Battery Recall for cars with E400 high-voltage battery.

[kimdavi]

kimdavi : that's a really large cell deviation for a battery that is only 4 months old. (0.28 volts at 80% charge)
My car's battery was changed 4 1/4 years ago and the variance between the maximum and minimum cell voltage was 0.10 volts.
From your range data I'm guessing that your usable capacity is still 100% of new, ie no degradation.
Hence the problem is not that the cells are degraded, it is the fact that they are not all the same.

The SOH number provided by any app other than Kia's official GDS will be nonsense. We do not have any knowledge of what the latest BMS algorithm is doing.

If I were you I would try trickle charging. Charge at a lower current for longer. Circumstantial evidence suggests this balances the cells much better than a standard L2 charge.

I'll drain the battery to 1% and slow charge it to 80% over a few days, then check voltage differences and share my findings.

 

[pgeller]

Thanks for the suggestion, I will do that and retake cell voltage readings to see if there is any improvement.
My sole chance of exiting power-limp mode is to strive to balance cell voltages, ensuring they differ by no more than 0.14 volts.

I suspect that the BMS will limit your SOC to 80% because it's in a fault condition (hence the CHECK EV and Turtle indicators), but it's worth a try.

This whole episode has been tiring, but I'm thinking KIA is doing the best they can considering the variables.

pg

 

[JejuSoul]

Is it still alive?

Yes. No Limp Mode. Everything normal.

 

[pgeller]

Yes. No Limp Mode. Everything normal.

That’s great. Congratulations. Hopefully that’s a sign that other cars can be salvaged, maybe allow battery replacement going forward. Here in Portland, the Kia dealers have not been notified of a software update.

pg

 

[kimdavi]

Yes. 1.4V does not seem possible. Either the number I was told is a mistake, or the BMS creates a fake value when something is badly wrong with the battery.
The highest number I have seen on a badly degraded battery is 0.36V. While driving a bad cell can go as much as 0.36V LOWER than the rest. The car will turtle. When charging starts that bad cell will instantly switch from being lower to being HIGHER than the rest. Charging will cease when that cell is 0.36V higher than the rest.

Your observations align perfectly with my current situation. Following a full day of Level 1 charging, I conducted another Soul Spy battery cell map and discovered that certain cells, namely 54, 61, 68, 71, 79, 93, 94, 95, and 96, exhibited a higher voltage of 4.18 volts compared to the rest, which showed 3.88 volts. To further collaborate, charging process halted prematurely at 72.5% instead of reaching 80% due to .32volt difference (3.86v-4.18v).

 

[kimdavi]

I'm skeptical about the assumption that a high cell voltage difference is the sole cause of triggering BMS recall warnings. As evidence, I present a side-by-side cell voltage comparison between my original (left) battery and the replacement battery (right), both taken at a similar State of Charge (SOC) of 74%.

Despite the original battery pack exhibiting only a 0.02-volt Min-Max difference, it still triggered the BMS warning. In contrast, the recall replacement battery showed a larger difference of 0.32 volts. Both batteries indicated error code P1AD1. However, based on this comparison, it appears that the cell voltage difference may not be the sole cause of triggering the P1AD1 error.

 

[JejuSoul]

Kimdavi - thanks for adding the cell images above.
Some questions :
1. What do you mean by 'original battery pack'.
( The original pack for a 27kWh Soul is not subject to this recall because it does not contain E400 cells. )
2. How and when did you measure error code P1AD1.
3.Why do you state (SOC) of 74% ? The car scanner image you posted clearly shows SOC (display) = 100%
( We normally refer to the SOC (display) as SOC because that is the value the car shows. )
4. What do you think SOC (BMS) means? In your car SOC (BMS) varies a great deal from SOC (display). This has not been noted before.
( Maybe an explanation that doesn't make much sense to anyone but a battery engineer is :-

The Battery ECU computes SOC and Watt-hours remaining by 2 methods; One is by coulomb counting with a hall-effect ammeter, and the other is by watching open-circuit voltage of the pack and performing a complex calculation based on multiple inputs. Because the coulomb counting method has some drift due to the hall-effect system, Periodically the battery ECU will dynamically adjust these figures to keep them accurate.

The only person on this site who had a good understanding of SOC(BMS) was SoulEV2016. He stated

SOC BMS is always a formula between the max voltage and balancing stage (cell deviation) with a correction of Ah-meter.

We do know that SOC(BMS) is not a simple measure of the total capacity while SOC(display)
shows usable capacity. Despite this many people try to calculate total capacity using these two values.
5. I too doubt that a high cell voltage difference is the sole cause of triggering BMS recall warnings.
There are 11 different error codes that will lead to a battery replacement . Presumably they have different triggers.
This image is from the troubleshooting guide.

 

[kimdavi]

Kimdavi - thanks for adding the cell images above.
Some questions :
1. What do you mean by 'original battery pack'.

Original="E400 recall battery"

( The original pack for a 27kWh Soul is not subject to this recall because it does not contain E400 cells. )
2. How and when did you measure error code P1AD1.

1) After the BMS and Harness update, Kia diagnostics reported error code P1AD1 following the triggering of the warning.
2) I used a Car Scanner App to check the error code on my current battery (a replacement due to a recall triggering for the second time) and once again got the code P1AD1.

.3.Why do you state (SOC) of 74% ? The car scanner image you posted clearly shows SOC (display) = 100%

The Car Scanner App indicates a BMS SOC of 72.5%, whereas the Display SOC shows 100% (refer to the 2nd image on post 205). Initially, during full charges using level 1 charging, it would stop at 80%. Now, I've observed it stops much earlier, suggesting it's likely at 72.5% capacity based on charging duration.

4. What do you think SOC (BMS) means? In your car SOC (BMS) varies a great deal from SOC (display). This has not been noted before.

I've been noticing this has been happening recently. Not sure why they don't match. I'm annoyed that the maximum charging is now restricted to 72.5%, I can't fully charge beyond this.

( Maybe an explanation that doesn't make much sense to anyone but a battery engineer is :-

The only person on this site who had a good understanding of SOC(BMS) was SoulEV2016. He stated

We do know that SOC(BMS) is not a simple measure of the total capacity while SOC(display)
shows usable capacity. Despite this many people try to calculate total capacity using these two values.
5. I too doubt that a high cell voltage difference is the sole cause of triggering BMS recall warnings.
There are 11 different error codes that will lead to a battery replacement . Presumably they have different triggers.
This image is from the troubleshooting guide.

View attachment 101

 

[andrew1joe]

At 100% display SoC, your max cell is 4.18v and min cell is 3.86v. Low voltage cutoff is ~3.0v. That means your effective voltage range is 3.86v - 3.0v or 0.86v. A new battery should give you 4.18v - 3.0v or 1.18v range of potential. 0.86v / 1.18v = 72.9% of actual SoC vs display of SoC 100% which is where your 72.5% comes from.

All lithium BMS that's not for Lifepo4 has the following logic (Lifepo4 has different cutoff points):

Charge:
1. monitor all the cells
2. if *any* cell hits 4.2v (lower if there is top buffer), stop charging.
3. if after charging for x minutes, max cell voltage doesn't change, stop charging. Balancing happens during this time + after the min cell reaches its max voltage. This also handles the case where the entire pack gradually degrades and can never hit 4.2v.

Discharge:
1. monitor if *any* cell drops below ~3.0v (may be slightly lower or higher depending on chemistry) and shut off battery pack if any cell drops below threshold. Note that high acceleration can suddenly sag voltage on weak cells which is why cars enter limp mode. It's to protect the battery pack and give you a little more range before the BMS shuts off the battery pack.

 

[andrew1joe]

If you balanced the cells and the voltage differential doesn’t shrink, that means the degraded cells are refusing to go higher SoC/voltage and converting the extra energy into heat instead. The car also monitors if max cell’s voltage doesn’t change after charging for a while and stops charging. That’s why you see evenly degraded battery packs also stop charging at a lower voltage like 4.0v where none goes to 4.18v. Note this means that it doesn't matter if 1 cell is degraded or the whole pack is - the pack's total usable capacity is limited by the cell with the lowest voltage.

I have some loose 18650 cells I harvested over the years that no longer goes above 3.8v and just gets hotter after that point if I continue to charge it on my dumb charger.

Basically what's happening is that once anode and/or cathod degrades, it no longer has the ability to maintain 4.2v potential difference between the terminals.

 

[JejuSoul]

If you balanced the cells and the voltage differential doesn’t shrink, that means the degraded cells are refusing to go higher SoC/voltage and converting the extra energy into heat instead. ...

This has been seen in the past data. If there is one degraded cell then the module containing that cell will heat up much faster than the others.
The heat is caused by higher internal resistance in the degraded cell.

But in this case 5 of the 8 modules are much hotter than the other 3. Look at the SoulEVSpy image above. The module temps are in the top row. The voltages beneath them. The majority of the cells are at a lower voltage. These facts do not fit with our idea of degraded cells causing the heat, because the degraded cells would be the ones that got hotter. And they would be the ones at the higher charging voltages.

Also the range data suggests that this car is still a 'new' battery. It is only 4 months old after all.

My explanation for this is that this is a re-manufactured pack containing cells of very different capacities. Maybe some of the cells are new and some are old. Maybe all are new but they are of different types. No-one (including Kia) knows what is inside every re-manufactured pack.

 

[andrew1joe]

Regarding temperature, it looks like the second picture was taken well after the car finished charging (overnight?). Can you confirm kimdavi? I say that because in the second picture, the temperature is much lower than the first and the temperature difference between the modules aren’t that much different from each other after charging.

As for mixing different capacity batteries, I would be surprised if Kia engineers are that incompetent. That’s a big no no and basic knowledge for batteries of any type as the imbalance would cause the battery pack to fail much faster.

 

[IanL]

Going back to post #209, and trying to understand - what is the difference between SOC(BMS) and SOH?

 

[JejuSoul]

SOH stands for State of Health - it shows the degradation of the battery over the years.
SOC stands for State of Charge - it goes down as you drive the car, and then back up as you charge each day.

The car shows us SOC as a number between 0% and 100%
Without an app you cannot see SOH directly. You can approximate it by doing a range test.

Using an app you can see three values for SOC
SOC (display) - this is the one the car shows. It is the one we normally use and abbreviate to SOC.
SOC (BMS) - in a good battery this value is consistent and closely related to SOC but we don't understand what it shows in a very degraded car
SOC (precise) - we have no idea what this value represents.

Ordinarily we should only use SOC (display). The exception would be trying to understand what the other two are showing because it may be useful.
If you write SOC without specifying which one it should always be SOC (display).

 

[JejuSoul]

Regarding temperature, it looks like the second picture was taken well after the car finished charging (overnight?). Can you confirm kimdavi? I say that because in the second picture, the temperature is much lower than the first and the temperature difference between the modules aren’t that much different from each other after charging.

As for mixing different capacity batteries, I would be surprised if Kia engineers are that incompetent. That’s a big no no and basic knowledge for batteries of any type as the imbalance would cause the battery pack to fail much faster.

These are two different packs measured on different days. The first one does not have a high voltage deviation and the modules are similar temps.
It is the second one, the one that he is currently driving, that has the high voltage deviation.

Re-manufactured packs are made by a 3rd party not Kia.
Re-manufactured battery packs do fail much faster.
The relevance to this thread is that re-manufactured battery packs trigger 'limp mode' more often.
Do you have a better explanation for what we are seeing.

 

[kimdavi]

Regarding temperature, it looks like the second picture was taken well after the car finished charging (overnight?). Can you confirm kimdavi? I say that because in the second picture, the temperature is much lower than the first and the temperature difference between the modules aren’t that much different from each other after charging.

As for mixing different capacity batteries, I would be surprised if Kia engineers are that incompetent. That’s a big no no and basic knowledge for batteries of any type as the imbalance would cause the battery pack to fail much faster.

Yes, second picture was taken early morning after charging overnight.

 

[kimdavi]

My battery's charging capacity has been progressively decreasing. For instance, today I charged it at level 1 all day, yet it only reached 68% State of Charge (SoC) according to Car Scanner. What's intriguing is that the Display SoC indicated 100%, which is quite puzzling. Furthermore, I've noticed that the GOM range keeps dropping; earlier in the week, at full charge, it stated 93 miles, but now it's down to 80 miles.

 

[andrew1joe]

My battery's charging capacity has been progressively decreasing. For instance, today I charged it at level 1 all day, yet it only reached 68% State of Charge (SoC) according to Car Scanner. What's intriguing is that the Display SoC indicated 100%, which is quite puzzling. Furthermore, I've noticed that the GOM range keeps dropping; earlier in the week, at full charge, it stated 93 miles, but now it's down to 80 miles.

68% is the BMS SoC which tracks how much usable capacity your battery pack has remaining. Display SoC will always show 100% at full charge as it scale the BMS SoC. For example, at ~34% BMS SoC, your Display SoC will show 50%, etc. It looks like your battery pack is failing fast if it dropped from 74% not that long ago - it won't be long before it's in the shop again for a battery replacement.

 

[andrew1joe]

These are two different packs measured on different days. The first one does not have a high voltage deviation and the modules are similar temps.
It is the second one, the one that he is currently driving, that has the high voltage deviation.

Re-manufactured packs are made by a 3rd party not Kia.
Re-manufactured battery packs do fail much faster.
The relevance to this thread is that re-manufactured battery packs trigger 'limp mode' more often.
Do you have a better explanation for what we are seeing.

Sorry I worded poorly - I understood that first picture was original pack and second was remanufactured pack - I was just pointing out that the temp was low and difference small in the second picture which kimdavi confirmed was taken after charging overnight.

My guess is that they are using same capacity cells in spec for remanufactured cells but either (1) they are doing a poor job identifying and replacing degraded cells and/or (2) they are doing a poor job finding used/newish cells that on average matches left over capacity of other cells in the pack.

On the other hand, if they are using completely different cells, that would also explain premature failure but that would be shocking level of neglect on the 3rd party that's remanufacturing the pack.

 

[JejuSoul]

...
My guess is that they are using same capacity cells in spec for remanufactured cells but either (1) they are doing a poor job identifying and replacing degraded cells and/or (2) they are doing a poor job finding used/newish cells that on average matches left over capacity of other cells in the pack.

On the other hand, if they are using completely different cells, that would also explain premature failure but that would be shocking level of neglect on the 3rd party that's remanufacturing the pack.

I agree with this guess.
But they are doing many of these replacements. As one pack comes in they just strip out some cells and add in a few of the better cells that came from the previous bad pack. It would not be surprising that occasionally mistakes were made.
We seem to have an actual example of such a mistake being discussed here.