Chevy Bolt EV Forum banner
21 - 40 of 51 Posts

·
Registered
Joined
·
1,503 Posts
So far no 2017 bolts have been recalled
Only 2018/2019
NO!!!

ALL 2017 and 2018 Bolts were recalled for this possible battery fire + a subset of 2019 with campaign id 20V701000. 2017 CHEVROLET BOLT 4 DR FWD | NHTSA says it right there:
"General Motors LLC (GM) is recalling all 2017-2018 and certain 2019 Chevrolet Bolt EV vehicles. The high voltage battery could catch fire when charged to full or nearly full capacity."

You can read thru the associated docs, currently at 35 like the FAQs and letters to NHTSA like https://static.nhtsa.gov/odi/rcl/2020/RCAK-20V701-4648.pdf.

You can input the vehicle and model year at Recalls | NHTSA to see if there are recalls for given model year and model of automobile.

Umm... as for CA having most buybacks or whatever, maybe... but then again AFAIK, CA accounts for the largest % of EV/PHEV purchases of any state in the US and CA is also the most populous state.
 

·
Super Moderator
2020 Chevrolet Bolt
Joined
·
1,200 Posts
I would just note that the article mentions LG denies that as the cause, and elsewhere LG points to Hyundai's BMS:
 

·
Registered
Joined
·
1,958 Posts
Hey @Stv, you asked:

I am unable to read through 30+ pages of posts about this, so I apologize in advance if this has already been asked and answered:

Do the fires have any correlation with the number of charge cycles on the battery or number of miles on the vehicles that burned?
This is about the best summary of what we know:

Seems like part of the problem is that while some cars appear to be higher risk:
  • 2019 with Korea-made battery (versus 2017-2018).
  • Charged to full and/or parked at high state of charge for a while.
  • Frequently deeply discharged.
  • Has a cell with low voltage (probably based on one of the fires where the owner was complaining about reduced range before it caught fire).
there is not enough evidence to rule out fire risk in other cars:
  • Some 2017-2018 cars have caught fire, though at a far lower rate than 2019 with Korea-made battery cars.
  • While the rate of fires dropped greatly during the interim fix period (94% software limit or advisory to use target charge <= 90% or hilltop reserve), there was one (maybe two) fire in a car with the interim fix.
  • While frequent deep discharges appeared common among fire cars, it is unknown (to the general public) if any fires occurred in cars that were not frequently deep discharged.
  • Probably most low voltage cells are not related to the fire defect (since reports of low voltage cells seem to be a lot more common than fires), and detection of low voltage cells does not reliably prevent all fires (as indicated by two fires with the "final" fix), even if the fire defect sometimes reveals itself with a low voltage cell before a fire.
In other words, the population of cars at risk is still "all 2017-2019 with Korea-made battery", even though the risk levels apparently vary (to an unknown extent) within that population. The "final" fix was presumably intended to find the at-risk X% (X < 100) of recalled cars with defects before they caught fire so that they could be repaired; obviously, it did not work in at least two cases. So the at-risk percentage X% is still 100% as far as we (and probably GM) know.
We don't know the mileage on all the vehicles, but we do know that one of the 2017 models that burned was really high mileage (likely used for Uber, etc.) and given that many of the burned vehicles were regularly charged to full and deeply discharged, it seems likely they were being driven a lot.

I've shown that it's statistically unlikely that it's just a fluke that (Korean battery) 2019 models show more fires (even without factoring mileage or number of charges), but given that the 2019 models have been on the road the shortest time, their collective lower mileage isn't mitigating risk—if the 2017 models had the same rate of fires as the 2019s, there would have been 21 fires in 2017 models by July, 2019, and there had been no fires at all at that point. Thus there is definitely a defect introduced into batteries produced for the 2019 model year in Korea that radically increases risk.

We also know that all fires so far appear to have happened in the late stages of charging or shortly after charging. From that I conclude that if you don't charge every day, on the days you're not charging there appears to be much less risk. I'd also recommend using time-of-departure charging to finish charging just before you drive somewhere.

As a result, I've switched from charging every day to waiting until my battery hits 50% (and charging to hill-top reserve level, 88%), which is several days given the minimal amount I'm driving at the moment (still mostly working from home).
 

·
Super Moderator
2020 Chevrolet Bolt
Joined
·
1,200 Posts
Does anyone know how many physical cells are connected together before connected to the BMS?
There are 288 individual cells, connected 3 at a time in parallel, for 96 "books" connected in series, in 10 different modules. Some modules have 10 books, other modules have 8.

The BMS is able to monitor cell voltage at the "book" level. There are 6 battery pack temperature sensors.
 

·
Registered
Joined
·
969 Posts
I've shown that it's statistically unlikely that it's just a fluke that (Korean battery) 2019 models show more fires (even without factoring mileage or number of charges), but given that the 2019 models have been on the road the shortest time, their collective lower mileage isn't mitigating risk—if the 2017 models had the same rate of fires as the 2019s, there would have been 21 fires in 2017 models by July, 2019, and there had been no fires at all at that point. Thus there is definitely a defect introduced into batteries produced for the 2019 model year in Korea that radically increases risk.
The change for 2019 could possibly be related to the unannounced (but mentioned in a service manual or something like that) battery capacity increase for 2019.
 

·
Registered
Joined
·
143 Posts
could also be to reduce your charging rate to take it easy on the battery, thus giving you more time to enjoy your longer lunch when traveling and recharging. I can speculate all day long.
 

·
Registered
Joined
·
143 Posts
I get the arrangement but I don't get how many BMS wires there are to balance the 288 cells. Regardless the software can't measure individual cell voltage. So what keeps a single cell in a pouch or module from degrading? How can the BMS detected a bad cell that's wired in parallel with others? Seems to me (am not a battery engineer and barely understand LIPO technology) that the good batteries can pull-up the bad battery to the point where it can continue to degrade until it affects the others and detected by BMS. but it may be too late by then. What am I missing here? Shouldn't there be individual cell fusing or monitoring???
 

·
Registered
Joined
·
635 Posts
So far no 2017 bolts have been recalled
Only 2018/2019, and no buybacks offered
If your state has a strict lemon law GM might deal with you
So far California has the most buybacks
Since April this year
Your information is not correct. All 2017s were part of the recall and there have absolutely been 2017s that have been bought back/msrp swap.

The recall affected all 2017s, all 2018s and some of 2019s. In 2019 they switched from the batteries supplied by the factory in Ochang, South Korea to US manufactured batteries. The US battery packs are not part of the recall, just the battery packs from the Ochang factory.
 

·
Registered
Joined
·
285 Posts
I get the arrangement but I don't get how many BMS wires there are to balance the 288 cells. Regardless the software can't measure individual cell voltage. So what keeps a single cell in a pouch or module from degrading? How can the BMS detected a bad cell that's wired in parallel with others? Seems to me (am not a battery engineer and barely understand LIPO technology) that the good batteries can pull-up the bad battery to the point where it can continue to degrade until it affects the others and detected by BMS. but it may be too late by then. What am I missing here? Shouldn't there be individual cell fusing or monitoring???
All the cells in the same cell group are fused together. There's no way for them to individually monitor them, there are 96 datapoints, not 288. A bad or weak cell certainly is going to end up taking even more abuse because it's forced to follow along with the good cells.
 

·
Super Moderator
2020 Chevrolet Bolt
Joined
·
1,200 Posts
I get the arrangement but I don't get how many BMS wires there are to balance the 288 cells. Regardless the software can't measure individual cell voltage. So what keeps a single cell in a pouch or module from degrading? How can the BMS detected a bad cell that's wired in parallel with others? Seems to me (am not a battery engineer and barely understand LIPO technology) that the good batteries can pull-up the bad battery to the point where it can continue to degrade until it affects the others and detected by BMS. but it may be too late by then. What am I missing here? Shouldn't there be individual cell fusing or monitoring???
In general, you're correct that having 3 cells connected in parallel (the "book") limits what the BMS can detect. It can detect variations between "books," but not between cells within a "book." However, if a "book" has a weak cell, that may show up in variations in that particular "book" voltage over time.
 

·
Registered
2021 Nightfall Gray Metallic Bolt Premier
Joined
·
520 Posts
Any chance GM has figured out a way to monitor all the cells with software?
 

·
Super Moderator
2020 Chevrolet Bolt
Joined
·
1,200 Posts
Any chance GM has figured out a way to monitor all the cells with software?
The software is limited to the existing hardware sensors. The best they can do is look for patterns among those 3-cell "books" and try to infer what's happening at the cell level.
 

·
Registered
2021 Nightfall Gray Metallic Bolt Premier
Joined
·
520 Posts
Not arguing with you, but how do we know there's not a way for software to monitor each cell?
I've seen where the battery packs have been taken apart so I'm guessing if there was a way to monitor each cell the ability would have been found then.
 

·
Registered
Joined
·
635 Posts
Not arguing with you, but how do we know there's not a way for software to monitor each cell?
I've seen where the battery packs have been taken apart so I'm guessing if there was a way to monitor each cell the ability would have been found then.
Because the sensors are at the level of seeing the 3-cells together, it can't directly measure a single cell of those three. From the sensors view it only sees a single voltage/level from those three cells. Software can only work with the available inputs, in this case a single sensor output for each group of three cells.
 

·
Registered
Joined
·
143 Posts
There's no software solution with the present battery wiring design. You need to rewire battery at minimum maybe upgrade BMS too for more sensor inputs. Thinking there might be a software fix is grasping at straws. To hope that the software will detect a faulty cell when not each cell is monitored. What a joke. No one at GM raised their hands to say "what if one cell goes bad should we monitor them all or fuse them?" Can't believe it really. Bad initial design, bad judgment thinking that software can detect bad cells when they are not monitored individually. Poor engineer that said you know if we modify the software to detected .08 variance or greater that should do it. The next software fix is "let's start flagging bad modules when the variance is .05". Oh GM. They need to replace all with new battery design that electrically monitors and isolates marginal cells before they short or smoke or do what they shouldn't do. No wonder GM says the public can't understand the details of the problem, solution so is best not to confuse us. Yeah, it's a little embarrassing isn't GM? Build an electric car with LIPO batteries and not monitor each cell. ridiculous.
 

·
Registered
Joined
·
635 Posts
The way the batteries are built and where the sensors are is exactly why I was skeptical that a software update would be able to detect/stop fires. It only takes one cell to have a bad enough issue and you have a fire. They can't directly monitor each cell, just the 3 cell groups. So it is quite possible for their software to never detect an issue or only catch it once it has already reached thermal runaway and then it just sounds the horn.

Though I am curious if they made adjustments to the battery conditioning algorithms as well. Ever since the final recall even when only charging to hilltop the car does a lot of battery conditioning now. It almost never did it before the update.
 

·
Registered
Joined
·
143 Posts
From what I've read LIPO are most susceptible to heat up and do bad things at the extreme range of operation e.g. low or high cell voltage. So a quick work around is not charge fast, don't charge full and don't let it go below 20%. Ok, that all makes sense. Hopefully, is there is a bad cell going it won't explode cause I won't stress it while charging or discharging it. But that's not a real solution anymore than park your car outside is not a solution. BTW GM forgot to say don't park near trees and don't start a forest fire. Now that GM realizes (assuming the software fix was to tighten the tolerance on cell voltage variance) that the previous fix didn't work now what? Here's where the rubber hits the road. Let me see, does GM's commitment to customer satisfaction exceed their need to keep the board of directors happy? No. So what GM will do is no more than they will be force to do by the NTSA and they'll spin whatever they do as "customer satisfaction and safety is our main concern" bull. Too many run ins with big corporations, worked to many years with big corporations to start believing that anything is more important than profit. Customer satisfaction is only important because they figured out it affects sales, although when stressed they often forget.
 
21 - 40 of 51 Posts
Top