We’re still in the early days of modern EV infrastructure, so minor issues can lead to a full high voltage pack replacement given the lack of high voltage-trained mechanics. [Ed’s Garage] was able to source a Spark EV battery pack that had succumbed to a single bad cell and takes us along for the disassembly of the faulty module.
The Spark EV was the predecessor to the more well-known Chevy Bolt, so its nearly ten year old systems might not reflect the state-of-the-art in EV batteries, but they are certainly more modern than the battery in your great-grandmother’s Baker Electric. The Li-ion polymer pouch cells are sandwiched together with cooling and shock absorbing panels to keep the cells healthy and happy, at least in theory.
In a previous video, [Ed’s Garage] takes apart the full pack and shows how the last 2P16S module has assumed a darker color on its yellow plastic, seeming to indicate that it wasn’t receiving sufficient cooling during its life in the car. It would seem that the cooling plates inside the module weren’t quite up to the task. These cells are destined for other projects, but it doesn’t seem like this particular type of battery module would be too difficult to reassemble and put back in a car as long as you could get the right torque settings for the compression bolts.
If you’re looking for other EV teardowns, might we suggest this Tesla Model S pack or one from a passively-cooled Nissan Leaf?
Bad idea to wear ANY kind of jewelry when working with battery packs…
An with a high voltage DC device that can’t be turned of, I would really recommend isolating gloves!
He wears isolating gloves in the earlier video where he’s working with full pack voltage. Probably not a bad idea while the full module is still at ~60V though.
WRT Observer’s comment, I’m guessing that’s a silicone ring, so it won’t be conductive. Little fuzzier on the conductivity if it’s a carbide.
Maybe just your dog tags in case you get fried beyond recognition?
I’ve never owned a lithium battery that didn’t go dud after about 9-10 years. I’ve read about a LG chemicals long term battery test that lasted 13 years and the battery was still alive, but it was at extremely low drain relative to the capacity. I wonder how it is for EV batteries. Some videos on youtube show 15 year old Tesla Roadsters that get 12 miles out of 7 kWh charge (should be twice more) because the internal resistance has grown so high, but those were just slapped together out of off the shelf parts then with little consideration for longevity.
Currently have a 10 year old ev, works fine, practically no range loss. (manufacturer replaced the battery at 8 years due to a isolation fault, but that wasn’t affecting available range) :D
I wouldn’t expect any range loss either if you replace the battery before it gets old.
I’d very much hope a 2-year-old battery would be working fine, that doesn’t really prove much does it???
I have a 2011 iMiEV with 112000km on the clock. I have driven it the last 24000km in the last 5½ years.
The pack is made of 88 50Ah cells – which is advertised as 16kWh, but as the cells are only charged to 4.1V instead of the nominal 4.2V it is more like 14.2kWh original.
I get around 7km/kWh which means it has probably been charged with 16000kWh or more than 1100 full cycles – in reality it is probably more like 2500 cycles from 50%. It has almost never been fast charged.
Currently the pack is at 11-12kWh based on range – the BMS guess is 32Ah (10kWh), but that is based on an expected degradation since last calibration (full discharge/charge – I rarely do that, maybe 3 or 4 years ago).
The climate is usually -5 to +25 Celcius (extremes of -15 to +35C).
The car did complain a bit this winter at -15C when trying to use max. current (150A) and the voltage sagged quite a bit more than usual.
So this datapoint is 13 Years, 2500 50% charge/discharge cycles and capacity is now down to 80% of new.
I find that the capacity of an old battery is highly dependent on load.
I had a laptop battery that was nominally 45 Wh, tested at 43 Wh, but only gave out about 20 Wh in actual use – because the battery testing procedure just runs the computer on idle until the battery goes empty.
That’s consistent with the wear-out mechanism of lithium batteries, where the solid electrolyte interface (SEI) gets thicker and less porous until it starts blocking ion flow in and out of the electrodes. There’s charge, but it becomes unavailable unless you cycle the battery really slowly.
There’s actually two (or three) main mechanisms. One is that the lithium itself becomes unavailable due to side reactions – the other is the SEI growth.
The first is a product of time x temperature x state of charge. The second mechanism happens by cycling the battery and gets worse the older the battery is, due to the side reactions leaving non-reactive lithium hanging around, further blocking the surface of the electrodes. It’s supposed to be a self-accelerating reaction, so the battery goes dud faster and faster once it goes under 80% capacity.
As another datapoint, I have a Bioenno LiFePO4 battery for a radio go-box. Put into service 9 years ago this week, it has been gently used in benign temperatures (never frozen, never baked). It’s had roughly a hundred full charge cycle equivalents, but with several complete discharges to cutoff. Originally a 15 Ah battery, I measured it last month to have a capacity of 10 Ah. Still in daily use, but I do expect its days are numbered.
2014 Fiat 500E with 106k mikes on the clock here. Still has about half charge; the cells are getting a bit wonky but it’s still functional.
Unfortunately new battery rebuilt battery packs cost as much as the car did (7k USD) if I provide the labor.
My 2013 Nissan Leaf with 113k on the clock has about 75% of original capacity. I’ll try to keep it long enough to see what happens but I am tempted to replace it with something ~30kwh usable to fit my needs better.
Interesting to see this aftermarket effort of some technicians, like happens with any mass adopted technology whose parts will eventually come to an end of life.
Overseas fix-it.
https://arstechnica.com/cars/2023/11/the-reincarnation-of-totalled-teslas-in-ukraine/
And use your Spotify acount.
Speaking of “sparky” how well do battery packs do in flooded cars?
Ever have an engine hydrolock? Better than that I would imagine!
Donald Trump just went off on a rant about electric boats sinking or something like that. Then he went off on a ramble about preferring electrocution to being eaten by a shark. I’m not sure that would happen to a passenger, I mean what are you going to do swim down between the terminals? How likely is one to become the path of least resistance?
But anyway.. that’s totally what we need a presidential candidate to be thinking and talking about, no? Electrocution vs Sharks? Reminds me of Herschel Walker’s Vampire / Warewolf thing.
I’m not at all concerned about a politician/jester who mocks EVs with electrocution/sharks humor, but advocates a policy position where the public can buy/drive what they want.
I’m far more worried about the “rants” of politicians who think that EVs are “zero emission,” that they will “save the earth,” and who want to force exclusive adoption of a transportation mode for which we have neither the the electrical infrastructure, nor raw materials, to support.
Clowns of the first type may be annoying. Clowns of the 2nd type will ruin us, irrespective of their party affiliation.
When you submerge a battery in salt water, the current takes all possible paths around the battery to get from one terminal to the other, not just the shortest one. Granted, the current at a distance is greatly diminished, and it’s DC so it won’t kill you easily, but I would still stay away from it.
Without doing the calculations, I would guess that having 400 Volts DC bubbling away under water could still pass lethal currents at something like a meter away from the battery terminals.
A quick resistor grid simulation suggests that you should be able to light a LED at 1 meter away from the discharging battery under water. Don’t know how accurate that is. It’s probably not lethal, but the current density increases rapidly if you get any closer.
Hey at least he can string a couple sentences together unlike the current guy
My vague “political” comment was deleted by a moderator a few days ago.
Why is yours still up?
Ever have a hydrolocked engine? I’m sure they do at least better than that! Most water isn’t even all that conductive.
Dernald Drumpf recently went on a rant regarding a similar topic. I think it was something about being in a sinking electric boat. Then he drabbled on about preferring electrocution to being eaten by a shark. How is that supposed to happen? Would the covers all pop off and one go swim between the terminals in their attempt to “escape”? How likely is one to end up within the path of least resistance anyway? This is exactly what we need to hear leaders talking about, no? It reminds me of Herschel Walker’s Zombie vs Warewolf thing. Pretty dumb!
No better or worse than regular cars – EV batteries & drive units tend to be sealed and have coolant running through them anyway, so it’s mostly down to the sealing of the connectors which is pretty good across the board these days.
I’m kind of sad to see the Nickel-Zinc battery didn’t get traction, and just faded away. A variant of the NiMH, it has higher terminal voltage and better output current (lower impedance), and more energy per cell. I bought a bunch of them when they hit the consumer market about a decade+ ago. They were great, except for lifetime: they all died within a few years. I have to wonder if that would have been resolved if the rechargeable lithium cells hadn’t obliterated their market.
My NiZn had an issue where they unfortunately discharged durimg use to 0v and died. Seemed like the technology may just not have worked.
Yep. They also had high self-discharge rate, a somewhat poor efficiency, and a tendency to generate hydrogen gas. There were some experimental car models that used them, and burned down because the hydrogen kept pooling in the battery box.
“…it doesn’t seem like this particular type of battery module would be too difficult to reassemble and put back in a car as long as you could get the right torque settings for the compression bolts”.
I’ve never tried the following procedure, so caveat emptor. I would start by making a drawing showing every bolt. Then I’d use a fine-point Sharpie to mark the position of one face of each bolt on the surface that it’s holding in place. Following that I would crack each bolt free, then return it to the original position indicated by the sharpie lines.
The next step would be to remove each bolt with a torque wrench, making notes on the drawing of the torque values required to break them free. This should give an indication of the torque pattern – if there is any – of the bolts. Then I would do appropriate averaging of the torque values, leaving out any apparent outliers.
During reassembly I’d use the calculated torque value(s), but might add a 5 to 10 percent fudge factor to account for stretching / deformation introduced during the removal and reassembly process.
Then I’d cross my fingers and hope that what I had done was reasonable, and resulted in a safe battery pack…
The torque to tension factor is highly fudged by the mating surfaces under the bolt/nut head, because the friction by dirt or surface deformation, or grease and oil etc. changes the torque.
That’s why you sometimes see bolts intentionally torqued to the point of yielding, because at that point the stress-strain curve goes flat at a certain value. When you release and re-tighten them, they’re work-hardened and the resulting compression is greater, so you have to replace them with fresh bolts every time.
Good information to have – thanks! When you mentioned work-hardening I had one of those ‘Doh! I should have thought of that!’ moments.
Short of having a factory spec, is there any somewhat reliable way to figure out an appropriate amount of torque?
Nice! Battery hacker’s heaven. Seems there are 15 series cells. If this block was to be used for a 48V application, it would require 13S (typical 54V) or 14S (Telecom going to 58V) for float. Once the weak cell is identified, it could be discharged manually to 0V, then shorted to leave the remaining good cells in the string functional and intact in their nice case. Could this be viable? Or would the 0V cell expand and create pressure on the adjacent cells?
Wear SAFTEY GLASSES!!!
Of note is the complete lack of support by GM to Spark EV owners.
At the young age of 60000 km, my 2016 is on its way to becoming a large paperweight.
The Spark EVs are quite different from the gasoline Sparks and have specific parts: not just the battery and the internal chargers, but even parts like the radio antenna or even the shock absorbers that are beefed up to account for the heavier weight of the battery.
GM – without fanfare, obviously – took the decision to stop manufacturing those parts, and they are out of stock everywhere. They even publicly denied the move, but the situation is very real and can be verified by searching parts resellers online databases.
Contrarily to popular belief, and at least where I live (Canada), there is no obligation for the car manufacturer to produce parts for any length of time except some vaguely worded “useful life”. Those industrial behemoths do exactly what they want if they think they can get away with it.
Welcome to the world of disposable EVs…
I purchased a full module off of the guy in the Ed’s Garage video up above. Nice guy to interact with. It was the las full module the guy had. I initially had messaged him looking for 20 cells. His response was, “Unfortunately, I only have 4 cells left!” My disappointment was immediate as I had already done my research on the LG cells and not only were they perfect for my intended use but also, they were a great deal in my opinion @ $5.00CAD per cell and he lived just a 15 minute drive from my house! After a brief moment, Ed sent another message which stated, “I actually have a full module left and its my last one” and offered it to me for $150CAD. I was all over that so I drove down to his place and picked it up without delay!
I use the module in a 13S2P configuration to power a duo of 1800w boost converters connected in parallel and onwards the boost converters are connected to a 1054W Raycus diode pump source. Secondly, I use the last end 3 parallel sets of cells in a 3S2P configuration to power the auxiliary circuit comprised of the microcontroller, camera guided aiming system, status LEDs, sensors, gauges and fans ect. I just left the module as you see it assembled in Ed's video as I have no reason at this point in time to disassemble it. I just soldered a 6 gauge jumper wire with a ring terminal on the terminating end onto the 3rd copper bridge plate from one end of the module essentially "tapping" into the current path 3 series cells in. I bolted the ring terminal onto a pre-existing bolt that is plastic welded to the modules chassis that is used to secure the top protective plastic cover in place. This works great! The cover hold down bolt is the same size as the positive and negative terminal bolts in practically the same respective location in its own corner of the module, so not only does it look like it was meant to be, which it wasn't, I made it meant to be for me and it makes it easy for hooking up and unhooking the battery module when I need to move it around as my system is still under development.I couldn't be any happier with this module! I haven't tested the individual sets of parallel cells so I don't know first hand the exact capacity of the module as it sits today however, Ed did a capacity test on it already and the cells checked out fine. There is a drop in capacity from factory new obviously however, the drop is consistent for a lithium-ion polymer battery of its age. These little (quite large) LG cells are great for what I use them for and they could be used in all kinds of DIY projects! All I know for now is that they can power my kilowatt laser pump system for several hours of intermittent use before reaching minimum discharge voltage, needing to be recharged. Each parallel sets of cells all stay relatively close to one another in their respective voltage levels as well so this is a good indication that there's no drastic degradation or internal damage to any of the cells. I just keep my eyes on it as one should with these lithium batteries when charging and discharging it to mitigate a potential catastrophic failure which if there ever was such an event, it would likely be due to user error I'm sure.
I picked up this module from Ed back in December 2025 I think it was so that's almost 4 months ago and I've had no problems with is since. Like I said, I couldn't be any happier with it. The simple facts that these are not "aliexpress" cells that come with the risk of unknown quality, potentially exaggerated ratings, carry the risk of burning your house down and even carry the bizarre risk of buying cells full of sand (sad but true) in my opinion makes these LG made cells that have proven reliability a way better value even if they are used and suffer from age degradation! When I look at it in that perspective, factoring in and mitigating the risks, $150CAD for 32 of these "used but tested good" LG made cells is crazy cheap! The $$$ per A/h value here in this case is extremely good and I'm no battery expert by far! I'll probably never find another local deal like this ever again. It's likely one of my best overall value buys in over a decade! Thanks, Ed!