Reported battery SOH improving over time, rather than the reverse?

[BEC]

In the past year the estimated range of our 2016 Soul EV has dropped from the mid 80-mile range where it had been for three years or so to around 50 (or even as low as 47 after a really chilly night). Back in the fall, after I was able to get the BMS to recalibrate after the parking brake and BMS update recalls (prior to which I did get them to read the SOH), I saw the SOH drop to 69.7% per Soul Spy. I figured this wasn't enough margin below 70% to go running to the dealer to start the warranty replacement, but I've been dismayed to see the reported SOH actually go up slowly since then. Just now I read it at just over 77%.

Even though it's the chilly time of the year hear in the Seattle area, I know I've been able to get the battery to recalibrate a couple of times in between then and now....but I have been surprised every time that the reported-by-Soul Spy SOH has improved.

Has anyone else had this experience?

I really want to get the battery replacement process started because when I discussed this with the service department at my dealer last fall they made it sound like getting the new battery was a very long lead item (which doesn't surprise me).

In the meantime, we still drive it nearly every day, but as the range has fallen to where it is now, suddenly we actually have been having range anxiety over ordinary usage, and one trip I routinely make now requires me to charge away from home when I never had to before last fall.

 

[IanL]

I believe it is possible for the SOH to improve, but my evidence is tenuous.

I have a 2018 MY (30 kWh) car, and SoulEV Spy just reports 100% since its last update (before that, if I remember correctly, it could not produce a valid datum). After the update, it states the value is read out of the BMS, whereas prior to the update it was calculating from cell deterioration data, which did not work for the 30 kWh model. I think Spy still uses cell det. data for the early model, so I assume the BMS SOH value for that model is not available in the OBD2 data.

Because of the uninformative 100% reading I get, I developed a method to calculate SOH from dashboard data (SOC before and after charge, miles covered between charges, and displayed m/kWh, which has to be reset prior to the next charge, after the value is recorded). The resulting value of SOH is dependent on the accuracy of the values displayed, which is not high (SOC, miles and m/kWh rounded to whole integer value), so some variation can be expected, but it should even out over the long term, and show a valid trend.

My results do show an improvement from when I got the car at 2 years old (31,700 miles) of 94% to present day (4 years old and 39,200 miles) of about 104%. The values use a baseline of 30kWh=100%, so 104% equates to 31.2 kWh, i.e. the car is still using the upper margin of capacity.

Of course, I do not know why it has improved, but my theory is that its early life did comparatively high mileage, so longer journeys using all or nearly all the range, whereas my use is low mileage, charging 30% to 80% capacity. Incidentally, the DC socket still has the factory adhesive plastic film cover, so it has never done faster than a 7.2 kW charge. The first value, when I got the car, is probably affected by the car sitting around unused and unrecharged for some time between sale and purchase, which may also have a bearing. However, subsequent values showed 99% and 102% on the way up, i.e. a fairly progressive improvement.

The apparent improvement in SOH is supported by an apparent improvement in predicted range (i.e. GOM values adjusted for SOC when not charged to 100%). The January values have gone from 112 through 117 to 120 (7% improvement).

Having said all that, I am a bit doubtful of the validity of Spy data calculated from cell deterioration. If you want to try my calculation method before returning to the dealer, the method is detailed here

 

[JejuSoul]

-
The only relevant measurement of SOH for battery replacement is the one done by KIA.
Just take your car to them and have them measure it.

 

[SmurfUK]

Curiously, our 2016 27kWh smurf has recently had a BMS update whilst in for a routine service. The dealer didn't know what for. After a few empty-ish to full charges the SOH went from "unknown" to 108.9%, where previously it was hovering around 102 - 104% during most of 2021 (charging to full around once a month). UK winter range with a full charge remains something like 94 miles, summer range usually around 112 miles.

Wondering if Kia has started adjusting the SOH calculations?

 

[IanL]

I've not heard of a BMS reporting SOH in excess of 100%, so are these numbers from SoulEV Spy? For the 27kWh models, Spy calculates SOH based on cell deterioration data (presumably because the BMS does not provide SOH to the OBD connector), so if the BMS update has affected the cell det. data, that would explain it.

For the 30 kWh (2018 MY) model, Spy uses the SOH from the BMS, and that never exceeds 100% - it isn't clear whether the value is altered by Spy to never exceed 100%, or whether that is how the BMS provides it.

Btw, 108.9% SOH implies the capacity of the battery is 29.4 kWh, which is rather unlikely, unless the battery cells have been (recently?) replaced with those with improved chemistry used in the 30kWh model.

 

[JejuSoul]

...Wondering if Kia has started adjusting the SOH calculations?...

Yes. The calculation of the SOH made by the BMS varies depending on which version of the BMS you have had installed.
The only accurate way to measure your SOH is to take it to KIA and have them measure it.
I have two 2015 Soul EVs. The blue one has an original BMS. The white one has a newer BMS.
The method used to calculate the SOH has changed.

...Btw, 108.9% SOH implies the capacity of the battery is 29.4 kWh, which is rather unlikely...

For a 2016 Soul EV the total capacity of the original battery is 30.5 kWh

 

[IanL]

...Btw, 108.9% SOH implies the capacity of the battery is 29.4 kWh, which is rather unlikely...

For a 2016 Soul EV the total capacity of the original battery is 30.5 kWh

So, it's possible, but surprising comparing with the mileage the OP has stated. If the battery has only lost 1.1 kWh, i.e. 4% capacity, I would have expected itro 130-135 summer and 115-120 winter (UK). I base that on my figures, proportioned for the smaller battery, as mine also appears to have lost 4%. But driving patterns, etc, so maybe not.

I would be interested to know what the source of the SOH value is.

 

[SmurfUK]

I've been using Soul Spy since we purchased the car. The dealer printout when purchased (3yrs old with 12k miles on the clock) stated 100% SOH, whereas Soul Spy has always indicated above that, usually varies between 102 to 104%. Reasons for this understood from reading other threads. Would love to believe the SOH could improve but I'm thinking it's just the way it's calculated, hence my wondering/suspicions about why it seems to have jumped up (to a somewhat unbelievable figure) after a BMS update.

 

[BEC]

OK....then I just need to get an appointment and get the official measurement done again. As it's been warm the last week or so for winter in Seattle the GOM range estimate has improved a bit more to as much as 55 miles. I need to do another SoulSpy readout as well.

I really want it to be at the 70% threshold so I can start the "case" needed to get the battery replacement process going. In the meantime I intend to keep on driving it if I can. The deterioration seems to have flattened out...or maybe its just the warming weather making it look that way.

 

[IanL]

When you go to the dealer, arm yourself with a copy of TSB PS548, and insist that they do as it says. Otherwise, they usually reset the BMS, which will ensure the SOH reads very high and then takes weeks to re-learn your battery parameters. Of course, because the SOH has gone high, they think the reset cured it.

 

[jray3]

I'm in the same boat as Bernard, south of Seattle with a 74 kilomile 2016 that hit 63.4% SOH on 1/27/22 when I set my dealer appointment for warranty EValuation.
A job change meant that I went from over 151 miles per day of driving (with 4 DCFC) to largely virtual work, and usage dropped to maybe 12 miles per day average in warming weather. By the time the dealer measured SOH, it was a freezing cold day but the BMS reported a lousy 70.3%!! Faced with my SOULSPY data, KIA of Puyallup opened a TechLine case and tested again the next day, but Corporate told me to come back later with a reading below 70%.
Appointments are still running 6 weeks out, so it'll be mid-April before I take another crack at the SOH meaurement.
At least I got a car wash out of the deal.... :roll:

Hey Bernard, come on up to the Seattle EV Association meeting this coming Tuesday and we can compare notes. https://www.seattleeva.org/wp/

 

[IanL]

MY 2018 SoulEV reports the BMS SOH on the OBD interface, so SoulEV Spy would give the same value the dealer reads. The earlier model does not, and Spy calculates a value based on cell deterioration, which does not necessarily use the same method as the KIA test equipment, so may give a different value.

 

[jray3]

The saga continues. 63.4 SOH according to SOULSPY on January 27th became 70.3% at the dealer by March, and now on April 12th, it measures 70.6%! Warming weather and sitting at 100% seems to only boost this SOH algorithm.. For good measure I used it on a max range run the night before, and a full recharge from 15% SOC only consumed 15.1 kWH, The return trip down to 5% SOC was only 55 miles, with cautious 60 mph driving. if 15.1 kWH is 85% of actual capacity, that yields a capacity of 17.76 kWh, or 65.7% of the 27 kWh spec, but not low enough to satisfy this BMS..... :roll:

 

[BEC]

I just realized I haven't been back to this thread for awhile. Quick update:

The SOH continued to decline from when I last posted here, as more or less was expected.. I got the official test again in early July and the SOH was well below 70%, so TechLine case was opened. We continued to drive the car until this past Thursday. By then the full charge GOM range was down to 34 miles and the car started giving us the turtle both going up hills that used to be easy, and even a couple of times just cruising down the freeway at ~65 mph. State of Health as reported by SoulSpy at this point is down in the mid 30% range.

So, we finally gave up and took it in to drop it off and arrange for a rental car, which we picked up today (late model Toyota Corolla sedan). This is to cover having our 41-year-old second child, who is currently living at home, having trustworthy wheels for medical appointments and such while my wife and I take a two-week+ road trip in our Mazda3, starting a week from Monday

The last time I spoke with the knowledgable-about-EV-things Linda at KIA corporate (September 8th) we were 10th in line for a new battery and she estimated about two more weeks...but so far as I know as I type this it has not been released for shipment from wherever they check out incoming replacements in California. We are hoping that while we are gone the battery makes its way to Kia of Puyallup and we can get the Soul back after we get home on the 18th. That's the best-case scenario, anyway. Well, a miracle could occur and the crate with the battery shows up in Puyallup on Monday and we get the Soul back before we head off to NM, but that's really pushing it.

So we shall see.....