag.com/uk-shortlists-six-nuclear-designs-in-smr-competition-intends-to-award-contract-by-summer-2024/ .
U.S., Canada, and U.K. are cooperating on licensing,
https://www.powermag.com/u-s-uk-canada-ink-trilateral-memo-to-cooperate-on-advanced-reactor-licensing/ .
In 2023, the U.S. House and the Senate both passed legislation overwhelmingly supporting development of new generation nuclear:
https://thehill.com/policy/energy-environment/4495980-house-approves-bipartisan-bill-aimed-at-bolstering-nuclear-energy/ and
https://thehill.com/policy/energy-environment/4495980-house-approves-bipartisan-bill-aimed-at-bolstering-nuclear-energy/ .
The growth of solar over the last 3 years has been so fast that a 5 year view doesn't capture it.
China added 444GW last year.
That will probably make 800TWh in 2024 (slightly optimistic to make maths easier).
China has 1600 million people.
So 500KWh per person.
Pretty much up there with France and Sweden, for the world's biggest emitter.
The 444 GW solar deployment last year was GLOBAL. I think China deployed about 250 GW last year (which was more than the entire world in 2022 at 230 GW I believe).
Global installed production capacity for solar last year was over 800 GW but not fully ramped yet. So hopefully we see a 550-600 GW deployment this year!!!
Ultimately I think we will end up deploying over 2 TW of solar per year by 2030. With BESS coming online at the hundreds of GWh level within the next year or two it will turbo charge solar (and wind or any variable generation) deployments even more IMO.
It seems you are correct, and that would put China 2023 down to third in the list.
Of course, as pointed out in comments earlier, it's not really fair to compare China to Sweden, or even France. We can probably say that in 2023 China deployed more KWh per person of clean energy than Europe, or the USA, or India, have ever done in a year.
I was looking at what happens if the world can deploy 1 TWh of solar per year. In theory - depending on exactly how electricity displaces fossil fuels - it could "wipe out" global coal usage in a decade, global oil usage in another decade, and global gas usage in another decade.
It's not the rate of solar and wind which needs to be compared to nuclear, it's the rate of growth in storage technology particularily long duration energy strorage, to firm up solar and wind, so it is comparable in its ability to generate when needed. Without adequate storage we will still be reliant on generation driven by the weather and either too much and curtailment or not enough.
You don't need LT storage. The new system will be built with the same capacity requirements as the old system. You need to have sufficient capacity to produce enough on the lowest production days to meet the highest demand days. No more than 5-7 days of storage is required. If the lowest output day is only 20% capacity, you need to >5x capacity.
The US grid has capacity to produce over 11000 TWh per year but we only demand about 4000 TWh. Production ability has to be multiples of demand. Is so now, will be with the new system. You don't need seasonal storage.
5-7 days of storage isn't exactly peaking at circa 2-4hrs, what you have described is essentially long duration and we are no where near delivering 5-7 days of storage. As an example all the Pumped Storage in the UK would last 1hr at average grid demand. The UK also suffered 11 days of below 20% wind production so it also really depends on what the energy mix is providing the 20% and how firm it is.
Hanna, I love your work. But you tripped a pet peeve when you used the word "exponential" to describe solar progress when the more accurate term for describing technological progress is "sigmoidal." Sure, there are regions in a sigmoid where growth rates are the same as they would be for an exponential curve, but sigmoids illustrate that there are natural limiting factors that prevent infinite growth with continuing acceleration.
Technological growth is more accurately portrayed as a series of sigmoidal curves, but the start and stop are not predictable.
While it's certainly difficult to seen a resurgence in nuclear using the evidence accumulated during the past 20 years, that same statement could have been made in 1960 about nuclear and in 2005 about wind and solar.
Data scientists need accumulated data in order to make predictions. Innovators often recognize that they can create something new even there are no data providing that it can be done.
It's also a maritime axiom - Don't try to steer the ship by watching the wake. (Or maybe that is something I made up as a seamanship instructor.)
Well said. I have argued for a couple of decades against the SIngularitarian view that we should instead think of "surges" of progress that are not continuously exponential. Sometimes in some areas progress can slow -- as seen in US drug development despite the decoding of the human genome. Exponential views tend to ignore economic and regulatory factors.
Sigmoidal... A new word for me to start using instead of exponential 🤓👍
The problem with variable renewables is that due to their intermittency, you have diminishing returns/escalating cost for the last 20-30% of coverage unless you have a lot of hydropower.
You don't have the problem as much with a nuclear buildout, since you can simply replace the fossile plants on a one-to-one basis and be done with it.
Especially poorer countries might stop at 70-80% of the way to avoid the "last meter" cost.
This being said, the clean transition needs maximum utilisation of all supply chains.
I am very optimistic about nuclear. For the first time since many years, it now has the full political support of most economic powerhouses with the exception of, drumroll, Germany.
Nuclear's problem first and foremost were political, not economical.
Walter - there is good science and engineering that tells us that the area of diminishing returns starts much earlier than 70-80% penetration.
A better estimate is that diminishing returns become increasingly apparent as penetration levels approach the natural limits of capacity factor.
This helps stimulate thought about the decreasing value of electricity from generating sources that have correlated production schedules.
http://open-electricity-economics.org/book/text/07.html
Thank you for the link, interesting!
I agree/understand the reasoning re capacity factors and diminishing returns.
Combining wind and solar and having some hydro/pumped storage to start with and using modest amounts of batteries, I think you can delay the onset of significant diminishing returns to the 60-70% buildout in the German scenario. After that, the required effort escalates.
From my experience, the diminishing returns are not properly understood even by many interested people, and someone should make a good visualisation of this connnection.
Best regards
I worry about whether China will have too
Much control on the production of solar panels. And we will need nuclear fission power until we have reliably cracked fusion power.
In a few decades solar and wind plus pumped hydro batteries will make creating fusion power generation a waste of time. Fusion only heats water. Solar PV creates electricity instantly in a millimetre of material.
Solar PV does, indeed create electricity instantly. But you did not quantify your statement or acknowledge that many human energy needs are in the form of heat, not in the form of electricity. Though electricity can obviously produce heat, delivering the electricity from the point where it is created to the point where the heat is needed isn't trivial, especially when the amount of heat needed in certain factory processes is measured in the 10s of MWth in a piece of equipment the size of a shipping container.
Why pit PV against fission? Both have their uses in the enormous spectrum of energy needed for modern society.
If SMR fission can get its act together that's good. Terrapower shows promise, it's a pity Nuscale has problems. It's fusion which I think is a waste of time.
I tend to agree with you about fusion, though I am not confident enough in my understanding of the world to totally dismiss it.
But I am cautiously excited by the potential for new fission power in a wide variety of sizes, shapes and technologies. I spend my days reviewing technical documents, listening to presentations and talking with talented engineers/businesspeople. The coverage in the mainstream press is roughly equivalent to stories about the visible portion of icebergs that totally ignores the portion that is underwater and not visible unless actively searched.
TerraPower is a player that has made good progress, received major government support and has a media magnet founder in Bill Gates.
But there are many dozens of projects, some of which are moving at least as fast and with equivalently committed support.
For example - there is a company called newcleo founded within the past three years that raised >$300 M. A company called Naarea has 150 employees and is just one of more than a dozen French entities. X-Energy is working with Dow on a project in Texas and has received essentially the same level of federal support as TerraPower. But Kan Gaffarian is not as well known as Bill Gates.
Fission releases about 2 million times as much energy per unit mass as burning oil. There is a wealth of innovative ways to use that advantage. Watch the bow wave, not the wake.
Hi Rod,
I think Nuscale's problems dent some of the hopes around SMRs. Yes, factory built reactors can reduce cost ... but smaller reactors can increase increase costs.
At the end of the day, dealing with water at 160 bar is difficult and expensive. Hence all the hope in low pressure designs as from Terrpower, Moltex, Thorcon etc.
I tried to get more interest in the UK for Molten Salts a few years back, but none of the companies have the lobbying power of Rolls Royce, with their fairly conventional PWR design, and all the approvals processes are geared around PWRs.
France has woken up and realised that they need to replace a lot of nuclear power in 2 decades or so, and that the EPR is too expensive. They also have to cope with masses of cheap (or negatively priced) solar power from across Europe, most afternoons. That will favour high temperature reactors than can store heat for several hours.
We will still need fusion for of planet purposes. The issue w fusion might me the same as fission. It won't scale. Fission off course has a host of other issues. That being said I am pro-fission but just don't see it as a viable option. S+W+B will dominate by 2030 and it won't even be close.
Pumped hydro batteries require large and topographical correct land area and are not practical in most places. The whole growth of solar story ignores nighttime and intermittent and the large role of gas peaker plants and other necessary backup which nuclear doesn’t need.
It works in a lot of places
Global Atlas of Closed-Loop Pumped Hydro Energy Storage
https://www.cell.com/joule/fulltext/S2542-4351(20)30559-6
In your last graph, you compare 4 year solar intervals with 10 year nuclear intervals. I understand why you did that, but would like to add that dividing a group that you average over into smaller sub-groups will often create sub-groups with an increased average. If you take 4 year intervals from Sweden and France with the most reactor starts, you will most likely increase the rate for nuclear by a lot too.
Solar energy capacity grows roughly 20% per year (doubling every 3.5 years). Grid scale storage increases 100% per year.
I heard you today on the Ezra Klein show. You did a great job!
Another great piece!
My main concern is land requirements
http://www.mattball.org/2022/09/most-recommended-overview-re-climate.html
My view is that the land requirement issue for renewables is a red herring:
https://open.substack.com/pub/craigbonthron/p/clean-tech-fallacies-part-ii-energy?utm_source=share&utm_medium=android&r=t9w83
That's why I'm keen on offshore solar power.
From a European perspective, there is masses of space in North Africa with great sunlight. But submarine HVDC is expensive, which suggests maybe we should be getting hydrogen from North Africa.
But what about the Adriatic? Sunny climate, very calm waters (windy, just like a lot of lakes), nearby steel making, and maybe 200 million people and much of Europe's industry within 400km of its shores. You could put a TW of solar in there and only notice it from space.
That ignores the fact that solar panels are not immune to corrosion from sea water, and the dead coral and fish from the blocked sunlight. Has sea based solar ever been tried anywhere, other than to power small buoys?
It is being tried by a number of (mostly Dutch and Norwegian) companies in the North Sea.
I think a MW scale array is planned to go with a windfarm.
Probably because if it can work there, it can work anywhere.
There is also a company with several tropical lagoon installations powering hotels etc.
Best not place it on the coral reef, and the fish don't mind.
This is quite old now, but I thought I'd look at Chinese figures again, prompted by some news FUD. My notes:
Chinese energy production is hard to find (maybe not able to search in Chinese), but the amount of solar and wind energy produce is heavily understated.
The only figures I can find suggest solar produced about 880TWh in China last year, but this also seems to miss quite a bit - maybe unmetered solar. 1,000 TWh seems a more reasonable minimum given capacity and expected yields.
Estimates for wind seem to be based on capacity factor, 521GW x 23.4% X 8760 hours = c. 1000 TWh.
China is expected to treble solar capacity by end of 2029, which would raise solar output to 3000 TWh, and doubling wind to 2000 TWh, so these two would be providing more than electricity than coal.
Put another way, China's renewables over this decade will add about 300KWh per capita per year.
So only beaten by French nuclear (1979 - 89) and Swedish nuclear (1976 - 86).
A similar look suggests the Netherlands has been adding 400KWh per person of solar alone, but that was for a few years and thy may run out of space.
Low Carbon is the way forward.
Great post, thank you! 💚 🥃
Wind and solar is not a reliable solution for the main energy grid although it does have some off grid applications, and certainly wont support industrial activity.
The only real alternative is Nuclear, and we must go back to that for all on grid requirements.
A quick study finds that in full operation nuclear will be 50% less expensive as compared to W&S…(even with costly subsidies) … just compare energy costs in France (70% nuclear) versus Germany (no nuclear) . Plus the reliability is just a joke with W&S
Also, everyone must realize that fossil fuels are needed for many other requirements outside of just electrical energy and without it we will have an economy more like the 1900s.
This is interesting! Not the main point, but I was surprised at how low China falls on the measure of new, carbon-free electricity generation per capita. I would have thought they'd be close to the lead given their big boom in capacity additions, and that they've presumably seen more load growth than the more mature western economies.
I imagine it's some mix of lower capacity factors in China, the remaining economic gap, or maybe China had particularly high growth in carbon-free generation in the uncovered years 2015-2018.
Great blog 👍 seems to suggest that nuclear only grew at a rapid rate (equivalent to current global renewables growth) in a select few countries for a limited period. Nuclear is a great low carbon fuel but it's being dramatically outcompeted by renewables and batteries. It's diseconomies of scale and complexity (not to mention risks) relative to clean-tech make it a much less attractive option now than it ever has been.
https://open.substack.com/pub/craigbonthron/p/a-nuclear-waste-of-time-and-money?utm_source=share&utm_medium=android&r=t9w83