Is This LG Battery Pack Problem an Achilles Hill of EVs ?

[SageBrush]

My Chevy Bolt Battery Died Twice At Under 12,000 Miles: Here's Why

Brad Berman tells the story of his Chevy Bolt suffering two battery pack failures within ~ two years. I presume in part due to his job as an EV auto journalist, GM was (relatively) attentive to his problems and even went so far as to let Brad interview a GM engineer to ask some questions.

The answers are general and political, and since my understanding is limited anyway the best I could take away is that LG allowed a statistically low number of cells out that failed prematurely. The interesting part of the story for me was learning that one poorly performing cell would either kill or severely degrade the performance of the entire pack.

My question: Would a Tesla pack behave the same if it develops one bad cell ? I'm pretty sure this situation also applies to the LEAF. I'm not worried about my car, I'm more concerned (if universal) that it would be a barrier to the successful transition to inexpensive EVs

 

[mswlogo]

No. The Model 3 batteries will remove the bad cell with fusible links. A bad cell takes more current to charge and blows the fuse.
The pack can withstand a certain amount of failed cells. And probably a certain amount of cells not to close to one another. This is what Tesla is good at.

Not sure what happens with marginal cells.

I'm sure someone way more technically knowledgeable will provide more detail. There are some nice youtube videos on the topic.

This is what you pay for with Tesla.

 

[Cosmacelf]

I don’t think a fusable link on a Tesla cell could be used for when a cell has severe capacity degradation, which is what appears to have happened with the LG cells. Nonetheless, the module and pack architecture in a Tesla car is quite different from a Bolt and I believe the pack can withstand several individual cell capacity losses without much detrimental effect.

Before I get into the differences between the Bolt and Tesla cars, I’d like to highlight how useful Tesla’s over the air software updates are. Note that Bolts literally got stranded on the highway due to a cell failure. In hindsight, GM realized that they didn’t need to shut down the car when a cell dies, they could instead degrade gracefully and allow the driver to continue driving, albeit with warnings. But they had to recall cars to uodate their software. Maybe Tesla had a similar issue, but Tesla is able to send out a “hot fix” to their fleet before an identified issue can impact many people. This is a huge difference, and it is rather amazing that no other vehicle manufacturer does this even now,

Bolts use a smaller number of bigger cells, while Tesla infamously uses a large number of small cells. It’s about a 10x size difference, it isn’t small (the Bolt uses 288 cells versus about 3,000 for the mid range Model 3). Balancing a pack is much easier with smaller cells, and cell to cell variances don't matter as much with small cells. In addition, a small Tesla cell undergoing a catastrophic cell thermal runaway is easily contained, but a 10x sized Bolt cell doing the same thing is a much bigger deal.

The bottom line is that there are very good reasons why Tesla chose their battery architecture, and it is a bit mystifying why all other EV manufacturers (except Rivian) stay with, IMHO, an inferior battery pack architecture.

 

[ReddyLeaf]

I think that my 2011 Leaf has 192 cells. there have definitely been reports of “turtle mode” showing up when one of the cell voltages got too low. The low battery warnings are triggered by the low voltage, so if one cell goes bad, the car will shutdown prematurely. This problem is exacerbated by the the lower mileage available in the Leaf. This isn’t my expertise, so don’t trust my understanding. I did personally experience a strange turtle like behavior when my battery was extremely cold, about 5-10 F, and at a low SOC (30% or so). I had power limited (25 mph turtle mode), even though the battery was way above the Turtle threshold. I trickle charged to 100% overnight and all was fine the next day. However, it’s a good reminder to not trust the lower SOC, especially when the battery is very cold. Tesla provides an additional warning that the battery will contain less useable energy if the battery cools off.

 

[SageBrush]

Nonetheless, the module and pack architecture in a Tesla car is quite different from a Bolt and I believe the pack can withstand several individual cell capacity losses without much detrimental effect.

Am I am reading you correctly that the same effect exists in a Tesla battery but the effect of a bad cell is diluted ?
I'm having trouble understanding why since (I think) that all Li-ion cells have about the same voltage but larger cells have more capacity. I infer that all current 360v nominal car packs have about 100 cells in series but the Tesla pack has ~ 10x more parallel groups.

I can understand how safety in cell numbers would protect power output of the pack but I don't understand how voltage is protected.

 

[SageBrush]

I found a Quora discussion of the pros and cons of more or fewer cells in a pack and read this interesting answer:

Frank Schmitt, B.S. in Engineering, Post-bacc in Vehicle Design
Answered Jan 21, 2014

Another advantage of smaller cells is that manufacturing defects can generally be detected via non-invasive (electrical) testing. In larger cells the types of small defects that can lead to a shortened lifetime are too small (relative to the cell's capacity) to be reliably detected.

In the same discussion JBS was quoted saying that the main advantage of small cells was thermal control.

 

[wycolo]

'Granularity' might be a good term to denote a multitude of smaller cells vs fewer large cells. More granularity seems to imply that most control and monitoring tasks will be easier to perform or more effective. Having said that I must admit I have no idea how my 2014 Spark A123 cells (unique to that year) are configured, yet they continue to soldier on after 5 years.
--

 

[ggr]

I think that my 2011 Leaf has 192 cells. there have definitely been reports of “turtle mode” showing up when one of the cell voltages got too low. The low battery warnings are triggered by the low voltage, so if one cell goes bad, the car will shutdown prematurely. This problem is exacerbated by the the lower mileage available in the Leaf. This isn’t my expertise, so don’t trust my understanding. I did personally experience a strange turtle like behavior when my battery was extremely cold, about 5-10 F, and at a low SOC (30% or so). I had power limited (25 mph turtle mode), even though the battery was way above the Turtle threshold. I trickle charged to 100% overnight and all was fine the next day. However, it’s a good reminder to not trust the lower SOC, especially when the battery is very cold. Tesla provides an additional warning that the battery will contain less useable energy if the battery cools off.

Why is it "turtle mode", when clearly it should be "tortoise"?