Login
Solar+storage is so much farther along than you think
I chat with Kostantsa Rangelova and Dave Jones, authors of a new Ember report, who find that solar-plus-storage costs have declined so much that it can now provide baseload-level power in sunny cities for less than the cost of new nuclear or even new gas. We discuss why even energy pros are behind the curve on this, how quickly the technology is improving, and why most of the world doesn't see natural gas as a viable option the way the US does.
(PDF transcript)
(Active transcript)
Text transcript:
David Roberts
Greetings, everyone. This is Volts for July 16, 2025, "Solar+storage are so much farther along than you think." I'm your host, David Roberts. As anyone who follows energy debates knows, the biggest and most immediate objection to solar power is always: the sun doesn't shine at night. Here's the thing, though: it is possible to attach batteries to solar panels that will store some of their power during the day and release it at night. If you attach enough batteries to a solar panel, you can spread its output over an entire day.
Voilà: a 24/7 "baseload" source of power. Problem solved, right?
Well, no, skeptics argue. By the time you've added the cost of batteries to the cost of solar panels, the power plant becomes too expensive to compete.
In their defense, skeptics were correct about this until very recently. But, as I say over and over again on this pod, solar and batteries are both getting cheaper so consistently and so quickly that it is literally difficult to keep up. Even energy professionals have trouble keeping up.
And that brings us to a new white paper from the energy researchers at Ember. It takes the current price of batteries — which is, incidentally, down 40% in the last year alone — and runs the numbers on how costly it would be to power various sunny cities in the world purely on solar+storage.
The results are eye-popping. The sunniest US city, Las Vegas, could get 98% of its power from solar+storage at a price of $104/MWh, which is higher than gas but cheaper than new coal or nuclear. It could get to 60% solar+storage at $65/MWh — cheaper than gas.
Today I'm excited to discuss these and other jaw-dropping findings with the paper's authors, Kostantsa Rangelova and Dave Jones. We are going to get into just how cheap batteries are now, what it means for gas, and why utilities and policymakers need to catch up with the current reality.
All right then, Kostantsa Rangelova and Dave Jones, welcome to Volts. Thank you so much for coming.
Dave Jones
Thank you, David.
Kostantsa Rangelova
Thank you for inviting us.
David Roberts
So you guys have this new report out about solar+storage. The idea of solar+storage, I think, is well understood now among energy types, the Volts audience. But the reason I wanted to do this, and I suspect the reason you wanted to write this report right now, is that things are changing on the ground quickly.
Kostantsa Rangelova
Oh yes.
David Roberts
In this area like this, you really, like, your analysis is out of date by the time you've done it, practically. So I really don't think even people inside our world fully appreciate what's going on. And I know that people outside our world are just way behind on what's happening. So I want to catch everybody up. So having said that, I'll throw it over to you, Kostantsa. Maybe you can just say what prompted you to write this at this moment.
Kostantsa Rangelova
Absolutely. And you're right in saying our starting point was, "Okay, batteries are getting so incredibly cheap now that this probably changes everything around solar and battery economics." And we wanted to test that in some sort of a thought experiment, saying, "Okay, what if you don't put any limitations on how much solar and batteries you want to combine? And can you do 1 GW constant electricity with solar and battery? How would that work in different places and how much would it cost?" And we were ourselves surprised by really how amazing the results were. And also you were absolutely right to say, like, the prices we were using, like, already, we are probably going to see another huge fall this year already because we were using average 2024 prices.
So, yeah, this is a very fast-changing environment.
David Roberts
I want to get back to that at the end, the moving target of it all. But let's start maybe with just a little general discussion of the merits of solar+storage. Again, I think this is all quite familiar to us, but I'm not sure everybody has thought it through. Let's just go through sort of like why you would combine solar and storage, why you would attach storage to your solar. I think the obvious one that everybody sort of understands is it extends the amount of time that you can be generating, basically. You can save some generation in the middle of the day and let that out at night.
Kostantsa Rangelova
Yeah, absolutely.
David Roberts
And if there's enough storage and enough solar, you can save enough to be generating all day long, which we'll get to later. So I think that's the obvious one that everybody knows. But there's also this phenomenon of maximizing solar penetration at a given grid connection, which I'm not sure people get. So maybe talk through that a little bit.
Kostantsa Rangelova
Yeah. So picture a solar panel, and it usually produces for just a few hours. When the sun is highest, it produces a peak. And then you have a few, like, shoulder hours where you still have some generation, but it's not as high as the peak in the middle of the day, which means that the grid connection that you have should theoretically match that peak production that only happens for a few hours, but then for the rest of the time, you're not using the grid connection. But if you build even more solar panels and flatten the curve in a way that you're not producing just for a few hours, but you're producing constant electricity with enough solar and then the battery capturing that excess and flattening it out.
This means that you can use the grid connection all the time. And instead of 20%, you can go all the way up to 100% utilization and also sell a lot more solar into the grid.
David Roberts
Right. So you add a bunch more solar, and if you didn't add a bunch more batteries with it, that would push your peak up past your grid connection level, but instead you just store that extra and so, as you say, flatten it out throughout the day. So you have the same sized grid connection with basically five times more solar on it by adding batteries. And by doing that, I mean, to state the obvious, you're spending less on your grid, which, as we know right now, is the constraint, is the problem facing solar and clean energy generally — just space on the grid, the ability to hook up to the grid at all.
So this can take those existing grid connections and just max them out.
Kostantsa Rangelova
Exactly. And also, like, if you're displacing fossil generation, like with solar alone, you're displacing fossil generation only in certain hours.
David Roberts
Right.
Kostantsa Rangelova
But with solar and battery, you can displace fossil generation in all hours. And we are already seeing this happening to some degree in California, where you have so many batteries that you already are displacing gas generation not just when the sun is shining, but also in the evening peak when demand is typically higher.
David Roberts
Right. Well, let's talk about how people are using it now, because this is not a theoretical discussion. These prices are coming down, and so people are using it more. So let's talk a little bit about the use cases that people are doing. So, for instance, you cite some examples of big existing solar fields adding more batteries so that they can add more solar. Talk through that a little bit. What's a good example?
Kostantsa Rangelova
I'm sure Dave would agree that what inspired us in doing this kind of analysis was a real-world project that was announced in January this year. It was a project in the United Arab Emirates where Masdar announced a 1 GW 24/365 electricity project. So they're going to have lots and lots of solar, lots of batteries, and they're going to have 1 GW nonstop electricity coming out of that. And this is supposedly going to be online by 2027, so very, very soon.
David Roberts
Yeah, I read about that. And, you know, like, I see so many of these projects coming out of that part of the world that are so ludicrously big. You know, it's like that long city they're always saying they're going to build. I'm never totally sure how seriously to take these announcements.
Kostantsa Rangelova
Yeah, if you're speaking about the Saudi project, you also have another major Saudi project that actually is already online. So they have this mega project there with 16 hotel resorts and all the additional — it's like a mini tourist village with lots of hotels and stuff. And this is entirely powered by solar and batteries, like a small mini-grid. And this has already been online for a year now.
David Roberts
So the idea of the project is to have a gigawatt of continuous power output throughout the day. So give us a sense of how much solar plus how much storage equals a gigawatt of continuous output a day.
Kostantsa Rangelova
Yeah, so what they're building, they have 5.2 GW of solar. That's the DC capacity. And it's actually on tracking. So what we did in our analysis was fixed. So tracking solar panels produce a little bit more and they need a little bit less storage. So 5.2 GW and they have, if I'm not mistaken, about 19 GWh of battery. So that's a lot of battery.
David Roberts
Yeah, huge amount of solar and a huge amount of batteries to get that continuous output. And I think most people sort of intuitively, when they hear that, are just like, surely that can't be cheaper than just building like a gas power plant, you know, or a nuclear plant or something like that has that continuous output built in. But I guess the whole point of this exercise, the whole point of this report, is to say that "Yes, it is now at least in the neighborhood of cost-competitive."
Kostantsa Rangelova
Well, so basically they announced capital costs around $6 billion. And what we have with our configuration, even though I have to, like, specify that our configuration is slightly different, in our modeling, we have 6 GW of fixed panels with 17 GWh of battery. But it's around the same ballpark in terms of CapEx. And then when you produce the LCOE (Levelized Cost of Electricity) of that, it's actually not that expensive. It comes down to around $100 per megawatt-hour, which for 24/7/365 is not a lot.
David Roberts
Yeah, it's crazy. And what about — I mean, this will come up later too — but another thing I think that strikes people when they hear this is if I'm building a nuclear plant, to do that I just need like a couple of acres. But if I'm doing that with 5.2 GW of solar panels and then 19 GW of batteries, the land becomes an issue after a while. Do you know, like, that UAE project, have they said what the land consumption of that is going to be? I mean, they have a lot of big deserts out there, obviously, so it's helpful for them.
Dave Jones
Yeah, the land area is big. That's the case for solar. That's why everyone would always prefer solar on rooftops, to try to take the strain off the land. If you've got low-grade land, that's obviously a different story. Where potentially if you're a country as big as some of the Middle Eastern countries, that land is not a constraint in the same way. What's interesting is also the material footprint to go with it. So that when you look through specifically from a battery side, we talk about the cost of battery, but it's so much in how that battery is now designed that has got a smaller footprint than ever.
So it's got almost all grid batteries, LFP technology, which means that they're not using nickel, they're not using cobalt. They've got a really good lifetime now. So some of the older generation ones were only kind of several years. They didn't have that many life cycles they could go through, charging cycles they could go through. You're now in a position where many of them have 20-year warranties that go alongside them, with an expectation to last even longer than that, doing as much as a cycle a day.
David Roberts
Let me just highlight that point real quick so everybody gets it. So the old batteries used to be what's called NMC. They had nickel and cobalt and stuff in them. Those were the standard. And now, as Volts listeners are familiar, we have this new kind of battery that's iron-based, basically does not have cobalt in it, does not have the nastiest chemicals in it, called LFP. Generally less energy dense, which makes it not great for cars, even though it's being used in cars, but is great for big stationary grid storage because it's made of cheap materials. And the cycle life, as you say, you get a lot more cycles out of it, so you get a lot more life out of it.
All of which just means, like, building 19 GW of energy storage with LFP versus with NMC is materially lighter on the planet, I guess.
Dave Jones
Materially lighter. I think that people would have concerns if you're using an awful lot of nickel and cobalt within that. You're still using lithium within that. What's been interesting is just how, as we started looking for lithium, how much more lithium there seems to be available out there in the world.
David Roberts
It's cheap now.
Dave Jones
It's super cheap. And people are thinking, "Oh, that's just a kind of phase and it will go back up again." But the US Geological Survey doubled the amount of estimated resources out there compared to five years ago. When you start looking for this, you start finding it. Not just in countries and awkward places like in the ecologically sensitive places in Chile. You're finding it in more and more countries where it is accessible. It hasn't got that environmental footprint before, but as we go through the kind of next part of the battery revolution, we could well be using sodium-ion, so-called salt batteries, which then eliminate that need for lithium.
And some of the quality of the batteries that CATL are now producing at scale, the sodium-ion batteries that they've already got at manufacturing, they're coming up with some really good characteristics that are similar or in some ways better than LFP.
David Roberts
They're comparable to LFP, right? Like lower energy density, but more stable materials, cheaper materials, longer life, et cetera.
Dave Jones
Exactly.
David Roberts
And there's actually a — I learned through your report, because I've had sort of like the corner of my eye on sodium-ion for a long time — but I learned from your report there's actually a sodium-ion battery storage facility built now. They actually built one in China, or they're building it, it's underway?
Dave Jones
Yeah, they have commissioned already and it's not small. And some of that manufacturing capacity has started to step up already. I think what's really here, the reason why this is, I guess, just taking a little bit of a pause building up is that lithium has got so cheap unexpectedly that actually it's kind of gone back. And it was certainly a big US developer — I can't remember the name of it now — that withdrew all the R&D out of sodium-ion because you haven't got that price advantage against lithium when lithium is that cheap.
David Roberts
Yeah, this is — I feel like we learned this over and over again — that like it's just not a good idea to bet a lot of money on materials shortages.
Dave Jones
No. But the battery technology progression has been multifaceted. So you've got that footprint side. You've also got the fire risk side that's got two magnitudes better than it was five years ago for grid storage. The other major development in grid storage technologies is the containerization of them, that they've become in some ways almost kind of plug-and-play. So you have these big shipping containers. They've got everything built into them, all the electricity, all the power system technology, the inverters already. So you don't need expensive engineers to install them, to take months doing that. So the speed at which these can be rolled out and the price at which these can be installed really has taken a big shift.
And because the whole of that kind of shipping container of cells has had so much more development over the last, like, one to two years, this is still a very new area in terms of the scale of these batteries coming up.
David Roberts
They're cramming more cells in the containers now, aren't they?
Dave Jones
Yeah. So the ones at the end of last year were all in the range of five to six megawatt-hours in a 20-foot shipping container. And now the ones, there's new models that are coming out this year that are in the 10, pushing towards 12 megawatt-hours. So halving that footprint of what you would need to install battery, which is super good. I think they've probably hit the limit now of how much you can quite fit in. But that's a huge amount of power.
David Roberts
This is funny. This sort of makes the whole point of the pod — just like cutting the physical footprint of batteries in half in the course of one year. Like, how on earth are you supposed to analyze that industry? You know what I mean? Those are like major movements in price and performance.
Dave Jones
And I was speaking to a European utility last week about their projects that they had been working on that came online last year, and they were all installing the frames themselves into buildings. Like, even the whole concept of it being containerized quite like this is fairly new in itself. So the whole speed at which this innovation is coming along, it's just shifted so quickly.
David Roberts
Yeah, yeah. Wild. And so, Kostantsa, just to wrap up the land thing, did you, in calculating these costs and doing your model here, is land cost in there or do you just sort of put those aside?
Kostantsa Rangelova
So land cost would have been very difficult to do on a global average basis. So because for our modeling, we didn't want to do local prices for every specific place, but we just wanted to focus on the solar resource. So what we did is we added a soft cost surplus that is around 10% that we added on top of everything else. And, of course, depending on where you are, these soft costs include the land, but also some of the engineering and labor costs. So basically, this could vary greatly from one location to the next. So it doesn't really make sense to crank it up and focus on it too much.
Dave Jones
When you're looking at how battery prices are being quoted, you get a dollar per kilowatt-hour that normally, when you talk about electric vehicles, is the pack price. And then when you're talking about grid storage, it's quite a bit more than the pack price. And that's the whole containerized unit that includes the inverter and the power system that goes with it. So that when you hear about BNEF, Bloomberg New Energy Finance, and others talking about an energy storage cost, that's the cost they're referring to. And then on top of that, you've got EPC and you've got land costs that then add a little bit more to that.
So that's the way that different prices are being quoted.
David Roberts
Are any policy supports or subsidies built into this model or is that just separate completely?
Kostantsa Rangelova
As said, this doesn't include any specific policy setup in any of the countries. So basically what this takes is how much does it cost to install the solar panels, to have the batteries, all the equipment. Then we did a global average in terms of the EPC, the labor costs, the grid inspection, like all of these mechanical installation, like all of these things are based on global averages. So that we have a relatively good picture in terms of how much would it cost on average across the world. And then, of course, the caveat is that in some places it's going to be cheaper, some places it's going to be more expensive.
David Roberts
Well, before we get into the actual work, one other question about where these are being used, which is one of the things that's most interesting and intriguing to me about this whole area is it's not just power producers who are installing solar+storage for production purposes. Also big customers, big energy consumers. Talk a little bit about what that looks like. Like what sort of big consumer would install this on site and why.
Kostantsa Rangelova
Yeah, so it really depends. So in some of the cases we have, in terms of real-world cases that we found, so we have, for example, some data centers or some utilities that have a PPA with a data center that have a setup like that, or you could even have something big as a titanium smelting furnace that does also have something like solar and batteries in a microgrid kind of setup where it's powering its use through that.
David Roberts
This is, I guess, implicit, but I'll just draw it out. Like, I think people, when they hear solar, even solar+storage, they have, I guess, sort of almost unspoken associations of sort of like weakness and variability. Like it's not strong enough to do something like titanium smelting, you know what I mean? But, like, you generate the power and store it in the batteries, and that power is the same kind of power you'd get from coal. Power is power. So, like—
Kostantsa Rangelova
Exactly.
David Roberts
Yeah, solar+storage is perfectly capable of doing this always-on steady industrial. It's power like any other kind of power, I guess.
Dave Jones
I think people are still getting their head around what you can do with solar just by itself. Last month you had a tenth of the world's electricity on average across the month coming from solar panels. And that produced more electricity across the month than all of the world's nuclear power plants put together. Solar's already reached this scale already and you can see really battery tearing after that quickly. And yeah, there's a definite credibility gap in people's heads about what they can achieve. And it's only just becoming clear now. And that's why, I guess, when we set about writing this report, we were so inspired by this 1 GW baseload plant coming from the United Arab Emirates, which is just the same as a normal thermal baseload power station, but only with the sun.
David Roberts
Yes. And again, if you're in the UAE and you're getting this gigawatt of power output from this plant, it's not like some different kind of power. It's not some different or weaker or variable kind of power. It's power like any other kind of power. So, Kostantsa, let's talk about the actual modeling, the sort of interesting little kind of toy model you built here, which is illustrative, but you start by sort of figuring out, just like for a unit of solar, how much storage do you need to flatten that solar out so that it's a 24-hour supply.
Kostantsa Rangelova
Yes, exactly.
David Roberts
Talk a little bit about how you did that calculation and what you found.
Kostantsa Rangelova
Yeah, absolutely. So let's start with — you want to have 24 GWh across the day or whatever power it is. Like, your solar panel would only have to produce that electricity for, let's say, 12 hours across the day, but most of the time it will actually be a little bit less. So you need around 15 hours across the day in terms of battery to cover the time when the sun is not shining. And these 15 hours across the day that you need to cover with the battery, you need a little bit more in terms of battery capacity just because you can't really charge and discharge at 100%. Like, to maintain the battery's health, you need to have a little bit of a margin. So you, let's say you —
David Roberts
There are losses, right, going back and forth.
Kostantsa Rangelova
Yeah, exactly. So you have some losses, you can't discharge to zero, you can't charge to 100%. So with that in mind, you need around 17 kilowatt-hours of battery to cover these 15 hours of insufficient daylight.
David Roberts
So that's per solar panel?
Kostantsa Rangelova
Well, yeah, it depends on how much you want to scale it, but yeah, the solar panel will produce enough. So you capture those 15 megawatt-hours or kilowatt-hours or however you like it. So you capture these 15 hours from the peak of the day, and you redistribute them through the rest of the day to get 24 hours of constant electricity. So, yeah, the rule of thumb is 17.
David Roberts
So for a given 5-kilowatt solar panel, you need 17 kilowatt-hours of battery storage to make it into a 24-hour power source, basically.
Kostantsa Rangelova
Yeah. The way we want to think about it is if you want 1 kilowatt constant electricity, you need around 17 kilowatt-hours of battery. And then the amount of solar could be a little bit more also because they also have a little bit of losses. But yes, 5 to 6 kilowatts of solar.
David Roberts
Solar, and that relationship, 5 kilowatts of solar to 17 kilowatt-hours of battery storage, that is an average. So obviously that's going to be different from place to place, depending on cost and climate.
Kostantsa Rangelova
Because not every day is average, yeah. And you have cloudy days. And this is actually the big problem because, like, you can make 24-hour solar easily. Like, you just add battery and you can flatten that solar output to 24-hour constant power. The problem is that on a cloudy day you have less solar than on the average day. So you would need a little bit more battery to bridge that. And then the problem becomes that you need a lot more battery to cover several cloudy days if you want to maintain that constant 1 kilowatt of electricity that you want.
So it really depends on the place. And this is why we wanted to model 12 different cities across the world to actually see how much clouds have an impact. And the answer is surprisingly little.
David Roberts
Yeah, it's wild. So you have this sort of relationship in mind: 5-kilowatt solar panel, 17 kilowatt-hours of battery storage—that gets you a sort of steady kilowatt throughout the day. And you just sort of, like, got to multiply those up to the amount of power you need for a given city. So you ran this for several cities. Talk about what you were trying to find out. Basically, like, for this city, how close can solar+storage get to supplying its power needs 24 hours a day throughout a full year? Is that basically what you were—
Kostantsa Rangelova
Yeah, exactly. So basically we were using real-world hourly solar data. So we have access to this really amazing JRC database which has hourly output for solar PV. So you can click on the map and it gives you the hourly data for 18 years back. So from 2005 to 2023, hourly generation—that's lots of data points. And basically, we run those simulations and I just wanted to make a very small clarification: for our modeling we used 6 to 17 just because we wanted to have a little bit more margin on the solar for the cloudy days and what have you.
So we ran that for 12 cities. And the sunny places, think about Muscat in Oman, or Las Vegas in the US or Johannesburg in South Africa. So places like that get more than 90% of the way to 1 GW constant electricity with those cloudy days. In a place like Washington or Madrid, you're in the high 80s. So 88% in Madrid, if I remember correctly. And then for Washington, it would be somewhere there as well.
David Roberts
And then if you have the bad luck to be stuck in Birmingham, in the UK.
Kostantsa Rangelova
That's a cloudy place. And you barely have, like, two very, very sunny months, like in the middle of the year in summer when it's actually, it is actually sunny. Most of the time it's various degrees of cloudy. So even there you have more than 60% of the way to 24/365.
David Roberts
So when you talk, like, take Las Vegas for example, when you say that solar and batteries can get—what was the number for Las Vegas? 97%?
Kostantsa Rangelova
And Muscat in Oman, is 99. And mind you, there is some variability. So on a given year there is plus or minus 2-3 percentage points depending on whether it's like a super sunny or cloudy year. But on average across these 18 years, Muscat is at 99%.
David Roberts
Right. So when we say Las Vegas is 97%, like, physically you could get that to 100. Obviously, like, you could just continue building solar panels and batteries, sure, to infinity until you got to 100. So the point of the research is not that 100 is impossible. It's that you can get to 97% at a reasonable cost, like within some boundaries of cost. And what were the sort of boundaries? Like, where did you draw the line and say, "This is too high"?
Kostantsa Rangelova
So the way the costs evolve is very interesting because you have this very quick progression. If you just build 1 GW of solar with no battery, you have something like 21% in Las Vegas, just because this is just how much your solar produces. If you have 3 GW of solar with 7 GWh of battery, you get to 60%. And the cost is $40 for the 1 GW, $80 for the 3 GW. So $40 is for 20%, and then $80, actually less than $80, is for 60%. And then to get to 97%, you're already around 100.
So you have this, like, gradual progression. But to get from 97 to 99.4, the price jumps to 167. So there is this point where building more solar and more batteries to just fill these, like, last few small gaps just doesn't make sense. And when you think about it, it's quite clear, because if it's just like two, three very, very cloudy days and you're not going to be using your solar and your battery most of the time, like, they don't pay for themselves.
David Roberts
Right, right, right. And one of the things that's wild is sort of that line, you know, where that sort of gradual rise in price kind of hits an L and goes straight up. It's not a gradual thing. Like, it really is — seems like a point where you're like, all of a sudden, costs —
Start skyrocketing. Yeah, exactly.
Yeah. It seems like there is a very distinct point.
Dave Jones
Yeah, that distinct point is really driven by how much electricity you need to carry across days. And lithium is great, you can slightly overbuild it for what you need across 24 hours. But it's not reaching that point, or probably won't reach that point in a long time, that you would want to have a battery sitting there idle, like carrying across just for a few days of a year that energy onto a cloudy day.
David Roberts
Yeah, yeah. At some point, the expense of holding that energy in the battery just becomes—
Dave Jones
And in a way that's okay. You've got other ways to deal with that. Like maybe we talked about, maybe you don't need to have everything running and you can, like, turn your process off. But of course, there's a way to get a gas or an oil generator in there for a short period. Bear in mind that most of these, the average electricity across the day isn't going to zero. So you still got some of that solar coming in. You don't need to — for a GW of baseload, you don't need a GW of fossil backup in there.
You need kind of 200 to 300 megawatts — covers virtually all of that shortfall, in the case of Las Vegas.
David Roberts
You're Las Vegas and you've got to build gas plants that run 3% of the year, like, that's what they call a "champagne problem." So let's talk about the prices then. So Las Vegas, getting to 97%, I think — I don't have it right in front of me — but I think that got you to $104 per megawatt-hour.
Kostantsa Rangelova
Yes. And here comes the big disclaimer again just to make sure that nobody gets confused: this is global average prices for solar and battery, doesn't include local policies, local pricing.
David Roberts
Right. So it could be lower or higher in Las Vegas, depending on the local market and local incentives and whatever else.
Dave Jones
Yeah. Specifically from the CapEx perspective. So when we look through to the CapEx price of what a battery and of what that solar panel costs, the cheapest countries out there are the ones that can access Chinese tech, because that's the cheapest out there. So when we look through and we look through to see the cheapest battery prices out there at the moment that are being installed, they're in India, they're in Saudi Arabia, and they're in China itself, and they're predominantly using Chinese technology. And the prices coming through are eye-watering. And if you don't have access to that Chinese technology, you're going to be in a different price level.
David Roberts
Yeah. And that's a big delta. That's not, that's not like — the cheapest battery prices you found in the world are substantially cheaper than the more expensive ones. This is not a narrow range.
Kostantsa Rangelova
Absolutely. We're talking $50 to $70 per kilowatt-hour in those, like, cheapest regions. And the average, the global average we used is $165.
David Roberts
Right. So half the global average. And that is just access to cheap Chinese batteries. That's all that is. So, like, you have to assume eventually the world is going to gather around that lower point, right? Because eventually the cheaper tech is going to get out there and disperse it.
Dave Jones
Absolutely. And it is for those kinds of key countries that are bringing in a scale today. So really, I guess California, Texas are probably outliers on that when you compare them to the rest of the world.
David Roberts
Right. But $104 a megawatt-hour, you know, if I'm a Las Vegas policymaker — obviously, you know, with all the caveats, this is not precise, you'd have to go actually do the local calculations in Las Vegas to get a real number — but taking this as a ballpark, $104 a megawatt-hour is not crazy. It's less than new coal. It's less than new nuclear. But we all know the real competitor here is gas. And that's more expensive than gas by a decent chunk. Fair? Yes?
Dave Jones
Fair, if you're in the US. If you're not in the US, we don't have the same price as you do for gas, so —
David Roberts
Oh, that's true.
Dave Jones
So many of the world's gas generators have extraordinarily different economics to you and for sure will be economic against those prices when you look at the current LNG prices.
David Roberts
Do you all happen to know off the top of your head what the average price in the US is right now for gas power?
Kostantsa Rangelova
You mean the LCOE for new gas? I think it's around $76, $78.
David Roberts
In the US?
Dave Jones
I think it was around $70 for new gas. It was a weighted US average — from memory, but I might be wrong on that.
David Roberts
Right.
Kostantsa Rangelova
Yeah. I mean, they claim this to be a global LCOE, but they use US-focused data.
David Roberts
Yeah. Because, like you say, it's much more expensive in other countries. We're wallowing in it here, but, you know, if you have to import it, it's more. But just for Las Vegas — I mean, this is all by way of just giving people a sort of general sense of the quantities involved here. So $104 a megawatt-hour to get mostly solar+storage versus $76ish. So, point being, even for a sunny city, to get all the way there would be a substantial price premium over available alternatives. Yes. It's not—you can't get all the way there for cheap.
I don't want to pretend like we're saying that.
Dave Jones
It is, and I guess what we were trying to do with this paper was trying to explore the kind of, like, the barriers of what is possible, of taking it to the extent and being a little bit kind of almost cheeky with that.
David Roberts
Yeah, we should say no one is actually going to try to power their whole city off solar+storage.
Kostantsa Rangelova
Off one technology, period.
Dave Jones
And when you're looking through and you're looking at the solar farms that are being built out in the world today, it's not a question of whether they're built with battery, it's a question of how much battery comes with them. And every day that goes on, there's more and more battery that's being added to them. So it kind of started off at kind of an hour, then it rolled onto two hours. Now a lot of the US ones have got four hours in them. There's some in Australia with eight hours on them. So this is really a thought experiment of just taking that to the limit.
But already you see that that extension, the amount of hours being stored, is getting longer and longer.
David Roberts
I'm highlighting the fact that $104 is a lot more than $70. But you could also turn around and say that, like, powering your whole city with solar and batteries for $104 a megawatt-hour is something that would have sounded like science fiction, you know, five minutes ago.
Dave Jones
Quite. And there are two different models for how this plays out. We've been talking, like, envisaging this as a big 1 GW project, but really when we're thinking about this, it's also probably an industrial user or even a household user. And if you're supplying that just from solar and from battery, from yourself, the economics is relative to the grid price that you pay. And that's substantially more than the kind of numbers that we've got in our head when we're thinking about the wholesale price.
David Roberts
Yes, yes, that's true. It's much more — from the consumer perspective, I think the economics are much easier. And we should say I'm highlighting the $104 number because that's what it takes to get to 97% in Las Vegas. But as you show, getting to like 60%, you can do that cheaper than gas. Cheaper than gas, which again is mind-boggling even relative to a few years ago. So Las Vegas could get to 60% solar and battery power for prices cheaper than new gas, cheaper than any fossil fuel source.
Kostantsa Rangelova
And it's not just that, it's about the direction of travel of these costs.
David Roberts
Let's get into that. So what I wanted to ask about — so there are two sentences that in your report just made my eyes water, so I want to read them. So one is: "In 2024, average global battery prices dropped to $165 a megawatt-hour, down 40% from $275 a megawatt-hour in 2023." So 40% in one year. That's just raw battery prices. So then the LCOE of the whole project, of a whole solar+storage project together, has dropped by 22%.
Kostantsa Rangelova
Yeah. And this is because the CapEx of the battery is such a big part of the whole LCOE.
David Roberts
Yeah, right. In a year. So literally, like, if you had done this analysis last year, the results would be dramatically different. And it sort of follows that if you do this next year — you know, I've almost come to the point where I worry that these, like, snapshots you're taking, because what you're doing here basically is taking a snapshot of a rapidly moving target, right? And so, like, it's effective in that it conveys to people, "Look how cheap this has gotten." But I sort of worry that, like, you just need to emphasize over and over and over again that, like, "Hey, we're going to be back next year. And we're going to blow your mind all over again."
Kostantsa Rangelova
I couldn't have put it better.
Dave Jones
And we know that's going to happen because we can see the prices going through, like I said, Saudi Arabia, India, and China at the moment, and they're just between a third and half of what that price was at the end of last year.
David Roberts
Yeah, these are not small — I mean, like, 40% is like, you know, there are whole consumer product categories that scrabble and fight their entire lives for decades to get, you know, 5% cost drops. So, like, 40% in a year is just mind-boggling. So I sort of wonder, like, did you or would it be easy to just whack another 22% off the LCOE, just assume that what happened last year is going to happen next year, whack 22% off the LCOE, and then run the whole analysis again. Would it be easy to do that?
Dave Jones
It would. And also there's other parameters which have got better as well. It's very strange getting feedback on a report where you're a bit scared, you've done some peer review, some discussion, and then you release it to the world and you're kind of fearing the worst. And every single piece of feedback is, "You've been far too conservative on this, far too conservative on that."
David Roberts
It's genuinely hard to talk about this stuff without sounding crazy. Like, the reality of it is crazy.
Dave Jones
One of the bits that surprised me an awful lot was on the state of charge, where we know that historically people have only used 80% of the battery, kept the top 10, top 20% unused. And we're like, "No, from what we've read, probably you can leave 10% unused." And now I speak to battery developers and they say, "Oh no, no, you can go all the way up to 100%, down to 3%." So you've only got 3% of unusable battery in total. That's how far many of these parameters have got. And the whole technology as a whole has improved.
David Roberts
Right. And solar panel prices are still going down.
Dave Jones
Solar panel prices are still going down. And also I mentioned this earlier, but the cycles within them — so you get a warranty that's then related to the number of cycles, and that can be for two cycles a day, still well over 10 years. So that's the kind of length of time where degradation was such a major issue within batteries that everyone was worried about. But you don't see that happening in the same way now.
David Roberts
Right. I mean, solar panels are coming down, and that's for a commoditized — you know, like, those are largely commoditized now. But in batteries there's still, like, there's still research, there's still, you know, fundamental research going on. There's still, like, new chemistries popping up, there's still whole new — you know what I mean? I feel like we barely even scratched the surface on batteries.
Kostantsa Rangelova
Yeah. And the deployment is just starting to pick up, and this will definitely accelerate that learning curve.
David Roberts
Toward the end here, you mentioned the off-grid opportunity. And this is one of the things that really stimulates the old imagination around this stuff. So I want to just talk about that a little bit — what you mean by the off-grid opportunity and what sort of, like, what sort of vistas does this open up in those terms? Because everybody, you know, the background is every grid everywhere is overburdened. It's hard to get on the grid, it's hard to get connected to the grid everywhere. So anything you can do around the grid is helpful. So just talk about that a little bit.
Dave Jones
There are many different people that would find that there are different cases for this. The primary one that we had in our head was on data centers. You had this massive rise in electricity demand coming through. You can put them anywhere in the world. Why not put them somewhere where there's lots of low-grade land around that you can install the solar panels, the batteries at enough scale that you can power that whole other thing with maybe 97, 98% fulfillment of supply coming through, and you lose a little bit on the side. But if that's the way to get through, to be able to manufacture it that quick, to be able to bring that power source to market for yourself that quickly, that you don't need to worry about grids, then that's the way to do it.
David Roberts
Have you at Ember talked to the data center people? Because I'm sort of puzzled — I mean, this is slightly orthogonal to our whole topic here — but I'm just puzzled by the data center people. So I read that white paper about powering data centers with solar+storage, which lays out the case, saying it's available, it's cheap, it works, and yet the hyperscalers don't seem to be talking about it. Like, they're off having these idle fantasies about building multiple SMRs and taking over nuclear plants, reviving dead nuclear plants, all this sort of, like, wildly expensive stuff they're talking about. And this is sitting right in front of them—and it seems like they are resistant to it for some reason. So I wonder if you have any insight on what's going on there.
Dave Jones
I think it comes back to how much pressure there is on them to do this cleanly. There's a very easy option there at the moment that they don't need to think about, which is to put a gas power plant to run it from the inside. Don't connect to the grid because a gas power plant can do that for you. So, do you go with the easy option that's dirty, or is there another easy option out there as well which is economic and it's clean, and look how much more imagination does that actually need to take? And I think that even for the companies that come out with quite kind of green credentials on that, that's going to be really tested in the coming months as more investments come out on how data centers are actually made.
I can't begin to know exactly what's going on behind the scenes with all of this, but one of the biggest prerequisites obviously is an awful lot of low-grade land with good solar insolation. So I don't know how much that in itself would be holding back, because that is — if you look at the difference between that and gas, that's the big difference that you would need. But what's really exciting is that so much of this new build is happening within the United States at least — is happening in the South, where you do get opportunities for this in that there is so much better solar resource with land available.
David Roberts
Yeah. And once again, like, down in the Middle East somewhere, there's some mega data center project that's running completely on solar, isn't there?
Dave Jones
Yeah. And what is most exciting for us about solar and battery technology is certainly not what's happening in Europe, where we get huge seasonality. There's a real limit to how far solar and battery by itself can take you. In the States it's unlocked quite substantially, especially if you go further south. But you've got, like, whole massive regions across the Middle East, across the whole of Africa and Mexico, where this is a real possibility — that the whole of, like, so much of the energy system, not just the electricity system, but the whole of the energy system as you work towards electrification, can be powered by these two technologies alone.
David Roberts
Yeah. It's amazing. And sort of, like, I want to ask, like, what's the holdup? Why aren't people just herding to this? And I guess part of it is just that literally, like, it dropped 40% last year. So, like, you know.
Dave Jones
That's exactly it. And that's why, that's why I'm really kind of here excited talking to you about it, because I think that people don't realize how much that technology has got cheaper, how much it has improved, how much that footprint has reduced. And when you understand that, you can actually understand how big this potential for solar and battery by itself is going to be.
David Roberts
Yes, it's just like all that historical talk about how far you can get with variable renewables is just like, "Hello, we mostly solved that. Knock, knock. Hello, we solved that. It's on the ground now. Let's do this."
Dave Jones
So long as you don't live in Birmingham in the UK.
David Roberts
And if you live in Birmingham, you've got other problems.
Kostantsa Rangelova
But yeah, it's just so new and I guess it's already out there, but it will quite fast become mainstream.
David Roberts
Yeah, I mean, the numbers are so compelling, and this is — I come back to this on the pod over and over again — it's like, it was one thing when demand was flat and we were out there begging people to swap out existing fossil fuel resources for clean resources, which is where we were for, like, the last 20 years. But now, like, big companies with lots of money need lots of power very quickly, and they are not romantic about fossil fuels. They don't give a crap where the energy comes from.
So I just feel like the logic of the situation is going to pull these hyperscalers in this direction. I mean, it's got to, right?
Kostantsa Rangelova
Yeah. And it's surprising how quick these kinds of projects come online. Like, as we talked about that Masdar project—like, announced in January 2025, online by '27. That's fast.
David Roberts
And that's again, like, 5 GW of solar and whatever, 19 GW of storage.
Kostantsa Rangelova
Somewhere in the desert.
David Roberts
You can't even build a modest-sized gas plant that fast. You can't get a gas plant in service that fast. I mean, one of the things I meant to sort of touch on earlier — maybe we can just touch on it briefly now — is that while solar+storage is rapidly going down in cost, gas is running into problems. Gas is going the other way. So maybe just say a little bit about that. Gas is also a moving target.
Kostantsa Rangelova
Yeah. I mean, when you look at the LCOE for gas, like, it's around $70, but it used to be somewhere around $60, so the trajectory is kind of going up. Like, probably in 2025, 2026, it's going to keep going up. Like, it's not like it's going to start coming down suddenly. While the prices of solar plus battery are going to keep falling very, very fast.
David Roberts
And the supply chain for gas plants too is gummed up right now. So even if you have cheap gas, just getting the equipment for the power plant is difficult.
Kostantsa Rangelova
Oh yeah. In the US, waiting times are like five, six years if I'm not mistaken.
David Roberts
So if you're talking to a giant hyperscaler and you're like, $104 per megawatt-hour versus $70 per megawatt-hour, but I can get you the $104 power in a year and a half, whereas you have to wait five years for the $70 power — I bet every one of them would go for that.
Dave Jones
And I think that globally what's happening on gas is nowhere near as exciting as what's happening in the United States. If you look at China or in India, they got 3% of their total electricity generation from gas.
David Roberts
Well, they're just importing it from us, right? And we're trying to get them to import more. But the funny thing is the more we export, the more prices go up for us. So, like, the last redoubt of cheap gas is going to go away.
Kostantsa Rangelova
And for them, it's not cheap.
Dave Jones
And they don't want it. So it might look like someone somewhere along the way is making a lot of money because when it arrives at the shores of China or India, it's not cheap. So they don't want to be stepping in. This whole idea that China and India were going to be going through a coal-to-gas switch is just complete nonsense. What they're doing is they're building solar and wind and battery at huge amounts of scale, faster than anywhere else in the rest of the world, and they've got no interest in taking gas. When you look at Southeast Asia, you expect a gas boom coming up there.
That's not happening. When you look in Africa in the last eight months, according to our data, we're tracking solar panel exports — which is a really good proxy for how solar is picking up — in the last six to eight months, there's been this huge pickup in so many countries of solar panels being imported within Africa.
David Roberts
Pakistan, I did a pod on Pakistan a couple months ago. Amazing.
Dave Jones
It's amazing. And that's the first step with Pakistan is obviously the solar. You reach that stage where you get saturation during the daytime hours. And what you started to see is a pickup on the battery side from that. So the latest Chinese export data to Pakistan was in the last three months — in March, April, May — was this huge step up in battery exports towards Pakistan, where you can now see that they're going through that stage two, which is to be able to put batteries. So that is really going to make the issue for them so much more palatable than solar alone.
David Roberts
Yeah. Like, if I'm a policymaker and I'm in one of these places where I don't have a lot of abundant domestic gas, I'm looking at this choice: do I want to build a natural gas infrastructure to import this and distribute it and burn it, which would take, you know, 5, 10, 20 years? Meanwhile, the LCOE of its main competitor is dropping 22% a year. Like, it's just insane for anybody to start heading down the gas path. At the very least, you're like, "Well, let's just wait another year and see if it drops another 22%." Like, it's just crazy now to do long-term things in the other direction.
Dave Jones
And it must be so hard within the US to kind of—the economics are skewed both ways. You've got that gas cheaper than anywhere else in the rest of the world, and you've got the solar more expensive than anywhere else in the rest of the world. So it's really hard to then kind of get your head around why other countries are doing this. "Are they being forced to do this? Why the heck would you build solar?"
David Roberts
They've gone woke.
Dave Jones
You're building solar because solar is the cheapest electricity that humans have ever invented.
David Roberts
Yes. And now we have an openly malign administration that is attempting to exacerbate that very problem. So we're just falling farther behind. I mean, it's one thing to look a year ahead at another 22% drop in LCOE, but look five years ahead or 10 years.
Kostantsa Rangelova
It's like giving up.
David Roberts
You know what I mean? These bets that the US is making, there's just no serious person who thinks —
Dave Jones
Yeah. And they're based on the fact that the world wants to buy imported fossil fuels. And I think that if anything, the countries have learned over the last couple of years with the price volatility and the general instability in the world is they don't want to be bequeathed to importing more oil and gas than they are already. That's not what their achievement is. If you look at China, what China's done, it's trying to turn itself into an electric state. And the reason why it's doing that is it doesn't want to build up a big oil and gas reliance. The sales of oil-powered cars peaked back in 2017. They've been falling for almost eight years now. It's insane.
David Roberts
I mean, one of the just basic facts I think that helps to understand the world in general is just that the number of oil exporters is much, much smaller than the number of oil importers. So, the number of countries motivated to escape oil is much larger than the number of countries trying to stick us with it.
Dave Jones
Yeah, and the US used to be an oil importer, so it's kind of like, it used to be kind of like one of the rest of the world. And it used to be that 80% of the world's population lived in oil-importing countries or fossil fuel-importing countries. And then what happened in the last, I think it was three or four years ago when the US turned to be an exporter of fossil fuels. That number has now come down to 75%. But still, 75% of the world's population live as net importers of fossil fuels.
So just with that mindset, you have such a different mindset from what you would be if you were a petrostate wanting to sell your fossil fuels abroad.
David Roberts
Yeah, we really do seem to have embraced our status as sort of a slowly dying petrostate. The last thing I kind of wanted to throw out there, and I don't really know what my question here is, but this is just something I was thinking about. I mean, you are framing this, and a lot of people are framing this as solar+storage, as though that's kind of like a thing, a single thing, a single power plant with, you know, XYZ capabilities. But really what we're talking about is just adding batteries to grids. I mean, all the sort of capabilities you're getting come back to the batteries.
Right. I mean, it is adding storage to a grid that brings you fundamental changes, more so, I think, than solar, because the electrons coming out of solar are just normal electrons. Like, they're cleaner, but, like, batteries are letting you do different things.
Dave Jones
And what you've got is, on one hand, you've got the cheapest electricity you've ever had, but on the other hand, it's not the most valuable electricity you have because it's only there during the day. Like, it's got an inherent floor. If you can actually make that electricity more valuable so that you can store it and use it when you want it, that's how you unlock solar. And I think that when you're looking at how fast the solar rollout can continue, battery really unlocks a whole new way in which you'll use solar that we're only just touching on at the moment.
David Roberts
Yeah, yeah. I mean, you're looking at these sort of confined cases of like a single off-grid installation or a single city, but sort of like in the limit, theoretically in the limit, the endpoint here is a grid with so much storage on it that it basically renders the timing of generation irrelevant across the board. Do you know what I mean?
Dave Jones
Yeah, exactly. And the purpose of the report really is to show that that's not a pipe dream—that's already happening today and it's already economic today.
David Roberts
Yeah, wild. Truly wild. I mean, the difference between an electricity system where timing is literally everything, right, where, like, matching supply to demand is everything, to a grid where you generate power whenever you want, throw it on the grid, and we'll hang onto it until we need it, is just such a fundamental change.
Dave Jones
And one thing we haven't talked about in people's minds as well is about the grid stability underneath that. What's becoming more and more common now is inverters, grid-forming, meaning that they can actually provide that inertia back onto the system and be able to deliver the ancillary services second by second, millisecond by millisecond, to be able to help maintain the stability of that system. And what you're going to come out of this is not a weaker system; you're going to come out with a stronger electricity system.
David Roberts
Yes, I'm going to do a whole — I'm going to do a whole pod on this, Dave, so it's a bit of a spoiler. But, like, people don't get that, like, you can get the inertia and the frequency regulation and voltage regulation and all these services out of inverters, and it's not even that hard. It's like a firmware update for a modern inverter. It's just software. And you're going to go from having a limited number of spinning masses providing those services to every inverter providing those services. Meaning there's going to be, like, hundreds of thousands, millions of inverters spread across your system, all of which are providing stability services.
That's going to be a way, way, way more stable and predictable and reliable grid than the one today. That's an advertisement for a coming pod there.
Kostantsa Rangelova
Looking forward to it.
David Roberts
Well, thank you two for coming on. I feel like the development in this area, in both solar and battery separately, but then, like, deploying them together, is so fast. We literally do need to do, like, a yearly pod, just so people — you know, these are not small changes year to year. It's like a fundamental market-shifting change every single year. So what an exciting time to be following this stuff.
Dave Jones
I'm glad you've now caught battery fever as well.
David Roberts
Oh, and I've had battery fever for a while now. And it's just like seeing all my dreams coming true around me. What a crazy thing. All right, well, thank you, you two, for all your work and this report and talking through this with us, and maybe we'll talk again soon about more fundamental changes.
Dave Jones
Thank you so much, David.
Kostantsa Rangelova
Our pleasure to be here. Thank you.
David Roberts
Thank you for listening to Volts. It takes a village to make this podcast work. Shout out, especially, to my super producer, Kyle McDonald, who makes me and my guests sound smart every week. And it is all supported entirely by listeners like you. So, if you value conversations like this, please consider joining our community of paid subscribers at volts.wtf. Or, leaving a nice review, or telling a friend about Volts. Or all three. Thanks so much, and I'll see you next time.
