Noah Smith has an interesting substack on how America seems to have missed the battery revolution1. I think Noah, and to be fair most other commentators, are missing the point. We don’t want batteries, we want efficient energy storage.
We don’t want batteries, we want efficient energy storage.
There are three basic use cases for batteries and they are quite different. That means that there may be different ways to get the benefits required. What are these use cases?
powering (generally small / portable) electric/electronic devices
storing power from intermittent sources like wind to meet general electrical demand
moving things (cars, trains, planes etc.)
Only use case 1 requires a (generally lithium-something) battery. These batteries are of course now ubiquitous and they are what Noah is most concerned with when he talks about the US missing the battery revolution. That’s partly because he sees the same (or at least similar) battery technology being required for use cases 2 and 3. As I explain below that’s not actually true.
Note I am not, in this article, discussing whether replacing fossil fuel powered stuff with non-fossil fuel powered stuff is sensible to prevent glowball worming, running out of oil or whatever. Politicians of all sorts seem keen on reducing fossil fuel usage so the goal here is to see what is the best way to achieve that.
Little devices
It is worth noting that use case 1 is not even close to problem free because these batteries are extremely dangerous if they are caught in a fire and it is relatively easy to get many of them to combust on their own. See the recent I-15 battery fire that was caused by a truck carrying batteries overturning or the EV transporter ship fire last year and any number of other fires caused by lithium batteries of one sort or another (use your search engine of choice). Note that airlines/the FAA, for example, have banned lots of lithium batteries from hold luggage due to the fire risk.
It’s not just new batteries either. People casually throw them away and that regularly causes unwanted fires in the recycling/waste disposal sector.2
A new report from the National Waste & Recycling Association (NWRA) and Resource Recycling Systems (RRS) estimates that more than 5,000 fires occur annually at recycling facilities. Materials recovery facilities (MRFs) process single stream recycling and, increasingly, these facilities experience catastrophic fires due to lithium-ion batteries erroneously placed by consumers in with their recyclables.
Every day, MRFs receive dozens of lithium-ion batteries due to public misconceptions about how to properly dispose of them. As lithium-ion battery usage grows, so will the risk of fires.
“Lithium-ion batteries are in more items than we might think,” said NWRA Interim President and CEO Jim Riley. “Besides recycling facilities, these batteries are a threat to the entire solid waste and recycling system, from collection to disposal, and impact our members every day. Given the risk they pose, it is important consumers understand that lithium-ion batteries should not be placed in their waste or recycling bins and should be disposed of properly through the right channels.”
The report also noted that the increased risk of MRF fires has driven up the cost to insure these facilities. The rate of catastrophic losses has risen by 41% over the last five years. As a result, insurance has increased from less than 20 cents per $100 insured property value to as much as $10 per $100 insured, as providers began to understand the threat to MRFs from fires, according to members of the insurance industry. The risk of fires and the cost to insure against them is expected to rise in the coming years as the use of lithium-ion batteries continues to grow exponentially.
So the battery revolution is not without issues even for use case 1. Now you may well ask why do people use lithium batteries (in their various flavors) instead of something else if they are so dangerous. The answer is simply energy density. You can store a lot more energy in a smaller, lighter package than you can with other battery technologies. The old Lead Acid, NiCad etc. batteries are larger, heavier and store less energy. Typically at least an order of magnitude less - i.e. they store the same amount of energy in something that is 10x the size or weight. If you like your laptops running on battery for 12 hours instead of 1 or having a smartphone that weighs a few ounces not over a pound then you need a lithium battery to power it.
So it would be good if we could find a better, safer, and ideally more dense solution for use case 1 but that may be hard. However we can probably adapt to it and the fires it causes because in normal every day use the batteries are incredibly useful and they don’t often burst into flames given how many of them there are in the world.
General Energy Storage
Use case 2 really ought not to need batteries at all and even the best lithium something batteries don’t do a good job at grid scale energy storage. @ChrisBond’s recent substack on South Australia’s grid3 pulls in some real data from recent days (during Australia’s current winter) to see what level of storage is needed to remove all non-renewable energy sources from the grid. He made this graph:
The red line is shortfall. In the current environment this is/would be filled in by power from other states and/or non-renewable fossil fuel generation, but the goal of his analysis is to see what happens if those aren’t there and you want to avoid power cuts and blackouts.
18 July 2024 is where renewables failed. On that day, South Australia needed around 2GWh of power that renewables didn’t have and which were not made up by the existing ~200MWh of installed battery. [Note I may have misunderstood his work, but even if I’m wrong in the specifics I’m sure I’m correct in the general sense of what he is pointing out]
He notes that in addition to the current 200MWh, South Australia is planning to add more in the next couple of years. Both he and I are a bit confused about exactly what is being added but it looks like less that 2GWh. Probably less than half. He notes that there are larger batteries around, like this monster in California
The project is a true renewable energy behemoth, spanning 4,600 acres, comprised of 1.9 million First Solar panels. It holds a capacity of 875 MWdc solar, and nearly 3.3 GWh of energy storage. It has a 1.3 GW interconnection capacity.
This solar power plus energy storage thing was not cheap
The project’s first phase added 346 MWac of solar modules and 1.5 GWh of battery storage. Financing for the the first phase was closed in 2021 and included $804 million senior secured credit facilities. This includes $400 million construction and term loan facility, a $328 million tax equity bridge facility, and a $76 million construction and revolving letter of credit facility. J.P. Morgan is providing the tax equity commitment for the initial phase of the project, with Deutsche Bank leading the construction and term financing.
In 2022, Terra-Gen closed a nearly $1 billion project financing for the second phase of the project. It included $460 million construction and term loan facility, $96 million construction and revolving credit facility, and a $403 million tax equity bridge facility. BNP Paribas, CoBank, U.S. Bank, ING, and Nomura led the funding.
In total, the two financing rounds for the project totaled over $1.7 billion.
$1.7 billion could build quite a lot of alternatives. It should be noted that batteries are not the only way to store energy. Traditionally utilities have used pumped hydro to do this sort of thing - see for example Dinorwig in North Wales. Dinorwig has a capacity of about 9GWh or 3x the capacity of the California $1.7bn battery pack. You have wonder why California could not build something similar to Dinorwig and how much that would have cost. Mind you California seems to prefer removing dams or letting them nearly collapse for lack of maintenance4 than to build new ones.
Having said that, I am aware that there are plenty of places where pumped storage is not possible, but with a bit of thought one might be able to come up with other ways to store energy for times when it is needed. The ideal would be to store it in such a way that it can use existing installed infrastructure as much as possible.
Vehicles
Ignoring the fact that they cost much more than gasoline/diesel powered vehicles, EVs are generally heavier and shorter ranged. That’s because, fundamentally, an internal combustion engine plus a full fuel tank weighs much less than a battery and electric motor of similar general performance.
We can put a pretty precise number on how much by looking at the Nissan Sakura/Mitsubishi eK X EV and the gasoline powered equivalents. These cars are identical in external dimensions and trim etc. The 2WD EK-X weighs 830kg. The 2WD EK-X-EV weighs 1070kg. That’s roughly 30% more. The EV has a range of 180km officially. The gasoline one appears to have a range of slightly over double that - possibly more, it was hard to figure out the tank size, I think it is 20l but it could be larger. The official fuel consumption is over 20km/l (23.3km/l - but I assume that’s optimistic). But 20l at 20km/l is 400km or 2.222x the 180km range of the EV
So the electric version weighs 30% more and goes less than half as far between required recharges. Its hard to see based on that, why anyone would voluntarily choose an EV unless it was a lot cheaper (spoiler alert: it isn’t cheaper) because it’s a lot less convenient. And the inconvenience gets worse in cold weather where EV range drops precipitously while gasoline engines lose perhaps 20%. It is true the EV has greater nominal acceleration but it is rare that a few seconds difference accelerating from 0-60 makes any effective difference to journey times.
Unlike acceleration, the cost difference is significant. And also strongly against the EV. The gasoline car comes in at around two thirds of the price of the EV (JPY 1,664,520 list vs JPY 2,568,500). Perhaps that difference can be made up in operational cost? Given the use cases of these cars are short commute/local errand running for the most part, the likely monthly driving distance is probably going to be around 1000km. At the current JPY 175/l and generously saying 20km/l fuel consumption, 1000km is 50l or JPY8,750. Which works out as an annual cost (for 12,000 km) of JPY105,000.
It will take about 9 years at that price for the gasoline car and fuel to equal the price of the electric car and this utterly ignores the cost of the electricity required to charge the EV - though to be fair I’m also ignoring the costs of oil changes and other maintenance. I also have no idea whether either will need new tires or other repairs. But we do know that after about 10 years of such use the EV will likely need to replace the batteries or be limited to much shorter journeys than the rated 180km (which is not necessarily a killer in the likely use case), while the gasoline vehicle will keep on just fine.
So the gasoline car weighs less, costs a lot less and goes further. Plus it works in the cold and if, for some reason (such as an unexpected snowstorm), it runs out of power in the middle of not much then a gallon of fuel will get it to somewhere where it can be properly refueled. Unlike an electric car which needs a generator and complicated charging electrical equipment to be brought to where it is if it is out of juice. Similarly if a gasoline powered vehicle is driven in places where fuel supply is potentially limited an additional 5 or 10 gallon jerry can can easily be carried at the cost of slightly less storage space. There is no good equivalent for EVs. This recent substack lists 95 reasons why EVs are bad.
There are EV enthusiasts who claim that in a few years batteries will be lighter and more energy dense and so (some of) the problems of EVs will be solved. But when pushed to point to specifics they end up extrapolating trends rather than pointing to a specific technology that will provide us with lighter more energy dense batteries.
The Ideal Solution
What we need for use case 3 (vehicles) is basically gasoline or similar (e.g. LNG/ethanol) that doesn’t come from the ground. If we have that we can keep all the advances in internal combustion engines etc. that have been developed over the last century and a bit and we’re fine. We can, if we want, have small KERS like hybrid systems or perhaps go for interesting systems where the ICE runs to charge capacitors/batteries that drive electric motors (modernized versions of the diesel-electric train locomotive) or we can decide not to bother - there are some good reasons to want electric engines but no good reasons to want to cart around batteries to power them.
For use case 2 (grid scale energy storage) we want the most compact way to store energy and we, ideally, want to be able to use it to remove the instability that widespread use of renewables adds to any grid. That means we want the renewables to store whatever power they generate in something that is not necessarily directly connected to the grid. Ideally this stored energy is also both easily dispatchable and storable for a long time without losses (think storing the energy of summer sunshine for winter nights).
By far the best way to store energy is to create some kind of hydrocarbon fuel with the energy. That fuel can then be transported or stored using existing technologies, all it does is replace the fossil fuel coming out of the ground. For both the power generation/grid energy storage use case and the vehicle use case synthetic hydrocarbons are far better than batteries. Hydrocarbons are (as noted above) extremely energy dense. We also have over a century of experience in transporting them safely and storing them safely. Not to mention lots of experience in putting out the fires when something goes wrong. Lithium batteries are far less energy dense and have way worse characteristics when they start to burn.
This is not a fantasy, there’s actually a company that does this - HIF. Their initial pilot plant in southern Chile takes wind power and turns it into gasoline. There would seem to be no good reason why this should not be extended to other places. I suspect that right now this is an expensive solution but it seems exactly like the sort of thing that will see the cost fall after a few iterations of process engineering optimization.
The synthetic fossil fuel solution removes the need for EVs and reduces the requirement for lithium and other expensive battery constituents. For use case 2 it solves the intermittency problem almost completely - any time there’s excessive wind you start creating fuel - and it potentially reduces the need for expensive complicated grid interconnects to remote deserts and mountains because you can just have a tank of synthetic fuel that gets emptied into a tanker truck periodically. If it produced (m)ethanol then it could even address some of the use case 1 tasks by means of a (m)ethanol powered fuel cell.
But is it efficient?
Possibly not very. But… not it may not be notably worse than other storage methods. Pumping water uphill is not that efficient, though I have been unable to identify exactly how efficient it is. Charging batteries particularly for the last 10% ish) is not very efficient. You can tell this by the way lithium batteries get hot as you charge them up. In terms of straight electrical energy in vs electrical energy out Lithium-ion batteries seem to be about 86% efficient at best - and at the wrong termperatures this can drop considerably. You can of course use some of the power to run a heat pump to maintain optimum battery temperature but this reduces the efficiency.
Further more other parts of the system also have inefficiencies. Conversion from AC to DC (and back to AC) is not efficient, neither is changing the voltage of DC or ensuring that the voltage from . I know people have put a lot of work into reducing the losses in both these areas but you can easily see 10% of theoretical power output be lost to each of these required conversions. Chucking all that together and a large lithium battery for energy storage can easily end up only 50% efficient when you measure energy supplied to it vs energy it supplies.
Given that synthetic hydrocarbons are stable, energy dense and easy to transport and store the fact that you use more energy to make these things than you get from them should not be a reason to not use them.
The problem of course is that there’s a lot of grift and government subsidy sloshing around in the lithium battery world so which technically creating sythetic fuel may be a winner, politically it’s going to be a tough sell. That doesn’t mean it should not be attempted though.
"For use case 2 (grid scale energy storage) we want the most compact way to store energy and we, ideally, want to be able to use it to remove the instability that widespread use of renewables adds to any grid. That means we want the renewables to store whatever power they generate in something that is not necessarily directly connected to the grid."
I agree with some of what you write in this piece but I think you're begging the question in favor of power-to-fuel conversion a bit.
When it comes to grid-scale storage, compactness isn't particularly important at all, and portability can be provided by the grid itself. Power demand is generally very predictable in both time and place, which is why the grid works in the first place. If you need to move a GWh of power from Arizona to LA, the electrical grid is by far the easiest way to do that.
Most of the problems with renewables right now are time-shifting in the domain of 12-48 hours, to handle demand smoothing and cope with normal variations in output. Lithium-ion batteries are becoming cheap enough to be economically compelling here, though other technologies like iron-air may be a lot better in this application.
Also, the more willing you are to accept some fossil fuel use or nuke power, the less power you need to bank to make the math work. This is doubly true the wider your grid can be geographically, so if we get a 12-day storm and freeze in the northeast, we can call on conventional generators in the midwest and southeast.
I have looked at installing a NG genset to handle the occasional (3-24 hour) outages I get 1-2x per year, and for not a lot more money I can get a battery system that could do that *and* store solar output. At current New England electrical rates, such a system reaches break-even in about 12 years, even with our far-from-optimal weather. This is due partly to our very high rates, but also to how much the cost of batteries and panels has fallen, and both continue to fall.
Also, the efficiency of AC:DC conversion has improved quite a bit in recent decades as ultra-high power semiconductors have evolved, and long-distance transmission via DC is in some cases more efficient than AC. This increases the potential for demand smoothing across more of the country, which if done correctly could do a lot to increase reliability and redundancy.
As for vehicles, everyone I know who owns an EV as a second car seems very happy with it. If you're a 1-car household, a plug-in hybrid seems like a better bet unless you don't drive a lot of miles, and I tend to think we are over-promoting EVs relative to hybrids.
Of course if electric:hydrocarbon conversion becomes a LOT more efficient, then sure, it could do a lot, but so far all the processes for fuel synthesis have god-awful efficiency and the economics are much worse than batteries.
Synthetic hydrocarbons, gasoline, interesting but but but, everybody knows, because everybody's been told over and over and louder and louder that CO₂ is destroying the planet!
If only the market were allowed to make these decisions instead of the bureaucrats.