2015 Soul EV: My Battery Replacement Story

I originally posted about my Soul EV in the battery aging thread, but I'm re-posting here to keep my story separate and easier to find. It's my hope that my experience will be insightful for your own situation.

I purchased a 2015 Soul EV, with roughly 37k miles in Dec 2017. I'm the second owner and I live in SoCal. The car is a 3-year lease return and the original owner lived in Northern Cali and drove the 37k miles. When I initially bought the car, the GOM read around 70 miles, but as I drove it more, it dropped 70-> 65->60->55->50 miles. There seemed to be some serious battery degredation. I decided to do a range test to see how far I can drive on a full charge (and how much kwh can be pulled out). Here are my results.

Here are Torque readings from before the range test:



The test was done at 80mph on flat highway with no wind, cruise control on. Low battery warning came on at ~45mi (when range went to ---). Ultimately I was able to drive ~51 miles until I hit 1% (last few miles @ city speeds). Consumption was around 2.5mi/kwh on the dash.

Here are Torque readings from after the test:





Torque reading from full charge to 100% the next day:



Since neither the car nor Torque provides the kWh used, the only guess is by looking at what it takes to recharge to full 100% from the EVSE. It is a very constant 6.6-6.8 kW charging speed, until the last 20-30 mins, then it tapers off. Note that EVSE kWh reading is from the wall. Please only look at the 1st row (other rows is from charging a dffferent EV):



Based on the kWh used to charge, (and assuming 100% charging efficiency, which it's not), 20.4 / 27 kwh usable means 75.5% SOH. (20.4 kwh * 2.5mi/kwh = 50mi on GOM seems roughly correct). I also want to note that based on several forum members here, my SOH displayed on Torque didn't drop from 83.1% -> 79.4% from driving fast or deep-cycling to 1%. The BMS simply re-calibrated itself to the proper SOH values after been driven down to low SOC and charging back up too 100%. No harm was done to the battery by doing this test. See the calibration thread for more details on this.

Based on these SOH values, several forum members here recommend that I take in the Soul EV to the dealer to get a SOH report. This will help us validate the Torque SOH numbers, and I will potentially get a battery replacement from Kia if my SOH is registered to be less than 70% on their end based on their capacity warranty.

Kish said:

...Since neither the car nor Torque provides the kWh used,...

Click to expand...

Torque does provide the kWh used. The method I have written just below this on how to read the kWh going into and out of the battery works on all the batteries we have tested so far. Those all have SOH=100%. You are the first person to test a battery on the Soul EV that is severely degraded. But this method may not work in your case. Could you show us these readings from your car. Or even better could you show a graph similar to the one drawn for the Kia Ray EV below.

Torque provides the Cumulative kWh charged and discharged into the battery. To get the input kWh for one charging session subtract the CEC value at the start from the CEC (Cumulative Energy Charged) value at the end. To get the output kWh for one driving session is slightly harder because you need to take into account the regenerated energy. You calculate that by finding how much charge went into the battery during driving using the CEC counter as before. The CED (Cumulative Energy Discharged) counters will give you the kWh used during the driving plus the additional regen.

I have tested on an old Kia Ray EV with serious degradation. - Re: Comparing BMS data on the Soul EV with the Ray EV
The Cumulative Energy counter shows an anomaly. It does not go up in a straight line as our non-degraded Soul EVs do. This explains why the Energy Counter cannot be used to measure the capacity of the battery

It was quite a challenge to find a dealer who is willing to service the Soul EV, even in SoCal. The initial dealer that I made an appointment with over the phone mentioned that they service EVs, but when I showed up to my appointment, they sent me away saying that they don't sell or work on EVs (they had several Hybrids/PHEVs for sale though!). After a lot of calling and searching, here are some of the SoCal dealers that service the Soul EV, for reference:

• Kia of Alhambra

• Kia of Irvine

• Kia of Orange

• Kia of Pomona

While waiting for my 2nd appointment to arrive at the dealer (which btw took a while to get), I checked the SoH values on the Soul EV again. And to my surprise the SOH was up to 91.7% (up from 79.1% 2 months ago)! Hoping that it was just the BMS being un-calibrated again, I ended up doing another range/drain test:



To 1% yielded only 47.9 miles! (Down from 51 miles last time):





After a full recharge, the Soul EV pulled 18.52 kWh from the wall and still shows 90.2% SoH! Based on the kWh used to recharge, the SOH should be 18.52/ 27 kWh = 68.59% SOH.



Now even more anticipating the visit to the dealer, really curious on what's going on with the SOH.

Finally the day has arrived, I took my car into the Kia dealership in the morning.

I explained my situation to the service advisor (I'm only getting ~50 miles on a charge). He initially misunderstood and thought that it was due to the onboard charger (OBC) not charging up to 100%, so I'm not getting a full charge. I told him that that wasn't the case at all; I have no problems charging at level 2 to 100%. He then checked in his "system" and indeed saw that the OBC has already been replaced once by another dealership, and confirmed that the OBC can't be the issue then (duh! :roll: ). I told him very clearly that I want some diagnostics done on the battery pack and I wanted a printed SOH report. I then left, after I was promised that the car would be returned by end of the day, after they run through their diagnostics. I asked for a loaner vehicle but was denied (loaner only provided if Kia approves a repair).

Later that afternoon I got a call from the service advisor letting me know that they ran their diagnostics (fault codes) and they weren't able to find anything wrong, so they can't do anything for me. In response I told him that I expected as much (no error codes), in that case I just want the SOH report (which is what I originally ask for anyway!), and I'll take the car back. He didn't like that at all-- he wanted $110 (1h of labor) to run that SOH report, and I said go ahead do it. (I didn't want the car at the dealership without any new info!)

Later in the evening I got another call from the service advisor saying that "Kia has requested more testing be done on the vehicle", and they need to keep the car for another few days, as their "EV specialist/technican" isn't currently available. (Wow, so they just read fault codes from a OBD II reader, not very useful! :roll: ). Still no loaner vehicle though-- same story "only provided if/when Kia approves a repair". So I'm sitting without a car for a few days, not so happy

A few days later I get a call back from the service advisor-- Kia has approved for my Soul EV for a free battery pack replacement under warranty! I tried to ask more details on the battery replacement conditions, but he just said that Kia was not happy with the test results. Naturally I also asked him how long that would take and he said a few weeks. Apparently the last time he did a battery pack replacement for the Soul EV, it took him 3 weeks from start to finish. How about a loaner vehicle? Yep, since the repair was approved, I get a loaner for the time being. Sadly it's no EV, just a regular ICE vehicle.

I'll provide an update once I get my Soul EV back with the replacement battery pack. I'm excited and curious what the Torque values will be!

-
Am glad you are getting a new battery. The old one was clearly problematic. You were right to insist on further checks.
If you happen to have the logfiles from your previous Torque tests, I would be happy to see them.
I commented upthread earlier today about looking at the Cumulative Energy data on Torque.
Too late to do new tests now.

Yeah, unfortunately I checked my phone's Torque logs under the dir ~/torqueLogs and it was empty. ~/.torque/tripLogs does have the logs from all my driving tests, but it only has the GPS data which isn't useful. Thanks for the pointer on CEC and CED, I will use those Torque values in my next driving test, in addition to using my EVSE data. From what I understand:

kWh used to charge
= CEC (after charging) - CEC (before charging).

kWh used while driving
= total kWh discharged - total kWh charged (regen)
= (CED (after driving)) - CED (before driving)) - (CEC (after driving) - CEC (before driving))

A representative from Kia HQ called me today to let me know that the replacement battery pack has been shipped to the dealer yesterday. That's essentially 1 day after they approved the repair ! I was expecting the replacement to take 3-4 weeks, but looks like I may be able to get my car back sooner than that. Out of curiosity, I asked them if they knew from where the battery pack is being shipped from (since some other forum members' packs were imported from Korea), but they did not have a concrete answer, and simply said that Kia has warehouses throughout the U.S.

I will follow your story.

thanks for comment

The service adviser at the dealership called me today, Soul EV is ready for pickup! Checked UVO and am seeing 92 miles @ 91% (This was ~50 miles before at 90%):



Of course these are just estimates, the real test will be to see what Torque says and a driving test. The battery replacement was completed 8 calendar days (6 business days) after the repair was approved by Kia. This was 7 days after the replacement pack was shipped to the dealer from the warehouse. I'm not quite picking up the car yet, I'm having the dealer look into one more issue (unrelated to battery).

Picked up my Soul EV from the dealership. Looks like they replaced both the battery pack and the BMS. Here is what the repair paperwork says:



It may be a bit difficult to read, so here is what is says: "Customer reports will charge to 100% but range will only give 55 miles. Cause: checked no codes present, checked BMS battery SOH and was at 66.5%. Also checked SOC and reading 93.5%. Contacted tech line xxxxxx and wrote diagnostics and sent screenshot of BMS data, and tech line advised to replace the high voltage battery and BMS module due to SOH below 70%".

It's interesting to see that the dealerships' own calculation was 66.5% SOH (which of course triggered the battery capacity warranty). This was very similar to my own calculations of 68.59% SOH based on EVSE charging data. However, Torque's estimated SOH values of 79.4% and later on, 90.2% were widely incorrect in my situation. In the meantime, the range now on the GOM indicates 98 miles at 100%:



I'll follow up later with Torque readings and a range test after driving. As far as my old battery pack, I do hope Kia will recycle it properly or put it into other good uses. I wish they had let me kept it somehow, as 18 kWh (66.5% SOH) usable would've been still great for a home battery backup solution.

As promised, here is the Torque info from the new battery pack:



Interesting to see that the SOH reads 110% at 29.7 kWh. Is Kia giving us "free" 2.7 kWh before the SOH reading kicks in? If so, then the battery capacity warranty is really at 80% SOH, since we'd have to lose 10.8 kWh to get to 18.9 kWh (70% SOH of 27 kWh). Regardless, I will do a drain/range test to see if 29.7 kWh is actually usable.

As an aside, looks like Torque is reading the Lost SOH / % wrong! It thinks that the battery has lost 10% of SOH instead of having it gained (which is a weird situation anyway). Should be easy to fix? -10% "SOH loss" instead of 10% when SOH is > 100%.

I did a drain/range test. Looks like the answer is "no" to my question earlier. I was able to drive a total of 81.2 miles to 1% SOC, at 80 mph (2.5 mi/kWh). The last 10 miles or so was at slower speed (55 mph). My SOH in Torque (110%) was unchanged after a full recharge to 100%.



EVSE recharging data:



Let's calculate the kWh actually used/recharged on the car's end:

Before driving:
CEC: 208 kwh
CED: 198 kwh

After driving:
CEC: 210 kwh
CED: 226 kwh

After charging:
CEC: 237 kwh
CED: 226 kwh

kWh used while driving
= total kWh discharged - total kWh charged (regen)
= (226 kWh - 198 kWh) - (210 kWh - 208 kWh)
= 28 kWh - 2 kWh
= 26 kWh

kWh used to charge
= CEC (after charging) - CEC (before charging).
= 237 kWh - 210 kWh
= 27 kwh

So it looks like I only used 26 kWh for driving 100% -> 1% SOC, which makes sense as the SOC was still at 1% (SOC BMS was at 2.5%). So there was probably still at least 1 kWh lingering in the last 1%. The car took 27 kWh to recharge from level 2 AC (which I'm assuming includes internal charging losses?). Finally, comparing this to the EVSE data shows that the OBC is 27/29.42 = 91.8% efficient.

Thus, even though the SOH shows as 110% (29.7 kWh usable), only 27 kWh is actually usable from a brand new battery pack.

-
Hi Kish. Thanks for posting the data. All looks good. It is the total energy that you can put into, and get out of the pack that confirms the usable capacity in your battery is 27kWh. This is the same as a brand new 2015 Soul EV. Hence you do not have a newer 2018 battery pack with 30kWh usable capacity.

The torque data showing zero deterioration may or may not be accurate. A BMS reset would set this to zero, just like the real zero on a brand new pack.

So there are a few possibilities about your new battery pack.

1. It is a brand new pack identical to the original. These had a total capacity of 30.5kWh, and a usable capacity of 27kWh.

2. It is a new pack that has been sitting in a warehouse for a year or more. It may have lost some capacity from the total but it still has a usable capacity of 27kWh. In a month or two there may be a sudden drop in the deterioration numbers as the BMS calibrates. You still have the 27kWh usable capacity and SOH=100%, but during the time it was stored it lost some capacity from the 10% buffer. This is often the case with new cars that have been sitting on dealer lots for a year or more.

3. It is a refurbished pack. They have found 'good' modules from broken batteries and put them back into a 'new' battery. I think this is unlikely because the labor cost involved is probably higher than just giving you a new battery.

4. It is a brand new pack based on the new 30kWh design that has been software limited to 27kWh. This would be the best for you. You still only have a usable capacity of 27kWh. But that will last a lot longer than usual because you will have a much larger buffer. I think this is currently less likely. But in future they will surely run out of the original packs, and they will have to do something like this.

Please keep reporting it will be interesting to have more data on this pack.

I will answer the questions about the Torque display showing 110% on the Torque thread. - Setting up Torque to show BMS data

There's an interesting article by a Leaf owner about his battery replacement here (click to download a pdf file) - http://bit.ly/2A2tJDc

Ah yes, I didn't at all consider the fact that the Torque PIDs for SOH could be providing the wrong info, since you guys are calculating this from the deterioration numbers from the BMS. Thus showing 110% instead of 100% SOH. Now that you mention it, it's definitely a similar effect to the usual BMS reset that the dealership does.

I also agree that scenario #4 is unlikely, because the 30 kWh usable Soul EV's wasn't available in the U.S. until recently, in the MY 2018. Thus it doesn't make sense for Kia to "waste" one of those newer battery packs in an older model 2015. Scenario #1-3 is likely, with #2 or #3 most likely. I will surely be reporting SOH readings as time passes. Let me know if there's other PID values you're interested in.

I DCFC'd my car for the first time today with the new battery pack. Three new pieces of information to report:

1) My car DCFC'd straight up to 94% without stopping! :shock: The 2015 Soul EVs are software-coded to stop charging at 84%, and we could manually initiate a 2nd charge which will bring it up to 94%, after then it will stop again. This behavior is also clearly mentioned in the 2015 Soul EV user manual. I verified this behavior multiple times with my previous (original?) battery pack, and it always stopped DCFC'ing at 84%.

I know that 2016+ Kia Soul EVs removed this original limitation of stopping at 84% and only stopped once charge has reached 94%. This could potentially mean that this new battery pack replacement that I've received was originally intended to go into a 2016+ Soul EV. Either that, or the new BMS that was installed is again from a 2016+ Soul EV so it was software coded to stop at 94% instead of 84%. (or both!)

As an aside, on 125A 50kW DCFCers, I notice that the charge from 0 - 83% is quite fast (~45kW), and it slows down from 83%+. Especially 90% - 94% is quite slow (< 10kW).

2) I took a 300+ mile road trip which required driving fast at highway speeds, as well as multiple DCFCs back-to-back (4 in total). Ambient temp was in the low - mid 70s. After the 3rd DCFC, I noticed that the battery temp was getting pretty warm at 100F. After the 4th DCFC, the battery temp was over 120F! :!:



The Torque reading above was directly after the 4th DCFC. You can see that the battery inlet temp (cabin temp) is at 78.8F, but the battery temp is at 122F, quite hot! I had the car turned off during the DCFC sessions, so the A/C was not running to help cool the cabin/battery. The battery cooling fan was pretty much spinning at full speed (95%?).

3) After the 4th DCFC session, I also noticed while driving on the highway that the throttle response wasn't that good. I quickly checked Torque and noticed that the motor output power is being limited to 63 kW! (Down from 90 kW usually).



The Torque reading above shows that both the motor output power and regen have both been limited to 63 kW. I'm assuming the BMS is doing this to protect the battery since it's so hot? Later in the day when I allowed the car to sit for a few hours, the battery temp had dropped down to 100F and the motor output power was back up to 90 kW. The ambient temp by this time was down to low - mid 60s.

-
I have a 2015 Soul EV with the original BMS software, but updated OBC firmware. It quick charges to 83%. The blue lights do not come on.
There is much confusion about the BMS update - BMS software update
The BMS software update is known as Service Action 297
Some people say it now quick charges to 94% in one session, some say it remains the same. Some people say the blue lights do come on.
Some people think that resetting the BMS magically increases battery capacity.
Perhaps this new firmware has a new algorithm for calculating SOH? Perhaps not?



Maybe there is more than one update available, or maybe it depends on a combination of upgrades. Anyway during the battery replacement many of the ECUs in your car must have been upgraded. Without knowing all the ROM IDs it would be hard to know exactly.

------------------------------------

Your issue regarding power reduction at extreme temperatures is new to this forum.
The issue of temperature / fast charging / very high ambient temp was discussed here - Palm Springs Summer TMS Torture Test!
Also see - TMS Behavior - for details on how the cooling fan works.

The previous known values for Fan Status, Fan Feedback and the temps at which they come on are

We can now add level 8. Your car has the hottest battery so far recorded!
We can see that at 50C the fan speed is level 8.

------------------------------------

In the Torque screenshot what are the 3 PID codes that are not found? Which Torque dashboard are you using?

Ah, I figured something out!

I noticed that Torque "charge power" reading always stayed at 0.0 kW (as well as charge volts/amps/output volts). Even during DCFC sessions, these values stayed at 0.0. Then today, while charging at level 2 AC, I was watching it and noticed that "charge amps" was going to 360 A or so, which is totally incorrect. So I figured this must have been switched with the voltage readings. Indeed I downloaded and swapped to the Kia_Soul_EV_OBC2016_data instead of the 2015 one I had previously installed and it worked! Now I'm getting correct readings for level 2 AC at 6.7 kW (as well as proper readings for amps/volts). This is a confirmation that indeed that the new BMS they installed must have been a 2016+ version, which in conjunction with new OBC firmware also explains why my DCFC session ends at 94%? By the way, my 3 blue lights on the dash does stay on while DCFC'ing.

As far as the dash I'm using, I thought it was pemessier_1080x1920_xxhdpi.dash, but looks like mine is slightly different. The layout and the values are identical, but some of the labels are different. Regardless, I loaded in the pemessier dash and the 3 missing PIDs are still not available. I also cleared and reloaded all 8 relevant PID files. What's supposed to be in those boxes?

Edit:
Looks like they're supposed be "OBC heatsink", "OBC inside", and "OBC Water". Yeah, not sure why these aren't being loaded (neither with 2015 or 2016 OBC PIDs).

Edit 2:
These values do show up using the other dash, SoulEV2016_720x1280_xhdpi.dash. Looks like the pemessier dash just needs to be edited to load in the proper PIDs?

Some piece of bad news. To quote you from before:

JejuSoul
According to the lab data 45C (113F) is the threshold when major deterioration starts to occur.

Click to expand...

Take a look at this:





My SOH % has dropped down from 110% (29.7 kWh) to 109.5% (29.6 kWh). "Max kWh" has dropped down to 28.1 kWh (from 28.2 kWh). We know that these readings aren't 100% accurate, but this does mean some deterioration has occurred, and full 27 kWh is probably no longer usable. Furthermore, the "Max Deterioration" has jumped up to 1% on cell #2. Earlier, the brand new battery pack had "Max Det" at 0% for all cells.

This deterioration is probably due to a combination of 1) back-to-back DCFC (at 50kW) + 2) highspeed driving on highway (80mph) + 3) loaded car with passengers (and luggage) + 4) moderate ambient temps (75F). I don't even want to know what would've happened if ambient was 90-100F+.
Needless to say, I'll probably take my other EV for longer roadtrips and avoid DCFC'ing the Soul EV back-to-back. I'll probably keep the Soul EV as an "around town" car, especially since the boxy shape isn't great for high speed driving (2.5 mi/kWh). It would be a shame to kill the battery pack right after I got a new one!

@Kish Thankyou for making the time to share this very interesting reading

A couple of things spring to mind
I wonder if the previous owner overheated the battery regularly as well by doing regular long trips, the cumulative result may have been the degradation you inherited.

I also wonder if yours is a one off, maybe the fan etc in the boot is faulty, air flow restricted etc?

Kish said:

Some piece of bad news. To quote you from before:

According to the lab data 45C (113F) is the threshold when major deterioration starts to occur.

Click to expand...

Take a look at this:

My SOH % has dropped down from 110% (29.7 kWh) to 109.5% (29.6 kWh). "Max kWh" has dropped down to 28.1 kWh (from 28.2 kWh). We know that these readings aren't 100% accurate, but this does mean some deterioration has occurred, and full 27 kWh is probably no longer usable. Furthermore, the "Max Deterioration" has jumped up to 1% on cell #2. Earlier, the brand new battery pack had "Max Det" at 0% for all cells.

Click to expand...

I do not understand the "Lost SOH %" and "Lost SOH kWh". If would expect this to be 0.5% and 0.1 kWh (27 kWh = 100%, 29.7 kWh = 110%).
Or do you have a Soul 2018 with a larger batterypack or mixing up the Torque PID codes for Soul 2018 version?