Analysis of a replacement battery.

[JejuSoul]

...Is the BMS firmware different for the different battery chemistries? My '16 Kia Soul EV recently got a 37510-E4200R battery replacement. I noticed that the SOH still reads 68.4%, same as when I brought it in. I assumed that they would need to update the BMS firmware to handle the new chemistry, and that doing so would reset the SOH back to 100 (or 110?) percent...

Yes. Here is a list of the various ROM-IDs for each of the BMS Firmware updates on the original Soul EV.



This is badly translated from the Korean as is so often the case with Hyundai / Kia.
E4050 - 'Elevated Temp' actually means with Battery Heater.
E4000 - 'Non-Heating' means without Battery Heater.

That you have a 37510-E4200R signifies the new pack has the new cells, has no battery heater and is 're-manufactured'.
What 're-manufactured' means in this case is uncertain. Did they take old cells out of a broken 30kWh car.?
Did they mix and match old and new cells? We don't know yet. We just know there is a big price difference.

Looking at the Kia parts website - Kia Soul EV Re-manufactured Battery (37510-E4250R)
This battery costs $1,609.50 and fits all models both 27kWh and 30kWh.
It is just the cells, they reuse the case and other electronics so this is for cars with and without battery heater.

Looking at the Kia parts website - Kia Soul EV Replacement Battery (37510-E4250)
This battery costs $12,293.25 and fits all models both 27kWh and 30kWh
It is just the cells, they reuse the case and other electronics so this is for cars with and without battery heater.

They add a $2,500USD core charge to both.



That your SOH hasn't changed shows that they forgot to update the BMS. This is an essential step for non re-manufactured packs.
We don't know the requirements for 're-manufactured' packs. Presumably though it should be the same.
I would take the car back and ask them to update the BMS.
-

 

[shumanchu]

Ok, from that table it looks like I have probably have BMS ROM ID 5522 and what I need is BMS ROM ID 5722. This information is covered under Kia TSB ELE 155 (upgrade action 379.PS).

Getting this battery replacement has been a nightmare. After 7 weeks with no car, I get my car back with a remanufactured battery instead of new, and with the incorrect BMS firmware to boot. They also left the High Voltage safety disconnect cover open, and there are multiple grease marks in the interior. I'll try to get them to re-flash the BMS to 5722.

After they re-flash, I assume I should do a drive cycle/charge cycle (below 10% SOC and uninterrupted charge back to 100% SOC)?

 

[shumanchu]

So I took my Soul EV into the dealer today finally, and they were able to install the correct 5722 BMS firmware into the car. Took about 1 hour, and they washed my car too!

Now the SOH is 111.5, and the GOM reports 93 miles on a full charge!! I am very happy with that, even though it probably could have been higher if they used a new 37510-E4200 battery instead of the remanufactured one.

I will make another post after a complete < 10% SOC to 100% SOC drive cycle with the recalculated SOH.

 

[JejuSoul]

-
Great that they fixed your car with the proper software.
Your SOH isn't really 111.5%. That is just a fantasy number that pretends to estimate the buffer size. In reality your SOH is probably 100%.
The fantasy number was introduced to take account of the buffer size in the original 27kWh car with the original BMS.
The new BMS has a different algorithm. We don't know how it works.

I assume a brand new battery with a brand new BMS cannot be calibrated further, since by default it is fully calibrated. But am always appreciative of real world data that backs up such assumptions.

 

[IanL]

Assuming the remanufactured battery has the same cells as the non-R battery, I would expect the performance to be the same. If you have not done much distance yet, the BMS is presumably heavily biased to default value. It will be interesting to see how the GOM changes with use.

 

[shumanchu]

I'll keep posting with new GOM and SOH numbers each time I do a drive cycle. I plan to do one this weekend, so I should have some new numbers on Monday.

Jeju, are you saying that the SOH that I read will never report accurately with the new battery and BMS firmware? Is that because Kia changed the algorithm and now the Soul EV Spy Lite app doesn't know how to calculate it anymore?

 

[JejuSoul]

...Jeju, are you saying that the SOH that I read will never report accurately with the new battery and BMS firmware? Is that because Kia changed the algorithm and now the Soul EV Spy Lite app doesn't know how to calculate it anymore?

For the first few years yes. Given that the Soul EV Spy Lite app will never be updated, it will always assume you have an original BMS.
The simplest guess on how the new BMS behaves would be to assume it is like the BMS in the 30kWh Soul EV which always shows SOH = 100% until after the 'early-years' buffer has been lost. After that it is unknown. Maybe it will work, maybe not. I estimate it will be 3 more years until we know.
-

The latest data from my car after 19 months, 22,000km

 

[IanL]

With my 30kWh model, the current version of the Soul EVSpy paid-for app gets the SOH from the BMS, rather than calculating it from Max Det and Min Det like Torque does. If the 5722 firmware is similar, it may be worth upgrading to the paid-for version to be able to read the value from the BMS. I'm not sure of it always has done that, but it now states the source of the SOH (implying that some other source may be possible), so at least you would know the origin of what is being displayed.

 

[JejuSoul]

-
SoulEVSpy knows nothing about what firmware the car has had applied.
The current version of the Soul EVSpy paid-for app shows a fantasy number for SOH exactly the same as mentioned above.
Presumably all the other apps will do the same. I'll post below when I have a screenshot.

 

[shumanchu]

With my 30kWh model, the current version of the Soul EVSpy paid-for app gets the SOH from the BMS, rather than calculating it from Max Det and Min Det like Torque does. If the 5722 firmware is similar, it may be worth upgrading to the paid-for version to be able to read the value from the BMS. I'm not sure of it always has done that, but it now states the source of the SOH (implying that some other source may be possible), so at least you would know the origin of what is being displayed.

Can you post a screenshot of your Soul EVspy app that shows "Battery SOH Source" is something other than "Calculated from BMS min/max deterioration"? Or just type in what it says, I don't really need a screen shot.

If the new 5722 BMS firmware has SOH as a OBD2 PID, that is, calculated by the BMS and reported like any other PID, then we could probably work out which PID it is and put that into Torque to read out the SOH. I'd also really like to work out how to read the BMS ROM ID over OBD2.

 

[IanL]

 

[JejuSoul]

-
It is now 2 years since I had the battery replaced on my white 27kWh 2015 Soul EV.
On December 30th 2019 the car had driven 86,993km before getting a new battery.
On December 30th 2021 the odometer reads 116,590km. That's 29,597km on the new battery.

Here's the latest Torque reading

Clearly the new BMS does not show the same deterioration data as the original.
The original BMS began showing the deterioration from the start as capacity was lost from the early years' buffer.
The new BMS does not do this. Maybe it will only show deterioration after the early years' buffer has been entirely lost.
Maybe another 2 years before I have a real deterioration reading on this car.
At the last service a few months ago the SOH was 100% as expected.

My blue 27kWh 2015 Soul EV is now 6 1/2 years old and has driven 88,554 km.
This car has the original battery and the original BMS.
Back in 2017 I have the readings for that car after 29,769km.
CEC 6185.5kWh, CED 6011.8kWh
Max Det 8.5, Min Det 1.1

 

[MassDeduction]

-
It is now 2 years since I had the battery replaced on my white 27kWh 2015 Soul EV.
On December 30th 2019 the car had driven 86,993km before getting a new battery.
On December 30th 2021 the odometer reads 116,590km. That's 29,597km on the new battery.

Here's the latest Torque reading

Clearly the new BMS does not show the same deterioration data as the original.
The original BMS began showing the deterioration from the start as capacity was lost from the early years' buffer.
The new BMS does not do this. Maybe it will only show deterioration after the early years' buffer has been entirely lost.
Maybe another 2 years before I have a real deterioration reading on this car.
At the last service a few months ago the SOH was 100% as expected.

My blue 27kWh 2015 Soul EV is now 6 1/2 years old and has driven 88,554 km.
This car has the original battery and the original BMS.
Back in 2017 I have the readings for that car after 29,769km.
CEC 6185.5kWh, CED 6011.8kWh
Max Det 8.5, Min Det 1.1

Congrats, looks like you're getting good results on both cars now. Good to see you post again, for the first time in a while JejuSoul!

 

[ksoul2084]

Those are some truly impressive numbers.

FYI. My battery was replaced (well, the cell packs were swapped out) last August. Since then I've only put on approx. 6,000km of mileage, usually charging using Type 1 to 80% max.

Last night I did an uninterrupted charge from 7% to 100% to reset the SOH, and let the car balance the cells, etc.

Once completed the car reported all cells at 41.6v, outside temperature was about 11c. Battery SOH is 99.7%, though estimate range was only 120km (lower efficiency with cooler temps and heaters running, etc. I guess).

I have a fairly disturbing update on my 2016 Soul EV+. Battery cells were swapped in Aug. 2020, I had posted the last update above in May of 2021. I have noticed the estimated range has dropped continually over this summer and fall, but hadn't checked with SoulSpy until today.

Jan 4th, 2022
Mileage: 167731.9km (so approx. 11,000km since replacement)
SOH: 91.9%
battery.min_cell_deterioration_pct: 16.3
battery.max_cell_deterioration_pct: 18.3

I almost always charge to 80% using Level 1 120v charging at home. I have charged to 100% using Level 1 about 4 or five times that I can recall in the last year.

This is very disappointing to me, to have my replacement cells with this high a deterioration with so little mileage and lack of abuse on my part.

Any chance it just needs to be top-balanced aggressively?

 

[IanL]

What is the spread of cell-pair voltages?

 

[ksoul2084]

What is the spread of cell-pair voltages?

Here is a sampling of cell voltages taken today, the first two when the car was started after 30mins scheduled preheat cycle (it is -43c with windchill outside here today, approx. -8c in the garage), and the next two after running a few errands around town, maybe 5km driving at slow speeds.

 

[IanL]

When I got my 30kWh model at 2 years old and 31,000 miles, I didn't know about Spy, but I calculated SOH based on dashboard data and charging records. If my method is valid, the SOH was 94% and the winter range (winters are mild here) was 112. I did eight charges (using the 10A plug-in charger) to 100%, because I didn't know any better. Then I started a 30% to 80% regime. By the next winter, the SOH was 102% and the range improved to 120 miles. At that point, the cell-pair voltages were pretty even, with about eight at 0.02V below the others.

I carried on with 100% charges every 4 recharges or so, till I saw that all cells read the same, and I then stopped the balancing charges and haven't done any for 8 months. The cells have remained balanced. At 39,000 miles, the SOH averages about 102.5% and the winter range about 120 miles. During that time, I fitted a wall box charger which can do 32A, but I usually use 16A, because what's the hurry?

My home-grown SOH method may not be valid (Spy just reports 100% by readout from the BMS, not max & min det), but there's some evidence that a measure of recovery is possible.

 

[MassDeduction]

When I got my 30kWh model at 2 years old and 31,000 miles, I didn't know about Spy, but I calculated SOH based on dashboard data and charging records. If my method is valid, the SOH was 94% and the winter range (winters are mild here) was 112. I did eight charges (using the 10A plug-in charger) to 100%, because I didn't know any better. Then I started a 30% to 80% regime. By the next winter, the SOH was 102% and the range improved to 120 miles. At that point, the cell-pair voltages were pretty even, with about eight at 0.02V below the others.

I carried on with 100% charges every 4 recharges or so, till I saw that all cells read the same, and I then stopped the balancing charges and haven't done any for 8 months. The cells have remained balanced. At 39,000 miles, the SOH averages about 102.5% and the winter range about 120 miles. During that time, I fitted a wall box charger which can do 32A, but I usually use 16A, because what's the hurry?

My home-grown SOH method may not be valid (Spy just reports 100% by readout from the BMS, not max & min det), but there's some evidence that a measure of recovery is possible.

Good call. Charging that car at 32 amp would exceed the ideal charging speed. IIRC correctly, that battery pack is 2 modules of 40 Ah useable, so the full pack is 80 Ah. With the ideal charging amperage being 25% of the Ah rating, that means 20 amp is the maximum ideal charging speed for the 30 kWh pack. For the 27 kWh pack, the ideal maximum charging rate is 18.75 Ah. So yeah, if you're not in a hurry, don't exceed those numbers!

I try to store at a medium-ish state of charge, and then wait until the last convenient moment to charge it any higher. I do that with everything I own with a lithium ion battery (cell phone, e-reader, e-bike, e-car, you name it). I definitely do not plug my phone in and leave it on the charger all night anymore! In fact, my phone is never plugged in at night. And I don't fast charge it when there isn't a need, either.

Though this does mean, having never drained the Soul EV below 18%, nor ever charged it to 100% in all the time we've owned it, we have NO idea what the SOH is. But Soul Spy reports that the cells are still nicely in balance, so there's likely no need. If they ever get enough out of balance to require a conditioning charge, then we'll make sure to drop below 18% first so that we also get an SOH reading.

 

[IanL]

...Though this does mean, having never drained the Soul EV below 18%, nor ever charged it to 100% in all the time we've owned it, we have NO idea what the SOH is. But Soul Spy reports that the cells are still nicely in balance, so there's likely no need. If they ever get enough out of balance to require a conditioning charge, then we'll make sure to drop below 18% first so that we also get an SOH reading.

If you would like to try my calculation method, which does not require you to charge to 100%, it is here.

 

[MassDeduction]

...Though this does mean, having never drained the Soul EV below 18%, nor ever charged it to 100% in all the time we've owned it, we have NO idea what the SOH is. But Soul Spy reports that the cells are still nicely in balance, so there's likely no need. If they ever get enough out of balance to require a conditioning charge, then we'll make sure to drop below 18% first so that we also get an SOH reading.

If you would like to try my calculation method, which does not require you to charge to 100%, it is here.

Yeah, I remember reading that earlier. Thank you for reminding me about it! I expect your method would provide a pretty good guesstimate, for sure.