GBatteries (YC W14) let you charge your car as quickly as visiting the pump
techcrunch.com
techcrunch.com
> The product allows users to charge a 60kWh EV battery pack with 119 miles of range in 15 minutes as compared to 15 miles in 15 minutes today.
> The technology works with off-the-shelf lithium ion batteries and existing fast charge infrastructure by integrating via a patented self-contained adapter on a car charge port
First read says that they're delivering 60kWh in 15min, or they're pumping electrons at 240kW. This is just nonsense -- most deployed DC fast chargers are 50kW... the charger companies and the OEMs are experimenting with 200kW chargers, but those are liquid cooled. So physics says this is nonsense.
Okay, so second read: the 60 kWh is distracting technobabble, and what they're ACTUALLY doing is "adding 119 miles of range in 15 minutes."
The just-announced Nissan Leaf e+ has a 62 kWh battery with 226 miles of range. When you DC Fast Charge, going from 0-80% takes about as much time as going from 80-100% (EV roadtripping is about doing frequent small charging because of this phenomenon, not one-time top-offs like with gas).
The Nissan Leaf e+ is about a 60 kWh battery, and if they're adding "119 miles of range in 15 minutes", that's about half the Leaf's 226 mi range, so call it "30 kWh of charge in 15 min". That means they're charging at an average of 120kW. Okay, now we're back within the realm of physics.
Note the Leaf e+'s (yet unreleased hardware) still only accepts a maximum of 100kW DC fast charging. I think jaguar has experimented with 150kW charging on the ipace (but again, liquid cooled cables).
So they're saying they found a way to cycle fast charging and get about a 20% improvement in average charging rate from the current kinda-top-of-the-line tech, and only for the first 50% battery capacity.
Not nothing, but hardly "as quickly as visiting the pump", especially since whatever tricks they use likely won't continue for the next 50% of the battery.
We've demonstrated 5 min to 50% charge, 10 min to 100% - I’ll edit this to include a demo that we’ve done at CES 2019, and that we’re now exhibiting at the Detroit Auto show demonstrating.
[EDIT] Video here: https://youtu.be/kSLrqR4TfnU Note: Charging times can vary by ~1 minute here and there, because the algorithm is adaptive and characteristics of batteries change from moment to moment.
>First read says that they're delivering 60kWh in 15min, or they're pumping electrons at 240kW. This is just nonsense -- most deployed DC fast chargers are 50kW... the charger companies and the OEMs are experimenting with 200kW chargers, but those are liquid cooled. So physics says this is nonsense.
>Okay, so second read: the 60 kWh is distracting technobabble, and what they're ACTUALLY doing is "adding 119 miles of range in 15 minutes."
There’s a mistake in the article, in both cases it should be in 5 minutes; should be corrected shortly.
In this example, we’re talking about a 60kWh battery with 238 miles of range (Chevy Bolt). Right now the car charges 90miles in 30 minutes, or 15 miles in 5 minutes because the manufacturer limits the charging rate for the batteries in order to preserve their life. Our technology can enable these Li-ion batteries to charge in 5 min to 50%, or 119 miles. This example assumes that the car goes to a fast charging station that has enough power to charge the vehicle in this time; fast charge infrastructure that’s currently being put up by companies like Ionity (350kW), Electrify America, or Charge Point (500kW capability).
The problem with batteries today is not charging speed; it’s possible to charge any battery quickly, but the faster you charge a battery the faster it will degrade. Our technology is able to decrease the irreversible chemical reactions that happen during charging, so that the same batteries can be charged fast without compromising cycle life.
There was a lot of concern with the first generation EV's about battery life, a lot of which the data has shown to be misguided now that the first gen's are coming off-lease. How long is the process to convince the OEMs this won't degrade their battery packs?
In the battery industry there's a saying; liar liar battery supplier. That definitely doesn't help things move fast!
Automotive OEMs do move slowly, but once they see value things definitely speed up - A lot of it has to do with being in front of the right people.
(And if you don’t want to overhead the cables or the connector, you care about current squared, giving you an added incentive to charge at near constant current or perhaps to charge some cells at a different rate than others.
What’s the trick here?
Edit:
> Our technology is able to decrease the irreversible chemical reactions that happen during charging, so that the same batteries can be charged fast without compromising cycle life.
This seems to imply it's doing something to the battery chemistry? Maybe a brief pulse of high-rate discharge every now and then to help balance things out?
That's usually within the domain of the battery management system. There's not much the charger itself can do (there is intelligent communication between EV and charging equipment, but not as granular as the view the BMS has).
It sounds like this would push the battery harder than they're designed for. Does using your charger void the warranty?
Short of physically swapping out batteries, charging an EC will never be as fast as pouring gasoline into a tank. Transferring that much energy that quickly over wires is ... I don't want to stand anywhere near it. Fillup time is not an are where EC should try to compete.
Diesel has a specific energy of 48MJ/kg and weighs 0.832kg/L.
48MJ converts to 13.33kWh. My diesel tank will take 70L from nearly empty. I've never timed it, but it's definitely substantially quicker than 10 minutes to fill.
So I'm transferring about 776kWh in to my diesel tank in <10 mins.
70 * 0.832 * 13.33 = 776.33kWh
So it would take a 1MW charger three quarters of an hour to transfer the same amount of energy. <gasp>
Did I get any of that right?
The article was corrected to say 5 minutes. See also replies from timsher upthread.
https://www.greencarreports.com/news/1118575_gm-working-on-u...
I believe their “secret sauce” is allowing existing packs which charge slower to get “up to speed”. If so, that’s a pretty limited market as it’s likely that within a few years, all EVs sold will charge faster than the listed speeds. They’ll likely have a maximum tota addressable market of a few hundred thousand cars with only a few thousand real consumers. Most people buying electric cars with slower chargers are doing so because those vehicles fit their lifestyle and wouldn’t have a massive incentive to buy a new charger.
We've demonstrated 5 min to 50% charge, 10 min to 100% charge - below is a demo that we’ve done at CES 2019, and that we’re now exhibiting at the Detroit Auto show demonstrating this.
[Video here: https://youtu.be/kSLrqR4TfnU Note: Charging times can vary by ~1 minute here and there, because the algorithm is adaptive and characteristics of batteries change from moment to moment.]
The problem with batteries today is not charging speed; it’s possible to charge any battery quickly, but the faster you charge a battery the faster it will degrade. Our technology is able to decrease the irreversible chemical reactions that happen during charging, so that the same batteries can be charged fast without compromising cycle life.
We’re working on enabling existing packs to charge as fast as possible. We’re not able to fast charge at these levels without approval from the OEM as they have limits built-in. We need to either be integrated inside of the vehicle, or on the charger but have the vehicles "approval" for charging at these higher rates, as at Level 3 and up it's DC to the pack.
One of the other challenges you may face is the lack of consumer knowledge about how charging affects cycle life. Since consumers tend to charge batteries in ways that dramatically reduce cycles, this is potentially huge.
Figuring out how to create demand from consumers will be very important as OEMs don’t really have any motivation here. Cycle life isn’t published nor really considered by consumers when they buy devices with rechargeable batteries.
https://twitter.com/atomicthumbs/status/1032939652393521156
If it's to believed, the charge rate limit from a Supercharger is already maxed out -- they're pushing energy into the battery as fast as the charger can supply it (and taking some safety risks in the process).
This same company has been around for a few years with spurious claims about improving laptop battery life with their AI algorithms. So far nobody has seen evidence that it works, and nobody is lining up to buy them.
My car (2017 Honda Accord) has a 500+ mile range on a tank of gas. It'd take over an hour to charge to 500 miles.
Even at 15 minutes that's a lie. It takes no more than 2-3 minutes of actual pumping to fill my 500+ mile tank of gas.
Also in NJ it definitely doesn't take more than 5 extra minutes waiting for the attendant 95% of the time.
that said, even at a full tank, that's only 220 miles for me, so unless I'm leaving Oregon on a trip, a 15 minute charge sounds amazing.
the good news is that I've not seen a charging station that had an attendant.
Here's one thing that stinks about current/last gen electric cars. I drove with my wife from my apartment in Nob Hill to Santa Cruz, using up about 92% of my range. I was able to find a charger and get fully recharged for my trip back. All of that is fine. What was disturbing, was that while I was climbing those initial hills out of Santa Cruz, my range figure dipped way below my miles to go figure. As I left behind that initial climb, my range figure then went comfortably above my miles to go. This is one place where some judicious AI could help. (Along with integration with GPS/Maps and destination information. I should also update my electric car to technology ca. 2019.)
Here's an article from 2015 when they introduced that feature: https://www.teslarati.com/closer-look-tesla-firmware-6-1-tri...
https://insideevs.com/ionity-debuts-350-kw-ultra-fast-chargi...
Volkswagen probably will be the biggest producer of battery electrics within 3 years. They own 12 car brands (including Porsche and Audi https://www.volkswagenag.com/en/brands-and-models.html). VW also has a stake in Ionity and they own Electrify America and Electrify Canada.
If this charging technique really does work then I imagine VW will license it.
* Poor power density
* Relatively limited nominal power output
(disclaimer: not a subject matter expert)
Worked just fine back in 2013.
The people's habits were the only obstacle: owners wanted to retain their own battery pack, instead of getting pre-owned one left by somebody else, in unknown condition. And/or drive over a weird contraption.
[1] (loud music warning) https://www.youtube.com/watch?v=H5V0vL3nnHY
You have to trust the charging station that you’re getting an equivalent pack in trade. Are incentives aligned to ensure that?
The difference, of course, is that a scooter takes batteries small enough to swap by hand.
The same cannot be said of EV battery technology.
So why the Tesla demo video? According to [1] just doing the demo entitled them to $90 million of Californian "ZEV credits", and they never believed in or planned to roll out the technology.
[1] https://www.quora.com/Whatever-happened-to-Tesla-battery-swa... apologies for the Quora link
The big issue is that assembling loose cells into a battery pack is a hard enough operation. It must achieve good mechanical properties (the pack must be rigid enough), good electrical contact with the battery terminals (bad contact would ruin a pack, could start a fire as a bad contact could overheat) and good cooling (battery packs heat, so afaik all designs use liquid cooling, which needs very good thermal contact between the cells and the cooling assembly). Loose cells would essentially mean you're feeding the loose cells inside the pack and the pack has some robotic assembly arms that it uses to put the cells in the right places while it achieves the three objectives.
With some different, taylor made cell design you could make this self-assembly easier, maybe even make it mechanically passive - have the cells orient vertically and slide into their places, and then they're locked. But the current cell design is definitely not chosen for this.
Has this AI approach been used in other charging applications? Curious to hear more details on what's going on behind the scenes.
Here in Norway we have more electric cars per head than anywhere else and most people wake up to a fully charged car.
It is really rare to need to charge the car at a public charger unless you have nowhere off road to park your car. What is most needed in most places is somewhere to charge the car overnight or while you are at work. Such chargers need no new technology, no extra infrastructure beyond a simple socket in a post beside the road or in each parking space in car parks.
I mean, I know in the abstract sense that it's ideal if they ever get fully drained and, but that's not all that catastrophic. I'd consider it normal use in most cases—if manufacturers don't want the charge to ever dip that low, they should make the battery stop working earlier.
Which brings me back to my original question: how does one treat a battery poorly? Or perhaps more accurately, how would people treat batteries they rent more poorly than ones they own?
In a world where consumers were highly cognizant of maintaining battery health, and only fully drained/charged their batteries when absolutely essential, I can see how rented batteries would be a problem. In practice, I doubt many consumers actually think about this, if they're even aware of it to begin with.
Ergo, I don't forsee rented batteries dying significantly sooner than owned batteries, because most consumers aren't going to change their behavior.
It appears that it's for some of the same reasons why people don't want to just swap their car - when you have a large asset that can be in a variable condition, you don't really want to be swapping it for another one randomly.
My Tesla Model 3 gets 75 miles in 15 minutes at the supercharger.