I live in Houston TX. Noah, you have mischaracterized the Texas power situation. Wind is just as important for electricity power in Texas as solar, with natural gas. coal, and batteries acting as a backup.
The agency that runs the Texas electric grid, ERCOT, rates the monthly power capacity of various energy sources as follows:
Solar 25,961 MW
Wind 39,525 MW
Battery 8,779 MW
Natural Gas 67,066 MW
Coal 14,713 MW
Nuclear 5,268 MW
ERCOT can select which of these energy sources it will use at any given moment. My understanding is that they use solar, wind and nuclear first and then use other sources to fill in.
Interestingly, solar and wind complement each other almost perfectly. Sun shines during the day and the wind on the North Texas Plains blows most robustly at night. Generally speaking, wind produces as much power at night as solar does during the day, but last night appears to have been an exception.
Crunch time comes at dawn and dusk when both wind and solar are weak.
It's not just the cost of solar. It's the cost of transmission of that power hundreds of miles to a destination and it's also the cost of long-term storage of solar power for places that have no sun whatsoever during winter. It's also the fragility of all those systems. If you can get generate power close to home, perhaps by wind or nuclear, you avoid those costs and risks.
Places that get no sun whatsoever during the winter probably shouldn't depend on solar. But for areas that do get sunlight, hardly any "close to home" generation is as cheap, easy or effective as photovoltaic solar panels.
Texas electricity is equally dependent at this point on wind and solar. The wind blows most reliably at night and solar takes care of the day. The two energy sources are absolutely complementary for us.
My guess is that it depends on climate. I think that Scotland and northern Europe will always do better with offshore wind. Noah does not seem to appreciate how dark it is from northern France northward during the wintertime. You just can't add more solar. I spent one winter in Gottingen, and I was told that you cannot expect to see the sun between December and April. Overcast. I did not see the sun between December and late February when I left.
A lot of North Americans forget how far north Europe is: Madrid is comparable in latitude to New York City, Rome to Chicago, Nice to Toronto, London to Calgary, Liverpool to Goose Bay and Scotland to the Alaskan panhandle.
Hmmm. Low density. Efficient within only certain latitude ranges and certain climates. Zero production for half the day. Sounds crazy to me no matter how cheap panels get. It will be useful but only dominant in certain regions.
The ideal Solar location is low to mid latitude, located close to but a couple of time zones west of major population centers in a thinly populated region the residents don’t mind paving over with panels.
Western North Africa (Morocco, Western Sahara, Algeria, Mauritania) could be hooked into Europe’s grid and provide solar power after sunset to Germany and E Europe (actively and then via battery). The same could be said of Spain, but they’re likely less willing to destroy their countryside.
Seems very likely, but it's important to remember that construction is decided by market demand. The more wind and solar that goes online the more they have to compete with each other for prices, after all the sun is always going to shine at the same time for everyone in a geographic region and the wind is going to blow at the same time. In many places the electricity prices can already go to zero when the wind is blowing hard enough, which makes it less cost effective to build a new wind turbine. More battery storage should even these things out some, it's just important to remember that things don't remain static over time.
I’d like to see an article about long term battery storage. What is current what might be actual in a few years and long term.
What nuclear would have advantage over solar is a bad volcano. An asteroid strike or god forbid a nuclear winter. I don’t know what a war between Iran and Israel might look like. How many nuclear bombs might go off and what it might do to the atmosphere and therefore solar collection.
In thinking about the nuclear saver rattling going on none of the above are ridiculous.
While natural disasters might leave the electric grid basically unimpaired, it is widely expected that warfare between major powers would involve deliberate disruption of electric grids. That might make any means of electric generation ineffective. Not to speak of rendering cryptocurrencies useless.
I'd never really thought about that, Kathleen, but that really makes sense. Factories that can supply their own power locally place a country in far better strategic position than those that require centralized power distribution, which is easy to disrupt in a military conflict.
We really should be transitioning to a fully modular grid, with lots of hyper-local solar and wind integrated below the level of the neighborhood substation, where you're stepping down to 480V, and some batteries -- the size of a Megapack or two -- at each of those substations. The grid will then serve to transfer energy between the places that are overproducing and the places that are underproducing, equalizing how full each node's battery bank is.
The problem remains with IBM powering a data center off the Three Mile Island reactor is that a nationwide grid is still needed for the distribution of that data.
There is no _great_ solution for long-term storage yet.
My suspicion is that what's going to happen ultimately is that the learning curve is going to proceed so far that just absurdly over-building PV + storage to be adequate in winter will mean we're oversupplied with energy in summer, and that means you can start doing fun stuff like hydrolyzing water and storing hydrogen (long term hydrogen storage is also pretty expensive right now, but I'm optimistic people are going to figure out better solutions with stuff like metal-hydride honeycombs), or building hydrocarbons out of water and the CO2 in the air (which can then be processed through a fuel cell later).
There also are some interesting technologies that may pan out for long-term battery storage, but my sense is that unless PV and lithium batteries _slow down_, in terms of how rapidly they're getting cheaper, the financial logic is going to point to just building ridiculous amounts of that, rather than investing in fancier tech. But we'll see. As long as the politics and geopolitics don't screw it up, there's a bright future to be had, regardless of exactly which tech wins out.
I'm with you on nuclear power, but in the event of a nuclear weapons exchange large enough to cause global cooling for an extended period, even if your country were not a party to the exchange, whether you have solar or nuke plants would likely be the least of your worries.
I think market forces will indeed push most of the world to adopt solar + batteries. But rich countries that foresee their energy needs increasing exponentially in the next 20-30 years due to data center needs, EV adoption, and (hopefully) carbon capture technology, will probably adopt nuclear power as their main source.
Yes it's more expensive. But it is also far more efficient for base load needs.
I'm curious how and when this will happen. China is the only country building nuclear at significant scale, and their plan estimates 18% nuclear by 2060, a target they'll probably miss (for the happy reason of over-building of solar and wind.) The rich nations of the West aren't building at anywhere near that scale, and to achieve what you're saying on the scale of 20-30 years would require the buildouts start, basically, yesterday.
The elimination of most coal plants over the next 10 years (per EPA rules, if not overturned) will result in shortfalls in baseload power and there is no feasible way batteries will replace this in that timeframe.
I hope we build lots and lots more nuclear plants to replace those coal plants, but battery storage for the grid is exploding right now. California alone built like 8 GW in 4 years. I used to be pretty bearish on grid battery storage, but it seems to be working.
If you’ve ever spent time in the British isles, you know that’s not as pessimistic as it sounds. Americans might have a hard time appreciating how gray England is. Think of the whole country as Seattle.
I wonder if this is partly why the graph shows the British Isles is one of the few places where solar will not be the cheapest former energy in 2027.
When I constructed square miles of solar arrays in Saharan Africa in my mind's eye, I immediately saw AK-47 armed hoodlums holding the electricity supply of Europe for ransom. Political stability is a factor.
Yes exactly. Here in the middle of England we have months long lulls in sunlight every year. If we had enough battery capacity to cover a few weeks of darkness we would still have to overbuild the solar capacity by about a factor 7 to get through the winter. Of course wind helps because it has the opposite seasonality to the sun. But wind has less of a learning curve than solar because a lot of it is built on site, not in a factory. So I think we will still need nuclear.
What's the POINT of the "but ..." The policy implications are the same whether one is a "retrofuturist" or nor. We need to remove the regulations that do not pass a cost benefit test from nuclear power generation and distribution, just like we should from geothermal or solar or wind. And we should pay producers the same amount more for producing a KwH of zero CO2 emitting electricity by taxing the net emissions of CO2 whatever technology is used. Neither Noah or I know whether this will mean nuclear power will remain a niche energy source or not. If we coud predict which technologies will be invented, they would already have been invented. Maybe near zero marginal cost nuclear power will produce the energy needed to remove CO2 from the atmosphere so inexpensively there will be no need to "decarbonize" anything.
But however optimistic or pessimistic you are about technology X or Y, the policy is the same: give the correct disincentive for emitting net CO2 and methane and evaluate each investment decision according to cost benefit analysis.
All that said, given the systematic suppression of nuclear power development for decades, it's not absurd to think that the marginal returns to policy reform are greater for nuclear power than for solar or wind?
Sure, we should expect better policy to make it so our nuclear power plant costs are more in line with the costs to build nuclear power plants in other countries
Nuclear energy may not produce CO2, but it can produce nuclear waste that takes tens of thousands of years to lose its radioactivity. No one has been able to convince me that the anti-nuclear movement wasn't correct in rejecting this long-term form of pollution.
Regarding nuclear technology, these small modular reactors are the way of the future. Westinghouse has a nice blurb on their website about it. Of course theirs is for military use, but the technology is scalable. And it is in the process of being implemented with these new permits.
I agree with you that SMRs are the only reasonable nuclear future; the days of building giant concrete and steel plants are over, people need to let that go. It's totally reasonable to support the development and deployment of SMRs (at the relatively small, prototype-level scale they're being built at) and also to accept that adding solar, wind and batteries at large scale is the most productive use of most grid investment funds at the moment.
My pessimistic take is that if SMRs get cheaper, they're going to do so right around the time battery storage gets *really* inexpensive -- and that will eat a lot of the use-cases for SMRs. But that might not pan out, and having two alternatives is much better than one!
I have worked in nuclear and I think the SMRs seem a bit overhyped. Building a bunch of smaller reactors is unlikely to be much more efficient than building a single big one. And no matter how cookie cutter you supposedly make it, most of the plant will have to be engineered from scratch. Every location has it's own makeup water and discharge situation. For instance, when GE builds natural gas plants, those can be mostly cookie cutter because they just need fuel and air. But if you want to build it as a combined cycle with a heat exchanger, all that stuff has to be engineered for the location. I hope I'm wrong though, and we should definitely try.
I agree. A number of companies have developed modular reactors at a cost of about a £1 billion each. Far better use of the nuclear technology than the big installations that take years to install and then years to decommission.
The French government was forced to step in, bail out EDF and nationalize the private investment that had gone into the French nuclear fleet. At great cost to the country.
In some cases, the reason for the nuclear outages was delayed maintenance that couldn’t be performed sooner because of Covid, and in many other cases it was due to drought and summer heat: not enough cool water was available in the rivers needed to cool down the reactors, so they had to shut them down. Nuclear can be just as intermittent and weather dependent as hydropower, solar or wind.
If the French are being forced to shut down reactors because they're dependent on (overheated) rivers for cooling, maybe they ought to be building some cooling towers?
Maybe a nuclear engineer will correct me, but I imagine that if a reactor is designed to be cooled a certain way, you can’t change the cooling method after it’s built.
I mean...you could...but there's no way it would be worth it. I haven't seen any articles about drought affecting cooling for nuclear plants, just the ones about maintenance.
Cooling towers use air to cool hot water yes, but much of the water evaporates in the process, concentrating minerals in the water in the heat exchangers. So to prevent mineral buildup that would foul the exchangers you need a regular source of new fresh water and to discharge the old water.
Yes, and since the reactors in question take in water from rivers, it's available.
I was being a little cheeky though, because the OP asked about cooling, and I pretended the issue was about the hot water being put back in the river. France's need is for lowering the temperature of the tepid river water at intake, which is harder, and would reduce the energy efficiency of the plant.
I think the learning curve question has a simple and direct answer. Learning curves require reproduction of the same or almost exactly the same thing.
The failure of the US Free market nuclear industry model is failing to realize the immense power of standardized designs. The French went with standardized nuclear designs based around the US Navy standardized designs.
Ever wonder why America Navy operates 99 reactors and 79 nuclear-powered warships – sustained learning curve and design improvements around a central design.
Every new competitve design and build is a high risk one off. Standardize everything. Change N=1 to N= 10, then 100 has enormous benefits in lower cost, replacement parts, Standardized inspection and training. And common established risks.
Solar panels are taking advantage of "samenese" and consider how many PV cells are made annually? A solar cell makes about 1 watt. The world added about 400 billion watts. The cumulative amount is trillions. And that drags along productivity in invertors, heat management, etc.
It's also why mega plants will always be more economical to install ans maintain that roof top.
This is a nice story that completely fails to address the premise it starts with: if solar+batteries is so great, then why is Microsoft investing billions to restart a behemoth 70’s technology plant?
Because restarting a behemoth plant that was operating up until just a couple years ago can sometimes be cheaper and faster than building new plants, even if the new plants are better and more efficient once they’re up and running.
> Exelon was operating Unit 1 at a financial loss since 2015. In 2017, the company said it would consider ceasing operations at Unit 1 because of high costs unless there was action from the Pennsylvania government. Unit 1 officially shut down at noon on September 20, 2019.
>Unit 1 decommissioning was expected to be completed in 2079 and would have cost $1.2 billion, but in September 2024, Constellation Energy, the owner of the Unit, announced plans to invest $1.6 billion to bring the facility back online. The plant is expected to resume operations in 2028.
They are really rich, they invest in almost every viable form of energy production. There are contracts for small module reactors and enhanced geothermal if they ever come true, as well as a bunch of conventional solar and wind contracts.
Because we can’t build it fast enough. Lots of capital is needed and PJM has some of the worst lead times for connecting to the grid. Restarting TMI is a stopgap (maybe measured in decades) until we build enough.
So in 1994 this young Army officer with a master's degree in Nuclear Engineering was faced with a decision - continue on to a 20+ year career with the Army until retirement eligibility, or leave the service to work in the nuclear industry. I personally loved the challenges and really high end science of the American nuclear industry at the time, but this was also the era of skyrocketing costs and angst around nuclear power caused by Three Mile Island. I finally decided to stick with the uniformed service not because of nuclear safety or cost issues, but because the greatest problem that America could not solve the spent fuel storage problem. The spent nuclear fuel problem is a uniquely and exceptionally American problem like single-payer health care and repeal/redefinition of the 2d Amendment - successfully modelled in democracies around the world, obvious technological solutions available to solve severe and important national problems and utterly unobtainable due to the political design of these dis-United States of America. It was galling to me, for example, that the State of Nevada preferred active underground nuclear weapons testing than to open Yucca Mountain to spent fuel shortage. I wasn't going to hang the rest of my working professional life on that kind of political idiocy and cowardice.
I completely agree that solar and battery technology and cost improvements has made the nuclear renaissance less vital to fight climate change, although simpler, safer, common reactor designs could help a lot. The spent fuel problem is still (fort years later) unsolved. The much older, retired Army officer me has max permissible solar panels on my roof and a large battery backup system in the basement. It wasn't cheap, but it offers me long term outage backup, peak utility rate offset, and in a pinch I could go along haul with just the solar I product, although some of my appliances would be off-limits. I would love to see a revival of nuclear, but in the meantime I'm getting my electricity for almost free - that includes charging three hybrid cars.
Spent fuel is a radioactive health risk and costs money to guard and monitor. Who gets the accumulated waste of closed plants? It’s irresponsible can-kicking. Responsible leaders plan for the life cycle.
I think it’s largely a marketing thing. Bring technologists round on solar by rebranding it as high-orbit fusion power plus visible-spectrum beam transmission.
I appreciate your point and have been arguing for nuclear for a long time. Transmission lines take a long time. Land use, water use, is on the side of nuclear. It's foundational science. States or countries need to assess their own energy profile needs. Land, water, geography, population density. Going all in on one technology is like creating another energy cartel. I like my supply chains diversified and research dominance spread in many baskets. You omitted the obvious. Human rights.
That's an analysis of supply chains from start to finish with a cost/benefit analysis, no greenwashing, negotiated fair incentives. So we don't look like hypocrites. Nuclear will be a tool in the tool kit. The idea that we will be more responsible with fusion and somehow our human failings or past mistakes will makes us kinder and gentler people is silly.
Southward et al’s essay (which was excellent) seems to rely on 2012-ish data to support the contention that the cost for firming solar is really high. It’s not, and the latest (unpublished) battery prices I have seen are crazy low. I do think specifically in the UK, though, is probably going to be one of the best places to build at least a bit of nuclear due to the small land area and poor insulation.
BP means incredibly poorly insulated housing, resulting in heat going out through all sides. For North West Europe, wind plays counterpart to solar (you have wind when sun not out), so battery needs limited.
While I agree with the general point that renewables > nuclear for a number of reasons, I think you are a bit too blasé about the potential for extended periods of low renewables output. We can say with a good deal of certainty that there will be multi day periods where the wind will die down and there will be little solar generation over a large geographic extent (look up the term ‘dunkelflaute’. This will particularly be the case once we’ve electrified heating and peak conditions shift to the winter.
My firm (E3) and others have done quite a bit of work in this space, both long-term studies have of these dynamics and more near term studies of how to keep the lights on as we add renewables in the near term. On both timescales, the key need is to capture that the ability of renewables and storage to provide firm capacity declines for reasons that are hard to explain in a Substack reply. But the steep drop offs in those resources reliability contributions doesn’t come until high levels of deployment in most cases. So on the margin, solar plus storage is cheaper than nuclear for the foreseeable future. Longer term, there needs to be some form of long duration energy storage or other non energy limited resource (carbon capture, geothermal, hydrogen, biofuels, and yes, nuclear) to cover those low renewables periods mentioned above.
Frankly, I think dunkelflautes are overrated as an issue.
If we’re going to net zero, we’re going to need a lot of hydrogen/synfuel/biofuel, for things like aircraft, shipping, high power agricultural machinery, and sundry other applications.
If we’re making these in quantity, we can make and store a little bit more of them and feed them in to peaking plants essentially the same way we do now. Yes, they will be expensive to run, but that’s not a huge problem when you’re running them a few days a year on average.
I’d also point out if your energy system has sufficient electricity production to make synfuels in these kinds of quantities, just turning off the electrolysers during a dunkelflaute will free up a lot of electricity.
If large-scale direct air capture becomes a thing, that’s another potentially huge source of electricity demand that could just be turned off in a dunkelflaute.
Aren't dunkelflaute a big problem in northern Europe (like the Germany where the word comes from) because peak electricity demand is _already_ in winter there, largely because how short the winter daylight hours are in that part of the world leads to greater demand for lighting?
To give an example, where I live in northern England there is only 7 hours and 10 minutes of daylight on the winter solstice.
It definitely can be a bigger problem in Northern climates. But I think lighting is something of a second-order concern. The main issue is that on that day in Northern England, you'll only have 7 hours and 10 minutes of, very poor from a W/m2 perspective, solar generation. So, absent a very large overbuilt of renewables for the rest of the year, there won't be enough energy to charge your batteries. And that really becomes a problem given the potential for a sustained dip in wind production at the same time. As Robert Merkel noted, other forms of storage like hydrogen or synfuels could help here in a couple of different ways. As I noted, the bigger issue from a load perspective will be electrification of heating, which will have a much bigger impact on demand than lighting, particularly during very cold weather.
"Nuclear still has important uses — in particular, where land and sunlight are scarce."
To defend the foundations article on the UK, that does describe us fairly well. The LCOE map shows the UK as being one of the last countries to have solar + battery be our cheapest power. Every part of the UK is further north than every point of the lower 48.
A lot of our solar could come from interconnecters to Spain (roughly the same latitude as New York) or even north Africa (roughly Texas / california) rather than being built domestically.
Fission waste is only really dangerous for a few years. After that it is really difficult to be harmed by it unless you do something idiotic like stay very near it without shielding for days. What idiot would do that? The fact that no one has ever been harmed by nuclear waste should be a data point to take note of. Is there any other waste with that track record?
I live in Houston TX. Noah, you have mischaracterized the Texas power situation. Wind is just as important for electricity power in Texas as solar, with natural gas. coal, and batteries acting as a backup.
The agency that runs the Texas electric grid, ERCOT, rates the monthly power capacity of various energy sources as follows:
Solar 25,961 MW
Wind 39,525 MW
Battery 8,779 MW
Natural Gas 67,066 MW
Coal 14,713 MW
Nuclear 5,268 MW
ERCOT can select which of these energy sources it will use at any given moment. My understanding is that they use solar, wind and nuclear first and then use other sources to fill in.
Interestingly, solar and wind complement each other almost perfectly. Sun shines during the day and the wind on the North Texas Plains blows most robustly at night. Generally speaking, wind produces as much power at night as solar does during the day, but last night appears to have been an exception.
Crunch time comes at dawn and dusk when both wind and solar are weak.
https://www.ercot.com/gridmktinfo/dashboards
Wait...I didn't say wind wasn't important! :-)
Did you get the wrong version of that "Our World In Data" chart? The one that's currently posted ( https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff1994403-4baf-47fd-9171-abc0e646e3df_850x1063.jpeg ) doesn't have solar or wind on it at all, just fossil and nuclear.
I think I've seen the version you meant to have here in past posts.
It's equal to solar in Texas. Because wind and solar dovetail so nicely, we talk about *renewables* in Texas.
Here in the PNW the wind picture is very different. I've seen entire weeks with zero wind generation. It's not abnormal to have zero for more than 4 days at a time. https://transmission.bpa.gov/Business/Operations/Wind/baltwg.aspx
The costs of solar are decreasing much faster
It's not just the cost of solar. It's the cost of transmission of that power hundreds of miles to a destination and it's also the cost of long-term storage of solar power for places that have no sun whatsoever during winter. It's also the fragility of all those systems. If you can get generate power close to home, perhaps by wind or nuclear, you avoid those costs and risks.
Places that get no sun whatsoever during the winter probably shouldn't depend on solar. But for areas that do get sunlight, hardly any "close to home" generation is as cheap, easy or effective as photovoltaic solar panels.
Texas electricity is equally dependent at this point on wind and solar. The wind blows most reliably at night and solar takes care of the day. The two energy sources are absolutely complementary for us.
My guess is that it depends on climate. I think that Scotland and northern Europe will always do better with offshore wind. Noah does not seem to appreciate how dark it is from northern France northward during the wintertime. You just can't add more solar. I spent one winter in Gottingen, and I was told that you cannot expect to see the sun between December and April. Overcast. I did not see the sun between December and late February when I left.
A lot of North Americans forget how far north Europe is: Madrid is comparable in latitude to New York City, Rome to Chicago, Nice to Toronto, London to Calgary, Liverpool to Goose Bay and Scotland to the Alaskan panhandle.
Reminds of a conversation I had with a Californian in Detroit. He told of feeling depressed and found out why when the sun came out in Spring.
And Detroit (despite its frigid winters) wouldn't actually be that far north by European standards, which is what would matter where SAD is concerned.
It has a name— Seasonal Affective Disorder (SAD)— three out of six of my immediate family members have it.
Nothing like lived experience!
Hmmm. Low density. Efficient within only certain latitude ranges and certain climates. Zero production for half the day. Sounds crazy to me no matter how cheap panels get. It will be useful but only dominant in certain regions.
The ideal Solar location is low to mid latitude, located close to but a couple of time zones west of major population centers in a thinly populated region the residents don’t mind paving over with panels.
Western North Africa (Morocco, Western Sahara, Algeria, Mauritania) could be hooked into Europe’s grid and provide solar power after sunset to Germany and E Europe (actively and then via battery). The same could be said of Spain, but they’re likely less willing to destroy their countryside.
It might *sound* crazy, but in practice it works.
Someone said Morocco?
https://en.wikipedia.org/wiki/Solar_power_in_Morocco
Seems very likely, but it's important to remember that construction is decided by market demand. The more wind and solar that goes online the more they have to compete with each other for prices, after all the sun is always going to shine at the same time for everyone in a geographic region and the wind is going to blow at the same time. In many places the electricity prices can already go to zero when the wind is blowing hard enough, which makes it less cost effective to build a new wind turbine. More battery storage should even these things out some, it's just important to remember that things don't remain static over time.
I’d like to see an article about long term battery storage. What is current what might be actual in a few years and long term.
What nuclear would have advantage over solar is a bad volcano. An asteroid strike or god forbid a nuclear winter. I don’t know what a war between Iran and Israel might look like. How many nuclear bombs might go off and what it might do to the atmosphere and therefore solar collection.
In thinking about the nuclear saver rattling going on none of the above are ridiculous.
While natural disasters might leave the electric grid basically unimpaired, it is widely expected that warfare between major powers would involve deliberate disruption of electric grids. That might make any means of electric generation ineffective. Not to speak of rendering cryptocurrencies useless.
I'd never really thought about that, Kathleen, but that really makes sense. Factories that can supply their own power locally place a country in far better strategic position than those that require centralized power distribution, which is easy to disrupt in a military conflict.
We really should be transitioning to a fully modular grid, with lots of hyper-local solar and wind integrated below the level of the neighborhood substation, where you're stepping down to 480V, and some batteries -- the size of a Megapack or two -- at each of those substations. The grid will then serve to transfer energy between the places that are overproducing and the places that are underproducing, equalizing how full each node's battery bank is.
Right now the incentives are all wrong for that, for reasons David Roberts talks about a lot: https://www.volts.wtf/p/this-new-organization-wants-to-remake
The problem remains with IBM powering a data center off the Three Mile Island reactor is that a nationwide grid is still needed for the distribution of that data.
There is no _great_ solution for long-term storage yet.
My suspicion is that what's going to happen ultimately is that the learning curve is going to proceed so far that just absurdly over-building PV + storage to be adequate in winter will mean we're oversupplied with energy in summer, and that means you can start doing fun stuff like hydrolyzing water and storing hydrogen (long term hydrogen storage is also pretty expensive right now, but I'm optimistic people are going to figure out better solutions with stuff like metal-hydride honeycombs), or building hydrocarbons out of water and the CO2 in the air (which can then be processed through a fuel cell later).
There also are some interesting technologies that may pan out for long-term battery storage, but my sense is that unless PV and lithium batteries _slow down_, in terms of how rapidly they're getting cheaper, the financial logic is going to point to just building ridiculous amounts of that, rather than investing in fancier tech. But we'll see. As long as the politics and geopolitics don't screw it up, there's a bright future to be had, regardless of exactly which tech wins out.
I'm with you on nuclear power, but in the event of a nuclear weapons exchange large enough to cause global cooling for an extended period, even if your country were not a party to the exchange, whether you have solar or nuke plants would likely be the least of your worries.
I think market forces will indeed push most of the world to adopt solar + batteries. But rich countries that foresee their energy needs increasing exponentially in the next 20-30 years due to data center needs, EV adoption, and (hopefully) carbon capture technology, will probably adopt nuclear power as their main source.
Yes it's more expensive. But it is also far more efficient for base load needs.
I'm curious how and when this will happen. China is the only country building nuclear at significant scale, and their plan estimates 18% nuclear by 2060, a target they'll probably miss (for the happy reason of over-building of solar and wind.) The rich nations of the West aren't building at anywhere near that scale, and to achieve what you're saying on the scale of 20-30 years would require the buildouts start, basically, yesterday.
The elimination of most coal plants over the next 10 years (per EPA rules, if not overturned) will result in shortfalls in baseload power and there is no feasible way batteries will replace this in that timeframe.
I hope we build lots and lots more nuclear plants to replace those coal plants, but battery storage for the grid is exploding right now. California alone built like 8 GW in 4 years. I used to be pretty bearish on grid battery storage, but it seems to be working.
“weeks-long lulls in…sunlight“
If you’ve ever spent time in the British isles, you know that’s not as pessimistic as it sounds. Americans might have a hard time appreciating how gray England is. Think of the whole country as Seattle.
I wonder if this is partly why the graph shows the British Isles is one of the few places where solar will not be the cheapest former energy in 2027.
The solar doesn't need to be generated in the British isles. It can be imported from southern Europe and even North Africa. https://en.wikipedia.org/wiki/Xlinks_Morocco%E2%80%93UK_Power_Project
Yes but ... energy security. I would avoid too much reliance on undersea pipelines or cables.
When I constructed square miles of solar arrays in Saharan Africa in my mind's eye, I immediately saw AK-47 armed hoodlums holding the electricity supply of Europe for ransom. Political stability is a factor.
Yes exactly. Here in the middle of England we have months long lulls in sunlight every year. If we had enough battery capacity to cover a few weeks of darkness we would still have to overbuild the solar capacity by about a factor 7 to get through the winter. Of course wind helps because it has the opposite seasonality to the sun. But wind has less of a learning curve than solar because a lot of it is built on site, not in a factory. So I think we will still need nuclear.
What's the POINT of the "but ..." The policy implications are the same whether one is a "retrofuturist" or nor. We need to remove the regulations that do not pass a cost benefit test from nuclear power generation and distribution, just like we should from geothermal or solar or wind. And we should pay producers the same amount more for producing a KwH of zero CO2 emitting electricity by taxing the net emissions of CO2 whatever technology is used. Neither Noah or I know whether this will mean nuclear power will remain a niche energy source or not. If we coud predict which technologies will be invented, they would already have been invented. Maybe near zero marginal cost nuclear power will produce the energy needed to remove CO2 from the atmosphere so inexpensively there will be no need to "decarbonize" anything.
But however optimistic or pessimistic you are about technology X or Y, the policy is the same: give the correct disincentive for emitting net CO2 and methane and evaluate each investment decision according to cost benefit analysis.
All that said, given the systematic suppression of nuclear power development for decades, it's not absurd to think that the marginal returns to policy reform are greater for nuclear power than for solar or wind?
Sure, we should expect better policy to make it so our nuclear power plant costs are more in line with the costs to build nuclear power plants in other countries
Nuclear energy may not produce CO2, but it can produce nuclear waste that takes tens of thousands of years to lose its radioactivity. No one has been able to convince me that the anti-nuclear movement wasn't correct in rejecting this long-term form of pollution.
Thx
Regarding nuclear technology, these small modular reactors are the way of the future. Westinghouse has a nice blurb on their website about it. Of course theirs is for military use, but the technology is scalable. And it is in the process of being implemented with these new permits.
I agree with you that SMRs are the only reasonable nuclear future; the days of building giant concrete and steel plants are over, people need to let that go. It's totally reasonable to support the development and deployment of SMRs (at the relatively small, prototype-level scale they're being built at) and also to accept that adding solar, wind and batteries at large scale is the most productive use of most grid investment funds at the moment.
My pessimistic take is that if SMRs get cheaper, they're going to do so right around the time battery storage gets *really* inexpensive -- and that will eat a lot of the use-cases for SMRs. But that might not pan out, and having two alternatives is much better than one!
I have worked in nuclear and I think the SMRs seem a bit overhyped. Building a bunch of smaller reactors is unlikely to be much more efficient than building a single big one. And no matter how cookie cutter you supposedly make it, most of the plant will have to be engineered from scratch. Every location has it's own makeup water and discharge situation. For instance, when GE builds natural gas plants, those can be mostly cookie cutter because they just need fuel and air. But if you want to build it as a combined cycle with a heat exchanger, all that stuff has to be engineered for the location. I hope I'm wrong though, and we should definitely try.
I agree. A number of companies have developed modular reactors at a cost of about a £1 billion each. Far better use of the nuclear technology than the big installations that take years to install and then years to decommission.
“France’s nuclear reactors made it less vulnerable to the cutoff of Russian gas during the Ukraine war.” - FALSE
France’s EDF lost nearly 20 billion dollars in 2022 because half their nuclear fleet went offline and in order to avoid massive blackouts they had to purchase gas on the open market at a moment of historically high prices due to the Russian war of aggression: https://world-nuclear-news.org/Articles/EDF-posts-record-loss-in-France-due-to-reactor-out
The French government was forced to step in, bail out EDF and nationalize the private investment that had gone into the French nuclear fleet. At great cost to the country.
In some cases, the reason for the nuclear outages was delayed maintenance that couldn’t be performed sooner because of Covid, and in many other cases it was due to drought and summer heat: not enough cool water was available in the rivers needed to cool down the reactors, so they had to shut them down. Nuclear can be just as intermittent and weather dependent as hydropower, solar or wind.
By contrast, the investment made by many European countries (most importantly Germany) in solar energy allowed them to avoid billions in additional gas costs when prices spiked due to the consequences of the invasion of Ukraine: https://www.carbonbrief.org/guest-post-solar-power-saved-the-eu-29bn-this-summer/
If the French are being forced to shut down reactors because they're dependent on (overheated) rivers for cooling, maybe they ought to be building some cooling towers?
Maybe a nuclear engineer will correct me, but I imagine that if a reactor is designed to be cooled a certain way, you can’t change the cooling method after it’s built.
I mean...you could...but there's no way it would be worth it. I haven't seen any articles about drought affecting cooling for nuclear plants, just the ones about maintenance.
Here’s one that does a good job of describing the curtailment of nuclear output from high river water temperatures in 2022: https://www.catf.us/2023/07/2022-french-nuclear-outages-lessons-nuclear-energy-europe/
And it’s still happening, again this year: https://www.ans.org/news/article-6268/french-nuclear-plant-lowers-output-due-to-hot-river-water/
It's entirely possible to cool hot water using air. It's not as cheap as free, but it's not far off.
Cooling towers use air to cool hot water yes, but much of the water evaporates in the process, concentrating minerals in the water in the heat exchangers. So to prevent mineral buildup that would foul the exchangers you need a regular source of new fresh water and to discharge the old water.
Yes, and since the reactors in question take in water from rivers, it's available.
I was being a little cheeky though, because the OP asked about cooling, and I pretended the issue was about the hot water being put back in the river. France's need is for lowering the temperature of the tepid river water at intake, which is harder, and would reduce the energy efficiency of the plant.
I think the learning curve question has a simple and direct answer. Learning curves require reproduction of the same or almost exactly the same thing.
The failure of the US Free market nuclear industry model is failing to realize the immense power of standardized designs. The French went with standardized nuclear designs based around the US Navy standardized designs.
Ever wonder why America Navy operates 99 reactors and 79 nuclear-powered warships – sustained learning curve and design improvements around a central design.
Every new competitve design and build is a high risk one off. Standardize everything. Change N=1 to N= 10, then 100 has enormous benefits in lower cost, replacement parts, Standardized inspection and training. And common established risks.
Solar panels are taking advantage of "samenese" and consider how many PV cells are made annually? A solar cell makes about 1 watt. The world added about 400 billion watts. The cumulative amount is trillions. And that drags along productivity in invertors, heat management, etc.
It's also why mega plants will always be more economical to install ans maintain that roof top.
This is a nice story that completely fails to address the premise it starts with: if solar+batteries is so great, then why is Microsoft investing billions to restart a behemoth 70’s technology plant?
Because restarting a behemoth plant that was operating up until just a couple years ago can sometimes be cheaper and faster than building new plants, even if the new plants are better and more efficient once they’re up and running.
> Exelon was operating Unit 1 at a financial loss since 2015. In 2017, the company said it would consider ceasing operations at Unit 1 because of high costs unless there was action from the Pennsylvania government. Unit 1 officially shut down at noon on September 20, 2019.
>Unit 1 decommissioning was expected to be completed in 2079 and would have cost $1.2 billion, but in September 2024, Constellation Energy, the owner of the Unit, announced plans to invest $1.6 billion to bring the facility back online. The plant is expected to resume operations in 2028.
https://en.wikipedia.org/wiki/Three_Mile_Island_Nuclear_Generating_Station?wprov=sfti1#
They are really rich, they invest in almost every viable form of energy production. There are contracts for small module reactors and enhanced geothermal if they ever come true, as well as a bunch of conventional solar and wind contracts.
Because we can’t build it fast enough. Lots of capital is needed and PJM has some of the worst lead times for connecting to the grid. Restarting TMI is a stopgap (maybe measured in decades) until we build enough.
This is just further question begging. Microsoft has the capital. Its data centers are self-contained.
So in 1994 this young Army officer with a master's degree in Nuclear Engineering was faced with a decision - continue on to a 20+ year career with the Army until retirement eligibility, or leave the service to work in the nuclear industry. I personally loved the challenges and really high end science of the American nuclear industry at the time, but this was also the era of skyrocketing costs and angst around nuclear power caused by Three Mile Island. I finally decided to stick with the uniformed service not because of nuclear safety or cost issues, but because the greatest problem that America could not solve the spent fuel storage problem. The spent nuclear fuel problem is a uniquely and exceptionally American problem like single-payer health care and repeal/redefinition of the 2d Amendment - successfully modelled in democracies around the world, obvious technological solutions available to solve severe and important national problems and utterly unobtainable due to the political design of these dis-United States of America. It was galling to me, for example, that the State of Nevada preferred active underground nuclear weapons testing than to open Yucca Mountain to spent fuel shortage. I wasn't going to hang the rest of my working professional life on that kind of political idiocy and cowardice.
I completely agree that solar and battery technology and cost improvements has made the nuclear renaissance less vital to fight climate change, although simpler, safer, common reactor designs could help a lot. The spent fuel problem is still (fort years later) unsolved. The much older, retired Army officer me has max permissible solar panels on my roof and a large battery backup system in the basement. It wasn't cheap, but it offers me long term outage backup, peak utility rate offset, and in a pinch I could go along haul with just the solar I product, although some of my appliances would be off-limits. I would love to see a revival of nuclear, but in the meantime I'm getting my electricity for almost free - that includes charging three hybrid cars.
Spent fuel is not a problem. It does not take much space and is mostly stored next to nuclear plants.
Spent fuel is not a technical problem, it's a political problem.
Spent fuel is a radioactive health risk and costs money to guard and monitor. Who gets the accumulated waste of closed plants? It’s irresponsible can-kicking. Responsible leaders plan for the life cycle.
You can recycle solar panels. Recycling nuclear waste is possible in breeder reactors which are politically unpopular.
I think it’s largely a marketing thing. Bring technologists round on solar by rebranding it as high-orbit fusion power plus visible-spectrum beam transmission.
I appreciate your point and have been arguing for nuclear for a long time. Transmission lines take a long time. Land use, water use, is on the side of nuclear. It's foundational science. States or countries need to assess their own energy profile needs. Land, water, geography, population density. Going all in on one technology is like creating another energy cartel. I like my supply chains diversified and research dominance spread in many baskets. You omitted the obvious. Human rights.
https://www.sciencedirect.com/science/article/pii/S221462962300227X
https://ourworldindata.org/safest-sources-of-energy
https://thecommonwealth.org/news/blog-integrating-human-rights-renewable-energy-transition
That's an analysis of supply chains from start to finish with a cost/benefit analysis, no greenwashing, negotiated fair incentives. So we don't look like hypocrites. Nuclear will be a tool in the tool kit. The idea that we will be more responsible with fusion and somehow our human failings or past mistakes will makes us kinder and gentler people is silly.
https://www.energy.gov/ne/articles/doe-study-finds-replacing-coal-plants-nuclear-plants-could-bring-hundreds-more-local
https://www.eia.gov/outlooks/aeo/pdf/electricity_generation.pdf
The last pdf is really worth a look.
Southward et al’s essay (which was excellent) seems to rely on 2012-ish data to support the contention that the cost for firming solar is really high. It’s not, and the latest (unpublished) battery prices I have seen are crazy low. I do think specifically in the UK, though, is probably going to be one of the best places to build at least a bit of nuclear due to the small land area and poor insulation.
I think you mean inSOLation rather than insulation?
BP means incredibly poorly insulated housing, resulting in heat going out through all sides. For North West Europe, wind plays counterpart to solar (you have wind when sun not out), so battery needs limited.
Correct, I meant insolation - that’s autocorrect trying to be helpful! I don’t think I can edit the comment now.
While I agree with the general point that renewables > nuclear for a number of reasons, I think you are a bit too blasé about the potential for extended periods of low renewables output. We can say with a good deal of certainty that there will be multi day periods where the wind will die down and there will be little solar generation over a large geographic extent (look up the term ‘dunkelflaute’. This will particularly be the case once we’ve electrified heating and peak conditions shift to the winter.
My firm (E3) and others have done quite a bit of work in this space, both long-term studies have of these dynamics and more near term studies of how to keep the lights on as we add renewables in the near term. On both timescales, the key need is to capture that the ability of renewables and storage to provide firm capacity declines for reasons that are hard to explain in a Substack reply. But the steep drop offs in those resources reliability contributions doesn’t come until high levels of deployment in most cases. So on the margin, solar plus storage is cheaper than nuclear for the foreseeable future. Longer term, there needs to be some form of long duration energy storage or other non energy limited resource (carbon capture, geothermal, hydrogen, biofuels, and yes, nuclear) to cover those low renewables periods mentioned above.
Near term example
https://www.ethree.com/wp-content/uploads/2020/08/E3-Practical-Application-of-ELCC.pdf
Long term example
https://www.ethree.com/wp-content/uploads/2020/11/E3-EFI_Report-New-England-Reliability-Under-Deep-Decarbonization_Full-Report_November_2020.pdf
Frankly, I think dunkelflautes are overrated as an issue.
If we’re going to net zero, we’re going to need a lot of hydrogen/synfuel/biofuel, for things like aircraft, shipping, high power agricultural machinery, and sundry other applications.
If we’re making these in quantity, we can make and store a little bit more of them and feed them in to peaking plants essentially the same way we do now. Yes, they will be expensive to run, but that’s not a huge problem when you’re running them a few days a year on average.
I’d also point out if your energy system has sufficient electricity production to make synfuels in these kinds of quantities, just turning off the electrolysers during a dunkelflaute will free up a lot of electricity.
If large-scale direct air capture becomes a thing, that’s another potentially huge source of electricity demand that could just be turned off in a dunkelflaute.
I certainly agree that hydrogen and synthetic fuels are likely part of the solution to the dunkelflaute. Maybe most of the solution.
Aren't dunkelflaute a big problem in northern Europe (like the Germany where the word comes from) because peak electricity demand is _already_ in winter there, largely because how short the winter daylight hours are in that part of the world leads to greater demand for lighting?
To give an example, where I live in northern England there is only 7 hours and 10 minutes of daylight on the winter solstice.
It definitely can be a bigger problem in Northern climates. But I think lighting is something of a second-order concern. The main issue is that on that day in Northern England, you'll only have 7 hours and 10 minutes of, very poor from a W/m2 perspective, solar generation. So, absent a very large overbuilt of renewables for the rest of the year, there won't be enough energy to charge your batteries. And that really becomes a problem given the potential for a sustained dip in wind production at the same time. As Robert Merkel noted, other forms of storage like hydrogen or synfuels could help here in a couple of different ways. As I noted, the bigger issue from a load perspective will be electrification of heating, which will have a much bigger impact on demand than lighting, particularly during very cold weather.
"Nuclear still has important uses — in particular, where land and sunlight are scarce."
To defend the foundations article on the UK, that does describe us fairly well. The LCOE map shows the UK as being one of the last countries to have solar + battery be our cheapest power. Every part of the UK is further north than every point of the lower 48.
A lot of our solar could come from interconnecters to Spain (roughly the same latitude as New York) or even north Africa (roughly Texas / california) rather than being built domestically.
Fission waste is only really dangerous for a few years. After that it is really difficult to be harmed by it unless you do something idiotic like stay very near it without shielding for days. What idiot would do that? The fact that no one has ever been harmed by nuclear waste should be a data point to take note of. Is there any other waste with that track record?