2016 Soul EV replacement battery - range unchanged and voltage question.

[evsoulnow]

Okay I have scoured the forum and not found all the answers to this, it there's a relevant thread please just post the link (and apologies for missing it).

Just supposedly had a replacement High Voltage battery fitted by Kia under warranty to our 2016, 30K mile old Soul EV. It went in with a 50 mile range, and came out showing the same. Reading up this seems likely to be a recalibration issue, and I understand I need to cycle it a few times - the service bulletin I found linked here said below 20% SOC to above 90% SOC to trigger recalibration and it might take a few cycles. So that explains why range is wrong. Can I expect a big change on the first cycle, and then fine tuning, or is it a gradual change?

Second question. The 'new' (remanu) battery shows 4.02 volts on all cells using EVSpy Pro. I see references to 4.08 mostly here, but I know the old one was 4.02 - is that a battery generation difference?

EVSpy also shows battery.max_cell_deterioration_pct as 57.1 - is that just reading the BMS which has not been recalibrated? Or is the replacement battery as defective as the last?

Any insights appreciated. I am of course going to cycle the battery and see what happens, just looking for pointers ahead of that. As a final question for that process is the battery charge %age shown by the car when powering down safe to use as a guide to SOC without the recalibration? Or do I need to teach my daughter (the driver) to use/read EVSpy for each journey?

Thanks in advanced for any guidance.

 

[IanL]

I don't know for sure, but I think the BMS "recalibration" will be slow (i.e. require many cycles). My understanding is that the full procedure for replacing of the battery should include reset of the BMS, which implies it will use stored default values as its starting point. With all new cells, that should result in a much more rapid convergence to representative values. It sounds like this was not done in your case.

The 57.1 % value is, as you say, almost certainly a result of the "uncalibrated" BMS, as the value is provided by the BMS. Bear in mind that SoulEV Spy was developed without assistance from KIA, so many of the parameter identifications are best estimations of the authors.

I think the displayed SOC %age is based on the cell voltages (rather than values filtered by algorithms), so should be representative of range available.

 

[JejuSoul]

-
The full procedure for replacing of the battery should include updating the software of the BMS, which implies all values will be reset to zero.
You need to take the car back to Kia and have them do it properly.

-

 

[evsoulnow]

I appreciate your response, in conversation with Kia now.

 

[evsoulnow]

To complete this thread, in case others find it.

• Kia updated the BMS software when they installed a replacement BMS earlier this year

• After installing the new BMS, a new battery was authorised

• When the new battery was installed, projected range was unchanged

• We tried depleting to <20% and charging to 100% three times using our level 2 charger - finally changed projected range by 5 miles (it's still missing about 40)

• Kia reviewed again, but stated they could not force a BMS update as it was on the latest version

• We tried depleting to 2% (that was scary) and charging to 100% using 110V/13A (that was a frustrating 28 hours) - this added another 5 miles of range

• There is currently a recall for another software update to the BMS (relates to the recent battery recall I believe) - but this is not yet available
  
  So the car is gradually relearning - sloooooooowly - how far it can travel. By my estimates it is doing around 100 miles on the replacement battery. Sometimes we are seeing more than 1 mile per 1% of battery, but most of the time it's pretty much 1:1.
  
  Driving on %age and ignoring the miles is stressful (especially when the projected range goes to "---").

 

[IanL]

KIA have not made clear how many cycles are required to get the BMS algorithm reasonably close to representative output - they just say "multiple cycles".

I have consulted my records, using the summer/winter variation of GOM value (extrapolated to 100% SOC), and it seems that something like 10 cycles of (roughly) 30% to 80% were needed for the GOM to reflect the change from nasty cold days to nicely warm days. Because my usage of the car is moderate (about 270 miles/month), this resulted in my top GOM value (145 miles) occurring in mid-November!

Frankly, I think the GOM is pretty useless and pay it little attention. When it gets below 50 miles, I use the SOC as a gauge. After all, in ICE cars, no-one calibrates the fuel gauge in miles.

 

[evsoulnow]

It's interesting. Here in Colorado with a temperature range from -10F/-20C to 100F/37C (and more in both directions) - we found the GOM fairly close on the original battery - as in it dropped based on ambient during the winter and increased in summer (my daughter tries to avoid running the A/C unless really hot).

The new battery is now behaving really weird. When it was not reading well and I took it to Kia, it had 29% SOC - it took me more than 40 miles, driving it hard (90+ for some of that on the highway) plus running the A/C to get it down to 2%.

My daughter has been tracking actual miles vs battery %age. It managed 124 miles with 17% reported by the car when she got back from the last leg of that. Today (similar temperatures, similar driving style) it eat 35% of the battery in 44 miles.

Based on the first two runs after we ran it down to 2%, and did the marathon 110V/13A slooooow charge - it was looking like the range was around 140 miles (based solely on miles driven and available %age). Certainly it manage the one cycle of 124 miles without getting stranded and was reporting low battery (comes on below 25%) down to 17%.

I cannot find any battery Kia might have installed which would support this range though. Makes me really nervous that the reported 17% is wrong and that she was actually almost out of a battery including buffer.

So many questions - the main one being why on earth Kia would not give the dealer would a way to reset to default GOM (yes I get the risks, but you'd simply flag when the GOM was reset, so a prospective buyer could see it had happened and when - under EV on the computer).

 

[IanL]

It is likely that your reman battery is using the new-technology cells introduced in the 30 kWh vehicles, probably because the stocks of the original cells have been exhausted. The new cells have higher capacity, but you only have 96 cell-pairs, whereas the 30 kWh batteries have 100, i.e. the nominal 27 kWh when new is extended to 96/100x30= 28.8 kWh. So your range expectation should be 28.8/27x original range when new, e.g. 140 miles becomes 149.3, and 130 miles becomes 138.6.

So, when new, could the car do 130-ish miles under the same conditions?

 

[evsoulnow]

Nope, when new to us (only 5.5K miles) is was 93, delivering as much as 104 miles when GOM was ignored and battery SOC was used. So we are totally confused. Current run is on track for around 130 miles again. AFAIK Kia don't and didn't have batteries with this range for any of their line :?

 

[IanL]

When I got mine, it was 2 years old with 30,000 miles. Its GOM range was and still is 120 Winter (maybe not as cold as yours) and 145 Summer ( maybe not as hot as yours). If it had your reman battery in the same condition, those values would be 115 and 140. That seems to be consistent with what you are experiencing.

Your original battery, under the same conditions, should have given you 108 and 126 miles, so my guess is that your original battery had some degradation when you got the car. Your replacement battery does not, hence the improvement. Enjoy!

 

[evsoulnow]

Wonder if this is US EPA GOM vs elsewhere ... not sure when I Google "2016 soul ev range" the interwisdom is the 93 miles we had.
Oh well, there is an outstanding (but not available) software recall for the BMS which might shed some light on this.
As you say - she's just enjoying the reduced range anxiety.

 

[Locuturus]

It's interesting. Here in Colorado with a temperature range from -10F/-20C to 100F/37C (and more in both directions) - we found the GOM fairly close on the original battery - as in it dropped based on ambient during the winter and increased in summer (my daughter tries to avoid running the A/C unless really hot).

The new battery is now behaving really weird. When it was not reading well and I took it to Kia, it had 29% SOC - it took me more than 40 miles, driving it hard (90+ for some of that on the highway) plus running the A/C to get it down to 2%.

My daughter has been tracking actual miles vs battery %age. It managed 124 miles with 17% reported by the car when she got back from the last leg of that. Today (similar temperatures, similar driving style) it eat 35% of the battery in 44 miles.

Based on the first two runs after we ran it down to 2%, and did the marathon 110V/13A slooooow charge - it was looking like the range was around 140 miles (based solely on miles driven and available %age). Certainly it manage the one cycle of 124 miles without getting stranded and was reporting low battery (comes on below 25%) down to 17%.

I cannot find any battery Kia might have installed which would support this range though. Makes me really nervous that the reported 17% is wrong and that she was actually almost out of a battery including buffer.

So many questions - the main one being why on earth Kia would not give the dealer would a way to reset to default GOM (yes I get the risks, but you'd simply flag when the GOM was reset, so a prospective buyer could see it had happened and when - under EV on the computer).

Hihi I've got the same spec 2016 with a remanufactured battery (Dec 2023), but I have been consistently noticing that I'm ending up with more range (GOM is sometimes right and more often wrong, hence the name I guess, though Car Scanner says SOH is 110%). I have gotten it down to about 30% - though on that trip I was expecting to end up with 18% range. ABRP with an OBD kept creeping up the estimates as I drove (up the I-5 from Bellingham back to Vancouver). I haven't had the time to drive it down to the single digits to really get the calibration going, but my range is currently around the 160km (99mi) mark @100% SOC per the GOM with HVAC on. I'll try driving it down to 5% and measure the distance travelled, and trickle charge it back up, and come back with some values.

Vancouver is just wet all the time, so I imagine GOM is taking the road conditions into account. Lately the weather is hovering between -5 and plus 8 C (so around the 20-40F ish?) so maybe that has something to do with the GOM range as well. I'm also changing to all-weather tires today so that might change things slightly too (Nokian WRG5s)

 

[uspurs]

I would be very interested to see the energy usage you're seeing - miles per kwh; in our '17 with a reman battery fitted last summer, I pretty much only rely on the SOC and energy read-outs.
In the recent cold weather here in the NE, we saw as low as 74 on the GOM on a full charge, but SOC % and energy read-out was indicating a real-world range of 100+.
We just got the Soul back after the SC267 recall and the GOM is now showing 113

 

[Locuturus]

I just got my Nokian Remedy WRG5s installed, and consumption shot up to 26-27 kWh/100km - about 2mi/kwh LOL (my Canadian 2016 doesn’t have the option to change distance units, but does have temperature unit change…)

I’ll break the tires in first, and drive it around some to get the GOM to figure out I changed the tires….will keep this updated…

 

[Locuturus]

After about a week of driving (took the car up from Vancouver to Squamish and measured since I remembered to do it)

• distance of about 76.5km, and back so x2 = 153km total (per google maps)
• elevation is hard to measure, but there's a lot of big ups and downs (though if you just looked at endpoints overall I only went from 32m to 4m (104' to 3'))
• speed on the route ranged from 60-100km/h; avg around 70km/h
• consumption on the car ranged from 20-19 kWh/100km or 3.1-3.2miles/kWh per the car's consumption report.
• ABRP with an OBD2 plugged in, set to Kia Soul 2016 27kwh profile showed these stats: 74km (one way going up), averaged at 56km/h, 166Wh/km (16.6 kWh/100km), started with 76% and ended with 30%

I haven't gotten around to organizing a route+time to do a Bjorn style test (I can't run down to zero cos I don't have an Ecoflow, and chargers aren't as plentiful as in Norway), but I'd like to run it down to about 10% or below at least from 100%, so I can get an accurate number.

Edit: I'm not exactly sure why ABRP is wildly different from the consumption recorded from the car... but hopefully wearing the tires a bit more will help even out the number.