JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Analysis of a replacement battery.

Mon Mar 09, 2020 8:28 pm

-
I have a 2015 Soul EV that got a replacement battery under warranty.
This thread is for analysis of that battery.

In late July 2019 I bought a second 2015 Soul EV.The initial reading was 84.8% SOH. It had done 75,000km.

Image

By October 2019 the SOH on that car had dipped below 70% and a replacement battery had been authorized. In 3 months I had driven 6,500km The SOH dropped 15%. Shockingly rapid!

Image

In late December 2019 I had my replacement battery put in. In the 10 weeks I waited for the replacement I continued driving the car. I added another 5,500km. Oddly no further battery degradation was seen.
The new battery cells came in a large wooden crate. The battery case, the BMS, the wiring harness etc were all reused. The original cells were put into the crate and were sent back to SK Innovation.

Image

The new battery continues to have 96 cell pairs, but these are now the MY2018 version. An upgrade from the E375 to the E400.

Image
Last edited by JejuSoul on Mon Mar 09, 2020 10:23 pm, edited 7 times in total.
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Mon Mar 09, 2020 8:33 pm

-
Here's a couple of preliminary guesses at the pack's capacity. Based on the possible increase in the energy density of the cells from 37.5Ah to 40Ah. (Cell pairs in parallel have double this.)

A technical description of the E400 cell is here :- APR13 : Energy Storage - Advanced Battery Development
(for SK Innovation E400 see printed pages 35-37, on screen pages 13-15)

  • Validation of power density with over 600W/L and specific discharge and charge powers of over 460W/kg and 480W/kg
  • Development of 40Ah cell with 230Wh/L and specific energy 150Wh/kg
  • Successful life performance with over 2,000 cycles (expected)

-

I have no evidence what the new capacity of my battery pack is, nor how big the 'early-years' buffer may be.
But for now I will guess that updated BMS continues to work as before and that we have a 10% 'early-years' buffer.
The 'early-years' buffer will be lost first before any loss of usable capacity.
There is also a safety buffer that will never be lost.

I will also make a guess that the values for energy capacity have increased in the ratio 40 / 37.5
The original battery cells were 37.5Ah and the new ones are 40Ah. (Cell pairs in parallel have 75Ah and 80Ah)
These numbers suggest my new battery pack is a 28.8 kWh usable pack. That difference in size would increase my range by about 15km.
Over the last month I estimated the readings on the GOM to be about 20km higher than for a comparable 27kWh pack.
I'll wait for warmer weather in the spring to do more extensive testing.

------

I have just tested the capacity of the new battery and now believe it to be bigger than I guessed just above.
By charging from 9.5% to 100% I loaded 27kWh into the battery. (Using Torque CEC - Cumulative Energy Charged values.)
This suggest my battery is 29.83kWh. About 30kWh.

I am now guessing that instead of maintaining a 10% 'early-years' buffer, the new battery maintains a buffer of the same capacity as before.
In the 27kWh cars the 'early-years' buffer is 10%. That would be 2.7kWh.
The actual total capacity was 30.5kWh, hence the safety buffer is 0.8kWh.

Using the same guess that the values for energy capacity have increased in the ratio 40 / 37.5
The total capacity of the new battery is 32.5kWh.
If the buffers are kept the same size, (not same proportion) then the usable capacity is now 29kWh.

The two guesses for usable capacity 28.8 and 29kWh both seem too small given the GOM reading of 197km.
A third guess would be 30kWh, based on the capacity I just charged the car, which is also what the service techs suggested it would be.

-
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

ksoul2084
Posts: 93
Joined: Thu May 16, 2019 10:48 am

Re: Analysis of a replacement battery.

Tue Mar 10, 2020 11:31 am

JejuSoul wrote:-
...
The two guesses for usable capacity 28.8 and 29kWh both seem too small given the GOM reading of 197km.
A third guess would be 30kWh, based on the capacity I just charged the car, which is also what the service techs suggested it would be.

-


I would expect 30kWh, which is fantastic news!

Thanks for keeping us updated.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Wed Mar 11, 2020 9:50 pm

-
Using the GOM to estimate battery capacity is not the best method. But certainly the easiest.
I really should do proper range testing. That is something I want to do when the weather is warmer. (about 20C).
For now I'm just watching to see how far it goes over the next 3 or 4 days until I need to charge it again.
After 3 days use I have driven 105km, the SOC shows 50%, and the GOM shows 96km left.

My blue Soul EV just charged to 158km on a full battery.
In the last few days the climate, driving style, places driven are all roughly the same as the white car.
The blue Soul EV has an estimated SOH of 94%. Implying it has 25.38kWh of usable capacity left.
In charging from 13% to 100% I added 23.8kWh. This implies a 100% charge would take 27.25kWh.

The numbers do not seem reliable.
I think I can show why the energy required to charge a degraded battery is no longer easily calculated.
But I do not know why comparing the GOM figures on these two cars is so unreliable.

Anyway if I did use the comparison of the two GOM numbers.
The ratio of the battery capacities would be 197 / 158
If the blue Soul EV has 25.38kWh of usable capacity left, then the white soul EV has 31.65kWh.

This seems too large. A 2018 Soul EV has 30.0 kWh of usable capacity. It has 100 of the E400 cell-pairs.
If all the buffers were the same this would imply a replacement battery is 96/100 * 30 = 28.8kWh
This was my guess 1 above, which I rejected because the reported GOM figure seems higher than if this were the case.

---

Looking at the charging data from Torque.

Both cars drew 28.7Amps from the wall at 215V.
The Blue Soul EV took 3.85 hours to charge from 13% to 100%. The energy going into the OBC = 23.6kWh
The White Soul EV took 5.1 hours to charge from 9.5% to 100%. The energy going into the OBC = 31.6kWh

Image

The energy going into the battery must always be less than the energy coming from the wall.
Above is an instant value. For the calculations I use averages from the entire charging session.

For the The Blue Soul EV the overall energy efficiency was about 85%, hence the energy actually going into the battery = 20.0kWh
This is a much better number than the previous calculation. (It cannot be 23.8kWh - that number is greater than the amount pulled from the wall.)

The values in the Cumulative Energy Counter are not to be trusted in a degraded battery.


The EWP cooling system came on after about 90minutes in both cars and stayed on.

---

Repeating the calculation I did earlier but this time with a much lower value for the battery capacity of the older car.

The blue Soul EV has an estimated SOH of 94%. Implying it has 25.38kWh of usable capacity left.
But in charging from 13% to 100% I only added 20.0kWh. This implies a 100% charge would take 22.99kWh.

Also It seems that the range loss in winter for a degraded battery is more than the deterioration numbers suggest. A new battery suffers less at cooler temps. I'll have to wait a few months to see if the range on the blue car comes back faster than for the white car. That's my best guess at the moment.

So if I did use the comparison of the two GOM numbers.
The ratio of the battery capacities would be 197 / 158
If the blue Soul EV has 22.99kWh of usable capacity left, then the white soul EV has 28.66kWh

Again this value is close to my guess 1 above
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Sat Mar 14, 2020 6:07 pm

-
Some more guesses this time based on range data.

Over the last few days I drove 170km. The SOC displayed dropped from 100% to 20%. Average temp was about 10C.
The Energy Consumption displayed on the instrument cluster was 7.6 km/kWh. I had reset this at the start.
So another guess of usable battery capacity: (170/0.8) / 7.6 = 27.96kWh

The GOM showed 42km left. The GOM is guessing I have a range of 212km at this point.
From previous experience I know that the GOM underestimates the range.
But you only find this out by driving down to a low battery state, which I don't like doing.
Hence I am guessing that the actual range I can drive at these temps and speeds is greater than 212km.

-

Using Torque to measure the energy going into and out of the battery I use the formula.
Overall Energy Discharged - Energy Charged by Regen = 80% of the usable battery capacity.
(34.7 - 10.7) / 0.8 = 30.0kWh

The two times I have calculated the usable battery capacity using the actual numbers given by the BMS, once when charging and once when driving, the value I get is almost exactly 30.0kWh , which is also what the service techs suggested it would be.

If this is correct then the actual capacity of an E400 cell is 41.67Ah not 40Ah.
The actual usable battery capacity of a MY2018 Soul EV would be 31.25kWh not the 30kWh it has always been assumed to be.

Looking at the values for Cumulative Current I see that :
CCC increased from 1097.9 to 1172.0 when charging by 90.5%, which would give the cell's energy capacity as 40.93 Ah.
CDC - regen increased by 68.7Ah when driving 80%, , which would give the cell's energy capacity as 42.94 Ah.

-
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Sat Mar 14, 2020 6:12 pm

ksoul2084 wrote:...I would expect 30kWh, which is fantastic news!..

I agree.
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

IanL
Posts: 41
Joined: Tue Feb 18, 2020 12:57 pm
Location: Jersey, English Channel Islands

Re: Analysis of a replacement battery.

Mon Mar 16, 2020 3:18 pm

For what it's worth, https://www.evspecifications.com lists the "30kWh" model as 31.8 kWh, but useable capacity 30kWh. Perhaps there is 1.8 kWh "at the bottom" which is not used, i.e. the car shuts down at that point? I think it hinges on whether the SOC refers to usable capacity, or "entire" capacity. It should relate to useable, but one wonders how well the software engineers are briefed - errors abound.

I have given up on relying on Kia's documentation. I have a Spec sheet for the 2017 model (July 2017) which states the battery has 192 cells and 30kWh. I think they probably forgot to update it for the number of cells, but it illustrates how you have to take everything with a pinch of salt.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Mon Mar 16, 2020 6:42 pm

-
The value 31.8 kWh as the total capacity for the MY2018 battery is almost certainly fake.

I have moved the response to the last comment about evspecifications to the thread - What is the total capacity of the MY2018 battery.

-
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Mon Apr 27, 2020 9:37 pm

-
We have some data from a car in Switzerland that has recently had a replacement battery.

Image

Image

see :- Akku nun getauscht ehem. / Soul 2015 lädt bis 94% nach BMS Update / Antrag Akkutausch
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

JejuSoul
Posts: 1494
Joined: Wed Jul 08, 2015 6:47 am
Location: Jeju
Contact: Website

Re: Analysis of a replacement battery.

Wed Jul 01, 2020 7:07 pm

-
After 6 months my car with a replacement battery still shows 110% SOH. It has lost 0.1% from Max and 0.0% from Min.
This GOM reading is higher than when the car was new.

Image
The codes for using Torque Pro can be found by clicking the link in the website icon under my user name on the left.

Return to “Technical Discussions”