Lower-cost sodium-ion batteries are finally having their moment(arstechnica.com)
arstechnica.com
Lower-cost sodium-ion batteries are finally having their moment
https://arstechnica.com/cars/2024/12/lower-cost-sodium-ion-batteries-are-finally-having-their-moment/
25 comments
> Cheap batteries from earth-abundant materials change that completely.
Not really. Batteries need to be about 100 times cheaper than current ones to make a difference. The way to think is this: how many cycles do you need to recoup the cost of the battery? Let's say you get electricity in the summer for roughly zero cents per kWh, and sell in the winter at 50 cents, and you incur zero losses. Current battery prices are above $100 per kWh, so you need at least 200 cycles to break even. At one cycle per year, that's 200 years. If, however, you charge during the day and sell at night, you can break even fairly quickly, even if the profit per cycle is just a few cents per kWh.
So, current batteries are ok for daily charge-discharge cycles. But not for yearly ones, and no battery will be for the foreseeable future, because the cost needs to go down by a factor of 100, at least.
Not really. Batteries need to be about 100 times cheaper than current ones to make a difference. The way to think is this: how many cycles do you need to recoup the cost of the battery? Let's say you get electricity in the summer for roughly zero cents per kWh, and sell in the winter at 50 cents, and you incur zero losses. Current battery prices are above $100 per kWh, so you need at least 200 cycles to break even. At one cycle per year, that's 200 years. If, however, you charge during the day and sell at night, you can break even fairly quickly, even if the profit per cycle is just a few cents per kWh.
So, current batteries are ok for daily charge-discharge cycles. But not for yearly ones, and no battery will be for the foreseeable future, because the cost needs to go down by a factor of 100, at least.
This is why this isn't the plan. Instead there will be some amount of over capacity at abundant times such as summer for solar which means power will be very cheap or free during some months of the year and batteries will simply be used to match power usage through the day and across a few days.
If batteries were very much cheaper than having just enough power across the year would be the right strategy but we don't so it's not and in a lot of countries the overproduction is only +50% and the much cheaper option.
If batteries were very much cheaper than having just enough power across the year would be the right strategy but we don't so it's not and in a lot of countries the overproduction is only +50% and the much cheaper option.
Absolutely, I don't see batteries being helpful for annual cycles, just for daily. Hence my mention of getting through windless nights. The article I referenced was mainly about that too, suggesting we'd need about a week of battery storage.
For annual cycles, the easier method is simply to overbuild your generation. If you have half as much solar input during the winter, then build twice as much solar.
Here's an article covering a study based on US data, suggesting that four days of storage and a 2X overbuild is roughly optimal there:
https://caseyhandmer.wordpress.com/2023/07/12/grid-storage-b...
For annual cycles, the easier method is simply to overbuild your generation. If you have half as much solar input during the winter, then build twice as much solar.
Here's an article covering a study based on US data, suggesting that four days of storage and a 2X overbuild is roughly optimal there:
https://caseyhandmer.wordpress.com/2023/07/12/grid-storage-b...
> For annual cycles, the easier method is simply to overbuild your generation. If you have half as much solar input during the winter, then build twice as much solar.
At latitudes such as of New York or London the solar input during the winter is 5 times lower than in the summer, or more. That's because there are three effects that all go in the same direction: shorter day, lower inclination and thicker atmosphere for the rays to cross (because they come more slanted). If you build five times more solar than you need during the summer, it means you have to idle 80% of the capacity at that time. Long before this will happen the marginal profit of building a new solar plant will become negative for the investors, and long before that happens the profit will be positive, but the risk too high for the investors to secure financing.
At latitudes such as of New York or London the solar input during the winter is 5 times lower than in the summer, or more. That's because there are three effects that all go in the same direction: shorter day, lower inclination and thicker atmosphere for the rays to cross (because they come more slanted). If you build five times more solar than you need during the summer, it means you have to idle 80% of the capacity at that time. Long before this will happen the marginal profit of building a new solar plant will become negative for the investors, and long before that happens the profit will be positive, but the risk too high for the investors to secure financing.
Sure, if you're Canada, you probably need to go ahead and build nuclear. Even if solar keeps getting exponentially cheaper, it'd be nice to be reasonably efficient with land use.
What has surprised me is that a solar/wind/battery grid is starting to look feasible anywhere, even if not everywhere. It could probably handle a decent portion of the US now.
What has surprised me is that a solar/wind/battery grid is starting to look feasible anywhere, even if not everywhere. It could probably handle a decent portion of the US now.
How many years did it take for them to drop by the previous factor of 100?
Erm, never a factor of 100.
Lithium battery prices dropped from about $700-ish per kWH to about $140 or so. About 1/5 in 10 years and most of the drop was early (about a factor of 1/3 from 2013 to 2016).
Lithium battery prices will continue to drop, but the big gains are over. It's now going to be a slow industrial production decrease over time.
Lithium battery prices dropped from about $700-ish per kWH to about $140 or so. About 1/5 in 10 years and most of the drop was early (about a factor of 1/3 from 2013 to 2016).
Lithium battery prices will continue to drop, but the big gains are over. It's now going to be a slow industrial production decrease over time.
But the history of lithium-ion is thirty years long, and over that period, their price did drop by 99%:
https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-...
The drop has been a fairly steady exponential, so if you look at the last decade, of course most of the drop was early, since that's how exponentials work. As far as anyone can tell, the trend shows no sign of stopping anytime soon.
https://rmi.org/the-rise-of-batteries-in-six-charts-and-not-...
The drop has been a fairly steady exponential, so if you look at the last decade, of course most of the drop was early, since that's how exponentials work. As far as anyone can tell, the trend shows no sign of stopping anytime soon.
They seem (both the article and the press release) to conspicuously avoid mentioning any details about important open problem with sodium-ion: Cycle life/longevity.
> On November 22, China’s Huawei announced a new patent for sodium-ion batteries named “Electrolyte Additives and Preparation Methods, Electrolytes and Sodium-ion Batteries.” The company’s latest work has focused on improving the shortcomings of sodium batteries – such as low coulombic efficiency and poor cycle life – by optimizing the electrolyte formula.
Yes, you've "improved the shortcoming." By how much? The lack of numbers is concerning.
Sodium ion has been in the labs for a while, but at least as of last I saw, they weren't managing more than a couple hundred cycles to 80% capacity. Lithium-ion is now typically thousands of cycles to 80% capacity (and associated higher internal resistance - 80% is just the standard benchmark for "lifespan" given the range of issues you see by that capacity loss). That they're "fluffy" batteries in terms of density isn't a huge problem for a lot of cases, but that their cycle life is "a few hundred cycles" has been a serious issue for any significant application of them.
That they don't mention this in the slightest implies, rather strongly, that they've not fixed that particular problem with it...
For a 70kWh battery pack, 300 cycles at 3 mi/kWh (a midrange average for year round driving) gets you 21MWh cycled through the pack, or 63k miles. Not exactly a stellar lifespan for a car battery pack.
That same pack at a lithium-ion standard 1500 cycles gets you about 315k miles.
> On November 22, China’s Huawei announced a new patent for sodium-ion batteries named “Electrolyte Additives and Preparation Methods, Electrolytes and Sodium-ion Batteries.” The company’s latest work has focused on improving the shortcomings of sodium batteries – such as low coulombic efficiency and poor cycle life – by optimizing the electrolyte formula.
Yes, you've "improved the shortcoming." By how much? The lack of numbers is concerning.
Sodium ion has been in the labs for a while, but at least as of last I saw, they weren't managing more than a couple hundred cycles to 80% capacity. Lithium-ion is now typically thousands of cycles to 80% capacity (and associated higher internal resistance - 80% is just the standard benchmark for "lifespan" given the range of issues you see by that capacity loss). That they're "fluffy" batteries in terms of density isn't a huge problem for a lot of cases, but that their cycle life is "a few hundred cycles" has been a serious issue for any significant application of them.
That they don't mention this in the slightest implies, rather strongly, that they've not fixed that particular problem with it...
For a 70kWh battery pack, 300 cycles at 3 mi/kWh (a midrange average for year round driving) gets you 21MWh cycled through the pack, or 63k miles. Not exactly a stellar lifespan for a car battery pack.
That same pack at a lithium-ion standard 1500 cycles gets you about 315k miles.
Wikipedia's page on sodium-ion lists several commercially-available batteries capable of several thousand cycles.
https://en.wikipedia.org/wiki/Sodium-ion_battery
https://en.wikipedia.org/wiki/Sodium-ion_battery
It does. Can you find any independent confirmation of those claims? I generally regard battery spec sheets as "wishful thinking," based on having done a lot of work with lithium chemistries in the past. Some of them are counting "cycle life" at 80% Depth of Discharge, which... kind of "isn't how that measurement is done."
I don't have any special sources, but since sodium-ion batteries are already being installed in BEVs and power grids, I guess we'll all find out how it's going before long.
No reason a power grid can't restrict their battery storage to 80% discharge, if necessary. It just means buying 20% more battery.
No reason a power grid can't restrict their battery storage to 80% discharge, if necessary. It just means buying 20% more battery.
Afaik, an important component of battery degradadation is the deformation of the anode and cathode as ions migrate between the two electrodes. Sodium ions are much larger that Lithium so this issue is much more pronounced. So in addition to energy/mass, there is another metric that prevents them from catching up to Lithium.
Sodium has a gravimetric energy density of ~4 kWh/kg for complete oxidation (lithium has ~12 kWh/kg and gasoline has ~13 kWh/kg), so there's still plenty of room for improvement.
Dr. Shirley Meng has done some interesting work on anode-free solid state sodium ion batteries - current samples have 350 kWh/kg of gravimetric density, but only retain 70% capacity after 400 cycles.
Dr. Shirley Meng has done some interesting work on anode-free solid state sodium ion batteries - current samples have 350 kWh/kg of gravimetric density, but only retain 70% capacity after 400 cycles.
(typo in previous comment: current samples store 350 Wh/kg)
Another Better Battery Bulletin. Nothing to see here, time to move on.
“We tend to be skeptical of news releases from companies,” he said. He specified that his comment applies to all battery companies." -Venkat Srinivasan
“We tend to be skeptical of news releases from companies,” he said. He specified that his comment applies to all battery companies." -Venkat Srinivasan
Despite everybody's cynicism, batteries have gotten a lot better over the past three decades. It just hasn't happened with any sudden drama. No reason it can't happen for sodium batteries, just as it has for lithium.
I'm glad sodium is having its day in this configuration of using them in batteries.
What I'm really trying to understand is how in the blazes we aren't using Sodium _directly_ as energy transport mechanism.
Using the Castner Process ( https://en.wikipedia.org/wiki/Castner_process ) you run the process of (caustic soda + energy) -> (water, oxygen, pure Sodium).
This process is 'perfect' in that you don't get contaminants; you lose nothing in this reaction. (It's decidedly imperfect in that the reverse reaction occurs during, so H2 wafts off, and it needs ~330ºC to run).
This process has been done to death in the early 1900s, but humanity hasn't done this stuff for a 100 years since the invention of a different process that turns NaCl into Sodium and Chlorine gas, because the latter is _also_ valuable, whereas water aint.
It's... perfect. You ship caustic soda which is relative to other energy processes and carriers not all that dangerous (you don't need to store it pressured for example), cheap, extremely abundant to the desert, there you have a massive solar farm that turns into pure clean water and pure Sodium. That's even better than a hypothetical amazing Electrolyzer, because you _need_ water with those, whereas castner _makes_ water. I hear they can use that in the desert. Oh, and the anodes can be made from iron, which is cheap, easy, and abundant. Unlike the material you need to make your anodes from in state of the art water-to-H2 electrolysers!
Ship the sodium back. All you need is clean water. Toss a bucket of water at it and it poofs back into (Heat + H2 + caustic soda). Again, __perfectly__, no losses. You don't need a reactor or a catalyst or pressure; that process just goes automatically, all you need is a bucket and a vessel.
Storing sodium is, despite the youtube movies, easy. Some paper steeped in oil is all you need to wrap it in. It's stable under all 'warehouse temperatures' (-10 to +70), does not need pressure, and __has zero losses during storage__. It's also quite light.
Some back of the envelope math says that it's energy dense enough; a containership full of sodium contains more than enough energy to pay for the trip and then some - about as much H2 you can make with that as a ship with pressurized H2.
I'm guessing I must have messed up that math because this feels like it solves.. everything. Free energy for all. There's enough Sodium the world over, you can start right now (Sodium is already made in industrial quantities for other purposes, so you can just buy a containership's worth right away). Surely after 100 years of science we can improve on the already functional Castner Process. Solar panels are idiotically cheap and plentiful.
You can convert half the sahara, or the entire east coast of Spain, or the sun belt, into a giant solar farm, just ship coastic soda to it and sodium out. More energy than you'll ever need, and you solved the availability problem because you can store it, at no loss at all, effectively forever, at nearly no cost. Any warehouse can be converted into storage for peanuts.
I'd love to know what I'm missing here.
What I'm really trying to understand is how in the blazes we aren't using Sodium _directly_ as energy transport mechanism.
Using the Castner Process ( https://en.wikipedia.org/wiki/Castner_process ) you run the process of (caustic soda + energy) -> (water, oxygen, pure Sodium).
This process is 'perfect' in that you don't get contaminants; you lose nothing in this reaction. (It's decidedly imperfect in that the reverse reaction occurs during, so H2 wafts off, and it needs ~330ºC to run).
This process has been done to death in the early 1900s, but humanity hasn't done this stuff for a 100 years since the invention of a different process that turns NaCl into Sodium and Chlorine gas, because the latter is _also_ valuable, whereas water aint.
It's... perfect. You ship caustic soda which is relative to other energy processes and carriers not all that dangerous (you don't need to store it pressured for example), cheap, extremely abundant to the desert, there you have a massive solar farm that turns into pure clean water and pure Sodium. That's even better than a hypothetical amazing Electrolyzer, because you _need_ water with those, whereas castner _makes_ water. I hear they can use that in the desert. Oh, and the anodes can be made from iron, which is cheap, easy, and abundant. Unlike the material you need to make your anodes from in state of the art water-to-H2 electrolysers!
Ship the sodium back. All you need is clean water. Toss a bucket of water at it and it poofs back into (Heat + H2 + caustic soda). Again, __perfectly__, no losses. You don't need a reactor or a catalyst or pressure; that process just goes automatically, all you need is a bucket and a vessel.
Storing sodium is, despite the youtube movies, easy. Some paper steeped in oil is all you need to wrap it in. It's stable under all 'warehouse temperatures' (-10 to +70), does not need pressure, and __has zero losses during storage__. It's also quite light.
Some back of the envelope math says that it's energy dense enough; a containership full of sodium contains more than enough energy to pay for the trip and then some - about as much H2 you can make with that as a ship with pressurized H2.
I'm guessing I must have messed up that math because this feels like it solves.. everything. Free energy for all. There's enough Sodium the world over, you can start right now (Sodium is already made in industrial quantities for other purposes, so you can just buy a containership's worth right away). Surely after 100 years of science we can improve on the already functional Castner Process. Solar panels are idiotically cheap and plentiful.
You can convert half the sahara, or the entire east coast of Spain, or the sun belt, into a giant solar farm, just ship coastic soda to it and sodium out. More energy than you'll ever need, and you solved the availability problem because you can store it, at no loss at all, effectively forever, at nearly no cost. Any warehouse can be converted into storage for peanuts.
I'd love to know what I'm missing here.
Seems like there are some others wearing similar thinking caps, albeit 10 years ago
https://doi.org/10.1179/174892407X180928
https://doi.org/10.1179/174892407X180928
The author appears to be unfindable, the abstract is clearly talking about something close to this but also mentions adapting existing UK based fossil fuel powerplants, lowballs a bunch of estimates, and somehow brings up a seemingly completely unrelated process that is also well-trodden ground in the distant past but hasn't been used in a century: A way to take the MnCl2 (Manganese-chloride) that is the waste product of an old timey way to make chlorine gas, and get yourself the manganese back out so you can recycle the process.
Which flummoxes me because the castner process was the one that _does not_ involve any chlorine. Perhaps the author of this paper is referring to the general principle here, which again is a mystery to me because the castner cycle I am referring to (sodium as an energy transport; [sodium + water] -> [heat, h2, caustic soda] to 'consume' it, [caustic soda + energy -> water, sodium] to produce it) doesn't have waste product at all other than water.
How did you find this? It's.. fascinating.
Which flummoxes me because the castner process was the one that _does not_ involve any chlorine. Perhaps the author of this paper is referring to the general principle here, which again is a mystery to me because the castner cycle I am referring to (sodium as an energy transport; [sodium + water] -> [heat, h2, caustic soda] to 'consume' it, [caustic soda + energy -> water, sodium] to produce it) doesn't have waste product at all other than water.
How did you find this? It's.. fascinating.
Just occurred to me: if on the energy-producing side you're dropping sodium into seawater, then effectively you're getting desalination for free.
Yup. And deserts are pretty good places to build massive solar farms.
The more mainstream way of thinking is to electrolyze water into hydrogen and oxygen. Which requires pure water (extremely pure; trying to electrolyze water is difficult and corrodes the anodes and such as is; doing it with salt water? Currently, disastrous, and I believe tons of money is being thrown at that problem with so far no satisfactory results). That means you waste a bunch of your energy on a desal plant just to get clean water out in your desert, and this gets you an ecological disaster as the briny mess (a desal plant converts energy and salt water into plain water... and a boatload of very highly concentrated salt water of course. As the dead sea and such can attest, that stuff pretty much kills an ecosystem, whereever you care to dump it).
I did some math and it's not that much water. But, hey, _its water_. It's more than enough to provide free and plenty water to the few people that live in these places. It's not enough to start farming alfalfa or whatever, though.
The more mainstream way of thinking is to electrolyze water into hydrogen and oxygen. Which requires pure water (extremely pure; trying to electrolyze water is difficult and corrodes the anodes and such as is; doing it with salt water? Currently, disastrous, and I believe tons of money is being thrown at that problem with so far no satisfactory results). That means you waste a bunch of your energy on a desal plant just to get clean water out in your desert, and this gets you an ecological disaster as the briny mess (a desal plant converts energy and salt water into plain water... and a boatload of very highly concentrated salt water of course. As the dead sea and such can attest, that stuff pretty much kills an ecosystem, whereever you care to dump it).
I did some math and it's not that much water. But, hey, _its water_. It's more than enough to provide free and plenty water to the few people that live in these places. It's not enough to start farming alfalfa or whatever, though.
For grid storage, maybe the downside compared to batteries is you're limited to turbine efficiency, but it still seems like a good idea.
Cheap batteries from earth-abundant materials change that completely. Iron-air is another possibility, though I think it's less efficient.
[1] https://dothemath.ucsd.edu/2011/08/nation-sized-battery/