EV Batteries

[steveting99]

Here's another data set to look at. Recurrent Auto (a telematics company that analyses BEVs) has a pool of +50k vehicles and examined 1k rides that have done more than 150k mileage on the odometer. Below is a summary of high mileage rides and average range score - which is an indicator of the high voltage battery's Sate of Health (SoH).

Article from Recurrent Auto is here: https://www.recurrentauto.com/news/1-billion-miles-later

150k miles is equivalent to 241,400 kms - more than what most would do in in one CoE cycle (the exception being PHVers). While Recurrent Auto's dataset is heavily skewed to US of A and the makes/models there, can see that range loss and thus battery degradation is quite satisfactory even for high mileage rides.

Can ICE vehicles at least match these results?

The results of Recurrent Auto ties in with what Generational+ in the UK has been saying about EV batteries in their article here: https://www.generational.ac/2025-battery-performance-index/

Real world data on thousands of rides, with at least a decade of results is showing that EV batteries are lasting much longer than what people think. 

 

Edited February 26 by steveting99
typo

[steveting99]

For Grab & Co. as well as those leasing BEVs, GeoTab has done a analysis covering 22.7k vehicles in 2025. GeoTab started this in 2020 with just 11 models and have now expanded their study to 21 models. Their dataset has a shorter vehicle age of between 1 to 5 years. Below is a are the results on battery degradation based on each model type. 

 

GeoTab article is here: https://www.geotab.com/blog/ev-battery-health/. Interesting that GeoTab's results show that passenger rides (i.e. sedans/SUVs) do better than MPVs. 

The average battery degradation is 2.3% per year. GeoTab also observed the S-curve on battery degradation shown in the graph below from Recurrent Auto.

GeoTab says DC fast charging affects battery degradation the most. Both how frequent one uses DC fast charging vs. AC charging and the power output during DC fast charging are factors to consider.

The correlation is that the more frequent one DC fast charges and with higher DC power output, the faster the battery degradation compared to AC charging.

The above two curves makes logical sense as DC fast charging (high current and high temperature) stress the battery pack the most compared to AC slow charging.

The next largest factor is how frequently one charges the high voltage battery to achieve one full cycle. I.e. from 0% to 100% is counted as a full cycle. So if one does 20% top up charge every two days and after doing this five time over a 10-day period the high voltage battery pack has experienced one charge cycle - not five. Below is the curve on frequent charge cycles.

   

The 3rd largest affect on battery degradation is how often the batteries remain outside the 20% to 80% Sate of Charge (SoC) that vendors generally recommend. This is applicable to those high voltage battery packs that have NMC chemistry. Below is the curve.

 

 

So the basic question to ask is how long will EV batteries last? GeoTab says that based on their 2025 study of 22.7k vehicles, the average lifespan of EV batteries will be around 13 years or more based on observed degradation rates. Again, more than one CoE cycle.

GeoTab also says that modern electric vehicle (EV) batteries continue to perform strongly.

Geotab’s latest analysis shows an average degradation rate of 2.3% per year, affirming long-term battery viability for both individual and fleet use. 

[Ake109]

Interesting that MPVs degrade more.

Is it a sign that MPV drivers probably more fck care the battery care aspect? Anyhow charge, overdischarge, etc?

[steveting99]

Finnish startup - Donut Lab's solid state battery seems to be the controversy of the year in 2026 in EV batteries. Revealed in CES 2026, there's lots of skepticism and doubters within the battery industry - all with valid reasons. Some are even calling Donut Labs claim to be a scam.

First set of lab test on fast charging of Donut Labs solid state battery has been carried out by an independent testing lab, with analysis done by one of my favorite tech YouTubers (Two Bit da Vinci) here:

The tests results looks very similar to that of an NMC chemistry lithium ion battery pouch cell.  

Will be keeping a close eye on the remaining tests and more interested in how Donut Labs can scale up production of their solid state batteries.

It needs to be cost competitive with China's gargantuan battery production and supply chain management.  Otherwise it's going to be a very niche product with limited application(s).

Edited February 27 by steveting99
additional text

[Rm2s]

On 2/27/2026 at 10:02 AM, steveting99 said:

Finnish startup - Donut Lab's solid state battery seems to be the controversy of the year in 2026 in EV batteries. Revealed in CES 2026, there's lots of skepticism and doubters within the battery industry - all with valid reasons. Some are even calling Donut Labs claim to be a scam.

First set of lab test on fast charging of Donut Labs solid state battery has been carried out by an independent testing lab, with analysis done by one of my favorite tech YouTubers (Two Bit da Vinci) here:

The tests results looks very similar to that of an NMC chemistry lithium ion battery pouch cell.  

Will be keeping a close eye on the remaining tests and more interested in how Donut Labs can scale up production of their solid state batteries.

It needs to be cost competitive with China's gargantuan battery production and supply chain management.  Otherwise it's going to be a very niche product with limited application(s).

Donut's claim of 100,000 charge cycles is rather hard to fathom. 

This is more believable?

https://carnewschina.com/2026/02/24/changan-to-install-solid-state-batteries-on-vehicles-before-q3-2026/

Perhaps we will see solid state battery fitted on Deepal models coming to Singapore in couple of years?! 

[inlinesix]

On 2/28/2026 at 10:36 AM, Rm2s said:

Donut's claim of 100,000 charge cycles is rather hard to fathom. 

This is more believable?

https://carnewschina.com/2026/02/24/changan-to-install-solid-state-batteries-on-vehicles-before-q3-2026/

Perhaps we will see solid state battery fitted on Deepal models coming to Singapore in couple of years?! 

We might see new MG4 1st with semi solid state

[Sdf4786k]

https://www.facebook.com/share/v/1CHibuddzQ/

hmm.. if this continues, EV going to have better parking lots than ICE cars LOL

 

[inlinesix]

On 3/2/2026 at 8:56 AM, Sdf4786k said:

https://www.facebook.com/share/v/1CHibuddzQ/

hmm.. if this continues, EV going to have better parking lots than ICE cars LOL

 

Nowadays, why are we slower than Bolehland🤭

[steveting99]

On 2/28/2026 at 10:36 AM, Rm2s said:

Donut's claim of 100,000 charge cycles is rather hard to fathom. 

This is more believable?

https://carnewschina.com/2026/02/24/changan-to-install-solid-state-batteries-on-vehicles-before-q3-2026/

Perhaps we will see solid state battery fitted on Deepal models coming to Singapore in couple of years?! 

Yeah, I'd put Changan coming out with their 400W/kg solid state battery (called the Golden Bell pack or Jinzhongzhao) with a higher probability of success compared to Donut Labs secretive solid state battery.

Changan has the manufacturing ability to do a pilot line before scaling up to mass manufacturing, with a planned capacity of 150 GWhr. This is to bring costs down and allow mass adaption into B&B rides. China's demand for high energy batteries came from NMC and this was 144.1 GWhr in 2025. Changan can effectively capture a substantial portion of this market if their planned mass production of 150 GWhr produces battery packs that are less costly than NMC but a bit more expensive than LFP.

Manufacturing capacity and capability is something that Donut Labs doesn't have - even if independent lab results confirm their outlandish claims.

Interesting that Changan has eight battery cell variants, including liquid, semi-solid, and solid-state types. Will be keeping a close eye on how Changan goes with their Golden Bell (Jinzhongzhao) solid state battery pack.

Edited March 2 by steveting99
additional text and typo

[Rm2s]

On 3/2/2026 at 9:22 AM, steveting99 said:

Yeah, I'd put Changan coming out with their 400W/kg solid state battery (called the Golden Bell pack or Jinzhongzhao) with a higher probability of success compared to Donut Labs secretive solid state battery.

Changan has the manufacturing ability to do a pilot line before scaling up to mass manufacturing, with a planned capacity of 150 GWhr. This is to bring costs down and allow mass adaption into B&B rides. China's demand for high energy batteries came from NMC and this was 144.1 GWhr in 2025. Changan can effectively capture a substantial portion of this market if their planned mass production of 150 GWhr produces battery packs that are less costly than NMC but a bit more expensive than LFP.

Manufacturing capacity and capability is something that Donut Labs doesn't have - even if independent lab results confirm their outlandish claims.

Interesting that Changan has eight battery cell variants, including liquid, semi-solid, and solid-state types. Will be keeping a close eye on how Changan goes with their Golden Bell (Jinzhongzhao) solid state battery pack.

I just read that BYD's strategy is to use sodium batteries on B&B cars (to lower cost of cars) and solid-state on their performance models. BYD's timeline is later, full production will be in 2030s. 

[steveting99]

Donut Labs has released their 2nd YouTube video on testing of their solid state battery with an analysis carried out by one of my favorite tech YouTubers (Two Bit da Vinci)  here: 

Can't make much sense on the purpose of the test of the high ambient temperature - unless to prove that it's not lithium ion NMC chemistry. Is Donut Labs is trying to prove that their solid state cell is safe to operate at high ambient temperatures - compared to lithium ion NMC? Part of their solid state cell did fail at 100 Deg C.

The interesting bit is that most battery cell manufacturers test up to 45 Deg C, with only Samsung going up to 60 Deg C.

Speculation is that Donut Labs is using some sodium-ion layered oxide - similar to CATL's Naxtra sodium battery. The discharge curve for CATL's Naxtra cell looks similar to NMC chemistry and Donut Labs cell.

Manufacturer of Donut Labs cell is Nordic Nano. Apparently the anode material is Titanium oxide with carbon nanotubes to help improve conductivity with an electrolyte containing sodium. 

 

[Ct3833]

On 2/26/2026 at 3:38 PM, Ake109 said:

Interesting that MPVs degrade more.

Is it a sign that MPV drivers probably more fck care the battery care aspect? Anyhow charge, overdischarge, etc?

More like MPV are the group of higher usage people like parents fetching kids, going outing with friends and relatives etc.

[steveting99]

BYD is throwing down the gauntlet to the rest of the industry with their new Blade Battery 2.0 announced on 5-Mar-2206. Article from CarNewsChina is here: https://carnewschina.com/2026/03/05/byd-unveils-blade-battery-2-0-10-70-in-5-mins-10-97-in-9-mins-and-20000-flash-charging-stations-in-2026/

There are two types of Blade Battery 2.0. The short blade version that has around 5% more energy density, has longer cycle life and can do fast charging. The standout feature is that charging time has been substantially reduced to around 5 minutes (10% to 70%). As fast as pumping petrol into an ICE tank. Examples of charging time on some models are shown below:

The operating voltage is quite high at 1,000V and the current is also high at 1,500A. The cable will require liquid cooling as it'll get hot. Combining both voltage and current together to get power makes 1,500kW (1.5MW) of energy being pumped into the battery pack. Cold weather temperature performance (down to -20 deg C) is outstanding - but not relevant to Singapore. It will be relevant in Europe where sub-zero temperatures in the winter are an issue.

For those concerned about how long it takes to full, table below shows around 9 minutes.

Example of the T-shaped flash (or megawatt) charger is below.

Each flash charger comes with a battery pack from 200kWhr to 300kWhr capacity according to CarNewChina article here: https://carnewschina.com/2026/03/08/byds-flash-offensive-5-minute-blade-charging-cracks-nios-swap-legacy/.  Based on a 100 kW grid draw, one BYD flash unit can support roughly 40 rides per day, assuming an average of 60kWhr per charge. This is going to keep Grab & Co. happy and eliminate the complaints from PHVers of not being able to charge a BEV not as fast as pumping petrol into an ICE ride.

If BYD is going to retain their No.1 position in Singapore, this new flash charger is going to be rolled out to fleet operators such as Grab & Co. Who'll want to take advantage of this innovative technology and move their PHVers into environmentally friendly rides. 

There is also the long blade version of the Blade Battery 2.0, that has around +25% more energy density at 210Whr/kg compared to current generation one blade battery that's around 165Whr/kg. Waiting for details on this. The long blade version will be more appropriate for those who are happy to own their rides and do slow AC charging at home/estate.

[steveting99]

The design of BYD's new 1,500kW chargers is ingenious. The charging cable is suspended above on rails thus making the 'T' shape of charger.

Because the cable sits on rails, it will make it easier to pull the cable with one hand to the charging port so ladies won't find it difficult. Much better than current design where the cables are floor placed and at times a bit heavy.

Can see there are additional grey panels next to the 'T' shape charger. Suspect this is where the power electronics and batteries are stored.

[steveting99]

Test 3 from Donut Labs has been released with analysis carried out by Two bit da Vinci here:

Only thing that can be deduced from Test 3 is that Donut Labs battery isn't a super-capacitor with the 10-day charge and hold test.

Interesting that Ricky from Two bit da Vinci says that the charge and discharge curve characteristics from tests 1 to 3 shows a consistent pattern that implies the same battery chemistry. So Donut Labs is highly unlikely swapping around cells of different chemistry to cater for different tests. 

Waiting for results from test No.4

[steveting99]

This end of 2024 report called "Dynamic cycling enhances battery lifetime" logged in Nature Energy: https://www.nature.com/articles/s41560-024-01675-8#Sec6 explains why the battery in an EV lasts longer than what laboratory tests indicate. The test results from the lab is typically shown in the battery curve.

During a battery-cell aging test in a lab, researchers use a constant-current cycling technique, which involves continuously charging the cells at a constant rate for an hour, then fully discharging them in an hour, and then continuously charging again for another hour. It's referred to as the 1C charge/discharge cycle test. But this approach looks nothing like what happens in real life, where there's frequent acceleration, braking, and parking that EV batteries actually endure.

Researchers are under time pressure to find shortcuts when testing the lifespan of battery cells that can last upwards of 18 years. To hasten battery aging, academic researchers typically employ a constant-current cycling approach to battery testing (i.e. 1C), which compresses the process and mimics some EV uses. Examples are public transportation and industrial vehicles that operate continuously and rest for shorter periods of time compared to personal vehicles.

The Standford university researchers found that under dynamic scenarios where there's frequent acceleration and braking; EV batteries can last up to 38% longer than what the lab test under 1C conditions indicate.

[steveting99]

Battery Design website has a great article on the difference between BYD's Blade Battery 2.0 vs. 1.0 here: https://www.batterydesign.net/byd-blade-2-0-compared-to-1-0/

Summary table is below:

 

[steveting99]

Was curious on the data reported by Recurrent Auto regarding EV battery replacement. Most EV owners would be worried on how long their EV high voltage battery pack is going to last (hint: newer models will last much longer with LFP chemistry) and what are the risks of it failing? Below is a table given by InsideEV's Tim Levin based on the +30,000 EVs that are in Recurrent Auto's database that stretch back to 2011. 

Whilst Recurrent Auto's EV database are heavily skewed to the north American market, can see that the latest 3rd generation battery packs (from 2022 onwards) are holding up quite well. 

Recurrent Auto's article on battery replacement risk is here: https://www.recurrentauto.com/research/how-long-do-ev-batteries-last