Hyundai Kona Forum banner

301 - 320 of 353 Posts

·
Registered
Joined
·
317 Posts
The truth is, we do not have a detailed, easy to understand explanation form Hyundai what is the exact failure mechanism.

So the big question is, is the "fold" internal to the battery pouch, like it is shown on the x-section picture? In this case the "tab" is possibly a mistranslation of "stripe", and the fold is inherent in all batteries made in the implicated time frame. We have some Korean speaking members on the board, perhaps they could help out clarifying this.
Pictures illustrating the alternative theory of the "folded tab" would be a tremendous help. Before we try to understand the problem based on semantics, it would be good to know if we are dealing with a mistranslation.

OTOH if the "fold" is something external, and the "tab" description is correct, I expect that could be seen by inspecting the exterior of the cells, and perhaps a kapton tape hack like Kiwi suggests would be possible. Given the significant cost of replacing the entire battery pack, I think, if there was some way to repair these packs in the field, I would expect Hyundai to pursue that. There is a huge financial incentive to do so. But they are not doing that. Which leads me to believe that the flaw is internal to the pouch, and no field repair is possible.

Note that the problem develops over time and the onset can be pushed back by practicing conservative charge management. So, to me this also speaks to the viability of "sharp anode end punches through the separator layer, due to battery pouch thermal movement". Once there is a hole in the separator layer, the path is open to crystalline dendrite formation progressing between the anode and cathode and the battery starts gradually leaking current internally, This would explain how the condition develops gradually, and how the BMS can detect it in time to prevent a fire.

With the other theory ( "an external anode tab is accidentally folded, and this leads to the anode tab and cathode tab touching") I would expect that condition to be binary, and sudden onset. I am struggling with how the BMS could catch and prevent that condition from "getting worse". It is "touching" or "not touching", and I fail to see anything in-between
that could be detected as an early warning.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
Be warned, following is a heap of speculation:
There may be an interim condition where the fault resistance causes some current to flow but not enough heat is generated to make the condition worse, i.e. thermal runaway.

You could consider what the BMS might be looking at the pause in charging at 82%, a voltage drop in any of the (98) 3-cell parallel assemblies, of 0.02 V in 10 minutes. The capacity of three cells in total is 180 Ah over a range of 3.50 to 4.12 OCV (approximately) and you'd initially assume a linear energy/voltage drop relationship even it isn't.

Factor the change in voltage we're looking for against capacity: 0.02 V x (4.12 V - 3.50 V) x 180 Ah = 2.23 Ah. The equivalent current for only 10 min instead of one hour is 2.23 Ah x 6 = 13.4 A. The nominal voltage is about 4.10 V so energy dissipated would be 4.10 x 13.4 = 55 watts. That's the heat generated at the point where the short had occurred in one cell if it were severe enough to drop the voltage of all three cells by 0.02 V in 10 minutes.

By rough calcs that amount of heat produced is much more than would be produced from normal resistive losses in the entire body of each cell at the highest level of acceleration or charging. But the four temperature sensors each cover a thermal mass of about 294 / 4 = 73.5 cells so the temperature rise may not even register over that 10 minutes.

Once charging has finished and if the car is left parked the BMS appears to take measurements every 30 or 60 minutes for 4 to 6 hours, judging from 12V system activity. I think that's enough time to detect the same 0.02 open-circuit voltage change earlier in the game, at less-risky energy dissipation levels. That could justify the idea of limiting charge level to 80% instead of the 90% I've been recommended by my importer, and leaving the car parked for a few hours after charging so that it can monitor its condition.
 

·
Registered
Joined
·
18 Posts
Be warned, following is a heap of speculation:
There may be an interim condition where the fault resistance causes some current to flow but not enough heat is generated to make the condition worse, i.e. thermal runaway.

You could consider what the BMS might be looking at the pause in charging at 82%, a voltage drop in any of the (98) 3-cell parallel assemblies, of 0.02 V in 10 minutes. The capacity of three cells in total is 180 Ah over a range of 3.50 to 4.12 OCV (approximately) and you'd initially assume a linear energy/voltage drop relationship even it isn't.

Factor the change in voltage we're looking for against capacity: 0.02 V x (4.12 V - 3.50 V) x 180 Ah = 2.23 Ah. The equivalent current for only 10 min instead of one hour is 2.23 Ah x 6 = 13.4 A. The nominal voltage is about 4.10 V so energy dissipated would be 4.10 x 13.4 = 55 watts. That's the heat generated at the point where the short had occurred in one cell if it were severe enough to drop the voltage of all three cells by 0.02 V in 10 minutes.

By rough calcs that amount of heat produced is much more than would be produced from normal resistive losses in the entire body of each cell at the highest level of acceleration or charging. But the four temperature sensors each cover a thermal mass of about 294 / 4 = 73.5 cells so the temperature rise may not even register over that 10 minutes.

Once charging has finished and if the car is left parked the BMS appears to take measurements every 30 or 60 minutes for 4 to 6 hours, judging from 12V system activity. I think that's enough time to detect the same 0.02 open-circuit voltage change earlier in the game, at less-risky energy dissipation levels. That could justify the idea of limiting charge level to 80% instead of the 90% I've been recommended by my importer, and leaving the car parked for a few hours after charging so that it can monitor its condition.
Good speculations and details. When you say "...parked for a few hours after charging so that it can monitor its condition" do you mean with the system turned OFF, or while it is in Utility Mode?
 

·
Registered
Joined
·
317 Posts
Ah something interesting: Reading the description in the recall note, I think I find a clue:

"A folded Anode tab in the battery cell could allow the lithium plating on the Anode tab to contact the Cathode resulting in an electrical short."

If we were talking about an external "tab", used to connect the battery pouch to something else, that surely would not be lithium plated. That is likely a plain copper tab.

Only the inside part of the anode strip, (enclosed inside the pouch, and sealed in by the separator layers) is "lithium plated". On one of the pictures, a microscope photo x-section shot, there is a layer shown labelled "anode active material" Afaik, that is the lithium plating they are talking about. The "lithium plating" is the layer that actually stores the electrons. Lithium plating is a somewhat unscientific way to reference the "active layer" which is some lithium compound,
depending on the exact flavor of battery chemistry used.

I have seen a post here: Hyundai profit hit after Kona EV recall, but LG Chem seen bearing bulk of costs

The article mentions how LG and Hyundai revised their 4th quarter earnings predictions accounting for the expected cost of the battery replacement campaign. Given that they have just closed the 1st quarter, if I am guessing correctly, nothing major is expected to happen in the next 2 quarters (8 months) and I expect the bulk of the replacements taking place
in the 4th quarter. That is my best guess based on the news coverage.

When an international corporation announces earnings guidance, that is usually a close match with their intentions.
But who knows, I could be totally wrong.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
Yes, I just noticed the same text within the last hour. I must be skimming over things too quickly.

... When you say "...parked for a few hours after charging so that it can monitor its condition" do you mean with the system turned OFF, or while it is in Utility Mode?
Turned off so that the traction battery is disconnected and the BMS relies on the 12V ("aux") battery to wake-up and operate for a moment. You can see this on BM-2 logs, see image below which I annotated a few months back. To derive anything useful from "cell" voltages the pack must be allowed to settle undisturbed.
I'm unsure what would happen if the traction battery was in-use, such as driving.

5259
 

·
Registered
Joined
·
18 Posts
Ah something interesting: Reading the description in the recall note, I think I find a clue:

"A folded Anode tab in the battery cell could allow the lithium plating on the Anode tab to contact the Cathode resulting in an electrical short."

If we were talking about an external "tab", used to connect the battery pouch to something else, that surely would not be lithium plated. That is likely a plain copper tab.

Only the inside part of the anode strip, (enclosed inside the pouch, and sealed in by the separator layers) is "lithium plated". On one of the pictures, a microscope photo x-section shot, there is a layer shown labelled "anode active material" Afaik, that is the lithium plating they are talking about. The "lithium plating" is the layer that actually stores the electrons. Lithium plating is a somewhat unscientific way to reference the "active layer" which is some lithium compound,
depending on the exact flavor of battery chemistry used.

I have seen a post here: Hyundai profit hit after Kona EV recall, but LG Chem seen bearing bulk of costs

The article mentions how LG and Hyundai revised their 4th quarter earnings predictions accounting for the expected cost of the battery replacement campaign. Given that they have just closed the 1st quarter, if I am guessing correctly, nothing major is expected to happen in the next 2 quarters (8 months) and I expect the bulk of the replacements taking place
in the 4th quarter. That is my best guess based on the news coverage.

When an international corporation announces earnings guidance, that is usually a close match with their intentions.
But who knows, I could be totally wrong.
After reading the linked article I have some other speculations. It said that 82,000 Kona EVs were affected by Recall 196 and 72,000 are affected by Recall 200. Perhaps that means some of the battery diagnostics led to battery replacements including the upgraded battery monitoring system that most of the recalled vehicles likely now have. That could mean that there is a lower probability that the issue exists within the remaining 72,000 vehicles. The article mentioned at the end that the first recall involved an estimated cost of $900 million USD. My guess is that the second recall will eventually involve costs much greater than the first.
 

·
Registered
Joined
·
317 Posts
Keep in mind that we are looking at various news feeds, the journalists do not necessarily have in-depth understanding of the topic. At the time recall 196 was done, there was no clear understanding of the root cause yet, so the software update was rolled out to more cars than the battery replacement will be.

Some, but nowhere near 10K batteries were replaced already. 10K replaced batteries would have shown up in Hyundai-s and LG-s balance sheets as a noticeable write-offs, on the order of ( total) $130Billion. That kind of expense/write-off does not happen for a publicly traded company without a trace.

The article mentions the #192 recall cost. The $900B number is the estimated total cost of replacing all the impacted batteries. The article quotes the money already spent as "38.9 billion Won" . You have to divide that by ~1000 so you end up with ~$39M. The cost sharing agreement between LG and Hyundai has not been reached until fairly recently. At the time campaign #196 was rolled out, LG and Hyundai were still in disagreement as to what the root cause was. (if memory serves, at that time the prevailing theory was "dendrites").
As of now, LG is supposed to pick up 62 to 70% of the replacement cost, depending who is reporting.

There were 2 more battery fires after the 196 recall was rolled out. From the 1st one, the battery was too damaged to yield any clues, but the 2nd yielded the "folded anode" theory. One interesting (and a bit disturbing ) note:
These 2 cars both had the #192 software update. So, how reliable is the software update in detecting the impending doom? If the #192 software update worked as advertised, why did it not detect the whatever problems indicating impending battery failure and shut down the car, like we see it reported by some of our users on this board? So that kinds of contradicts the current narrative.
 

·
Registered
Joined
·
317 Posts
And also $130M not $130B. I arrived at that number by dividing the total estimated cost by 7.
If 70K batteries cost 900M to replace, then 10K batteries cost 1/7 of that. Just some "gross order estimate" math.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
... These 2 cars both had the #192 software update. So, how reliable is the software update in detecting the impending doom? If the #192 software update worked as advertised, why did it not detect the whatever problems indicating impending battery failure and shut down the car, like we see it reported by some of our users on this board? ...
I may have not summarised it very well, or at all, but that was the point of my 'heap of speculation' above. During charging, the pause at 82% would not, if my assumptions are correct, allow the BMS to detect the onset of a problem until 55 watts has been dissipated as heat. Since the partial short would be in small area, that's unlikely to be a recoverable situation, IMO.
 

·
Registered
Joined
·
18 Posts
I may have not summarised it very well, or at all, but that was the point of my 'heap of speculation' above. During charging, the pause at 82% would not, if my assumptions are correct, allow the BMS to detect the onset of a problem until 55 watts has been dissipated as heat. Since the partial short would be in small area, that's unlikely to be a recoverable situation, IMO.
I agree. Even if only one watt worth of heating occurred at the microscopic tip of a puncturing anode tip, or lithium dendrite, a runaway consuming event can take place. It is somewhat analogous to the threat, in an explosive atmosphere, of a microscopic discharge of some static electricity.
 

·
Registered
Joined
·
317 Posts
Just to spell out what I think the current narrative is: "As long you limit your charging to 80% you are safe from battery fires". In fact the #200 recall notice explicitly states that if you limit your charging, to 80% max, it is unnecessary to park your vehicle away from combustibles, or outside of your garage. This sounds to me like Hyundai did chose their wording carefully to
disclaim liability in case there is a fire and damage, despite of the recommended precautions, not that I am complaining.
 

·
Registered
Joined
·
18 Posts
Just to spell out what I think the current narrative is: "As long you limit your charging to 80% you are safe from battery fires". In fact the #200 recall notice explicitly states that if you limit your charging, to 80% max, it is unnecessary to park your vehicle away from combustibles, or outside of your garage. This sounds to me like Hyundai did chose their wording carefully to
disclaim liability in case there is a fire and damage, despite of the recommended precautions, not that I am complaining.
For me this 80% maximum limitation will only create some inconvenience when on extended remote driving trips which are rare.
 

·
Registered
Joined
·
317 Posts
Keep in mind the winter. When it is verry cold, it is surprising how much the range can drop. Of course your 2020 has a heat pump, which is more efficient than the resistive heater my 2019 has. If you are still ok, then indeed nothing to worry about. Well there is one thing... If your BMS decides at some date that your battery is now unsafe, then you need to deal with the long battery replacement wait and a long-term replacement vehicle.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
However, anecdotal evidence (from FB mostly) is that those who suddenly find themselves with undriveable Konas are in line for replacements far sooner than the rest of us. A few have been offered buybacks, mostly in the US of course.
 

·
Registered
Joined
·
28 Posts
"...the #200 recall notice explicitly states that if you limit your charging, to 80% max, it is unnecessary to park your vehicle away from combustibles, or outside of your garage."
fyi. My Southern California campaign 200 recall does not explicitly state it, but it does imply it.
 

·
Registered
Joined
·
11 Posts
However, anecdotal evidence (from FB mostly) is that those who suddenly find themselves with undriveable Konas are in line for replacements far sooner than the rest of us. A few have been offered buybacks, mostly in the US of course.
Just got our Kona back today. 80 days out of action, hire car provided, looking for petrol reimbursement too. While not ideal by any means Hyundai New Zealand have helped make sure this is as good as a bad situation can be.
Have been told treat the new battery as we did previously and definitely OK to periodically charge to 100%.
Early days but fingers crossed that all goes well from here on in.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
For those of us whose cars stay 80 or 90% operational, the down time for battery replacement should be under a day, but I expect I'll be waiting until next year for that to happen. As with most things, the squeaky wheel gets the grease.
 

·
Registered
9/18 EU base model
Joined
·
209 Posts
Hey peoples, did we see this video? It's partly UK-centric but the VINs posted appear to be global. My VIN is right in the thick of it, "GUK". If this info is actually true it means that the number of affected Konas is FAR smaller than we had been lead to believe. It also explains why they keep telling us they're trying to determine what examples are affected.

Kona EV (OS EV): From KMHK381GUJU000203 To KMHK581GULU072444

 
301 - 320 of 353 Posts
Top