nuclear is not useful today. It is too slow to change output as load changes. We need to focus on storage for all the excess power renewables give at the best case, shifting that to worst case-
Comparing nameplate capacity for generation methods with much different capacity factors is misleading. China generates the majority of its electricity from coal, and is still adding more. They're adding more in renewables than coal by nameplate capacity, but coal likewise has a higher capacity factor than renewables, so it's really about the same. Then they say "increasing the proportion of renewables" because the initial proportion of renewables was close to 0.
Coal is a baseload source but not one you actually want to use.
https://www.cell.com/joule/fulltext/S2542-4351(18)30386-6
Firm low-carbon resources consistently lower decarbonized electricity system costs
• Availability of firm low-carbon resources reduces costs 10%–62% in zero-CO2 cases
• Without these resources, electricity costs rise rapidly as CO2 limits near zero
• Batteries and demand flexibility do not substitute for firm low-carbon resources
Non-tropical equatorial countries don't have meaningful seasonality, so they don't need seasonal storage.
For countries far north of the equator, it's more challenging, but there are multiple tools to address this, including: over-building so you have enough in winter, using wind which is seasonally negatively correlated with solar, importing power over HVDC, and diversifying wind spatially to reduce correlations which drop more than linearly in distance.
For small countries very far away from the equator that have highly variable insolation and limited geography to decorrelate, nuclear may be better. But it cannot be asserted a priori without a simulation study tailored to the specifics of that location. When I said that nuclear is bad, I am talking in generalities about the common case (United States) at current market prices.
The paper that you linked is old, we are dealing with exponential change in the price of storage and solar.
" we are dealing with exponential change in the price of storage and solar."
But not in grid storage. That is still incredibly expensive.
No it doesn't. Why do you just say that? There are simulation studies like CSIRO's work which show that it's still cheaper than nuclear after you account for everything.
The price of grid storage is absolutely falling exponentially with respect to time.
You're welcome to go find your own data of course, you'll struggle to find more direct numbers because significant grid storage has only become affordable in the last few years.
The term vaguely makes sense if there are sources of electricity that output a constant supply that are cheaper than the dispatchable sources of power. Like nuclear was supposed to be (but in the end is not). Or in some very specific locations hydro (without a reservoir) and geothermal are. But as often bandied about as a "type of power that must be filled" it simply doesn't exist. The type of power that must be filled is dispatachable power, everything else is just "well what cheap non-dispatchable sources can we use to avoid using more expensive dispatchable power".
That's the most expensive one, especially if you both need it to be reliable (can't be a battery that can run out) and need it to not emit CO2 (can't be natural gas peaker plants). So what you want is to minimize the amount of dispatchable power that you need.
Which is what baseload does up to the point that it's generating the "base" (i.e. typical daily minimum) amount of load, because then you don't need any of that portion of the load to be dispatchable since the baseload plants will handle it ~100% of the time.
If baseload is half of peak load then you only need 50% of peak load to be dispatchable because baseload sources handle the other half. If you have no baseload sources then you need enough dispatchable sources to handle 100% of the peak load in case that's the only thing available some days/times, e.g. when peak load is after sunset and solar output is zero.
The difference gets even more significant when you consider the possibility of reducing demand through pricing. Suppose raising the price by a moderate amount can lower the daily average demand by 25%; past that you start hitting inelastic demand and would need unrealistically high prices. If you have baseload as 50% of generation then you now only need dispatchable sources to handle 25% of peak load because the other 25% can be achieved through pricing. Without baseload sources you need 75% -- three times as much instead of twice as much.
More than that, it reduces to 25% the amount of long-term dispatchable power you need, i.e. things that cost a fixed amount to have and are also expensive to use but can't run out of capacity. Things you might only use two weeks out of two years, like hydrogen fuel cells or expensive flow batteries; you don't want to need three times as much of those.
Meanwhile you can use short-term dispatchable power (i.e. normal batteries) to do things like handle the peak demand in the early evening while charging them using solar during the day. The period after sunset but before the load falls off is only a few hours long, so you only need 50% of total capacity for ~15% of the hours in the day, i.e. having only 7.5% of daily capacity would mean that you don't have to use demand suppression through pricing unless you have an extended supply shortfall from renewables. Whereas if you don't have baseload then you need enough batteries for ~50% of daily capacity because now you need enough batteries to handle the entire peak demand in the early evening instead of half and the entire demand late at night instead of none. On top of the long-term storage in case the batteries run out.
This is the assumption that justifies the term "base load". In most cases it is simply false.
> especially if you both need it to be reliable (can't be a battery that can run out)
No source of power is 100% reliable, in practice power plants have unplanned outages a single-digits-percent amount of the time. Batteries charged by non-dispatchable power easily match this.
Of course in reality you want a grid with many power plants so that when one is down the rest probably aren't. "Probably" does some work though, for example: https://www.bloomberg.com/news/articles/2022-04-29/half-of-f...
Excess dispatchable power helps make up for when that probably doesn't turn out in your favor.
> and need it to not emit CO2
If we're banning CO2 emissions then up until the last couple of years in most places your only option would be to overbuild nuclear until the point that it was dispatchable and not base load power. Naturally no one did this because it would be obscenely expensive. Now of course you'd have the options of any of the clean energy sources (nuclear, solar, wind, tidal, etc) powering batteries.
Which is the reason batteries can't make for a reliable grid by themselves. There is a strong correlation for every battery being unable to provide power at the same time because they all start discharging instead of charging when demand exceeds supply and the longer that continues the fewer batteries you have that still have any charge, and therefore the higher the withdrawal rate on the remaining ones to extract the same number of total watts, until you have none.
> "Probably" does some work though, for example: https://www.bloomberg.com/news/articles/2022-04-29/half-of-f...
This is why you want a diverse set of generation sources so that you don't get correlated outages. It's better to have nuclear and renewables than just one or the other. France is unique in the world for having more than the optimal amount of nuclear. And it's really more like nuclear is concentrated in France and connected to a European grid containing several other things, but then you can't say that 50% of generation was offline because of this.
Also notice that even the "unplanned" reduction in output was still pretty planned. They find an issue that requires mitigating at multiple plants, so more than the usual number have to be taken offline at some point in the same year, but they get to choose when and can do most of them outside the period of peak seasonal demand, instead of letting the weather choose when.
> Excess dispatchable power helps make up for when that probably doesn't turn out in your favor.
Suppose you had enough battery storage to run the whole grid for 24 hours instead of for 4 hours. Then you unexpectedly discover that half of your power plants (regardless of what type they are) have to be taken offline for two months for repairs/maintenance. How much good are you getting out of 20 hours when you essentially need 30 days?
> If we're banning CO2 emissions then up until the last couple of years in most places your only option would be to overbuild nuclear until the point that it was dispatchable and not base load power.
Hydro is even older than nuclear and there are places where it's close to 100% of generation. If you don't have enough dam sites for that but still have some then using nuclear for base load and hydro for load following has been a solid option with zero fossil fuels since nuclear became available.
But nobody is even proposing to use nuclear as dispatchable power. Its purpose is base load regardless of what form of dispatchable power you use, and the purpose of base load is to reduce the amount of dispatchable power that you need.
This is historically true. However these days natural gas is often cheaper. Of course prices fluctuate, but in general coal has been losing favor because it is expensive. Note that cheap is not just the direct fuel costs, it also includes overhead, that you can throttle natural gas often makes it cheaper overall even when the direct costs are slightly more.
Also unreliable wind+solar is - when working - vastly cheaper than anything else. As such running natural gas at twice the price vs a coal plant can still be a lot cheaper just because you don't have to buy fuel when the wind blows/sun shines.
Dispatchable power is the only sort of power that provides 24x7 power with predictable rates. If there's more demand, you produce more power (at the same cost). If there's less, you produce less so you can sell what you do produce at the same cost.
Those large industrial customers often also end up selling the electricity in the market. Power companies need multi-decade commitments before they are willing to build anything as capital intensive as a nuclear reactor. In that time, industrial customers need to modernize their factories to keep them competitive. Which is often the time they realize it's more profitable to invest that money somewhere else and wind down the old factories.
Eve China, the best nuclear power builders out there, are shifting away from massive nuclear to storage and wind and solar.
Without a major technological innovation in the nuclear power space, I don't see how it can compete, except at the poles and in niches with very poor renewable resources.
Grid storage is cheap enough that Texas, a purely profit-driven grid is now overtaking California in the amount of battery storage deployed. 58GWh of new grid storage was added in 2025 alone, and the growth is still exponentialhttps://seia.org/news/united-states-installs-58-gwh-of-new-e...
I see now that your original post had a fantastical claim that we need weeks of battery storage, which is a fantastical claim. In reality we will need variable amounts of battery but a "week long" battery is not supported by a single detailed grid study I have ever seen.
When I have asked Pell to justify claims of "weeks long battery" the only justifications have been "I heard it from someone else", or napkin math that contains many errors, and in places where there are not errors choices are made to estimate an upper bound rather than a lower bound, indicating that the calculator doesn't understand how napkins math can be useful.
And for super cheap infrequently used storage, here's a recent purchase at $33/kWh of a 30GWh battery by Google:
https://news.ycombinator.com/item?id=47176841
I don't expect such batteries to be used much, despite being a fraction of the cost of current LFP batteries, because we really won't need much storage with such a low power:energy ratio.
To make sure we are on the same page are you envisioning a grid powered 100% by wind + solar + batteries?
Don't forget existing hydro, new advanced geothermal, existing nuclear (until they all reach end of life), etc. etc. etc.
Also, 5 was a typo, it should have been 4 of course! Form energy has been making a major splash for years, having first modeled out the grid case for this type of battery, and ensuring that all materials and battery chemistry allowed a potential route to success. It's iron-air.
Plus, referring to this in terms of "hours" isn't exactly the most informative reference, the more underlying characteristic of storage is if it's meant for daily cycling, meaning that it makes mine every single day from arbitrage, or if it's multi day which is not a super profitable form of arbitrage at the moment. Daily cycling makes a ton of sense with solar. Longer term cycles make more sense paired with wind.
If you use electricity to synthesize gas and then burn that later to generate electricity that is still cheaper than nuclear power.
https://theecologist.org/2016/feb/17/wind-power-windgas-chea...
Nobody builds nuclear power because it's cost effective or green. They either have nukes like China or have purchased an option on nukes (like Iran or Poland).
For that scenario, nuclear is still marginally cheaper (at today's prices at least).
There is still nowhere in the world nuclear power makes economic sense.
The problem with nuclear today is just that it simply hasn't kept pace with the lowering cost of alternatives.
its really not. The new(ie 90s) french reactors are about as fast as Combined cycle gas turbines. Even if its not, it works well enough, spain has shit all battery capacity and manages well enough
but if you have lots of renewables you need batteries ideally, which means the hypothetical argument of "its too slow" goes away because batteries are there to even out the supply.