Replacing the single cells in the battery pack?

[JejuSoul]

A good video by TeslaBjorn replacing one bad cell in a degraded 24kWh Nissan Leaf.
He manages to improve the SOH from 67% to 72%.



and

 

[kojairvs]

Thanks JejuSoul, every little helps.
Messing with the temperature sensors is definitely not good idea.
For me it was always a case of cheating with the cell’s voltage, which is less dangerous only if we stop using bad cell full stop. If we still use a bad cell and tell the BMS that it has higher voltage that cell will drop under 2.5V and possibly overheat and present a danger.
Soon I’ll try my theory about disconnecting one cell from the circuit and close the circuit with the conductor and finding the best way of cheating the BMS by connecting the voltage sensor to the neighboring cell. I’ll post here my results.

 

[IanL]

Thanks JejuSoul, every little helps.
Messing with the temperature sensors is definitely not good idea.
For me it was always a case of cheating with the cell’s voltage, which is less dangerous only if we stop using bad cell full stop. If we still use a bad cell and tell the BMS that it has higher voltage that cell will drop under 2.5V and possibly overheat and present a danger.
Soon I’ll try my theory about disconnecting one cell from the circuit and close the circuit with the conductor and finding the best way of cheating the BMS by connecting the voltage sensor to the neighboring cell. I’ll post here my results.

Have you thought out how the BMS will behave with your crafty cheat when it balances cells?

 

[kojairvs]

Really good question IanL, which made me look back in to my OBD data before trying to make my own opinion on this issue.

My trip to work (50km) after all night plugged in with 240V/2kWh charging cable.

Knowing that at work I charge 8kWh from 40% to 100% tells me that my battery can take around 13kWh which is probably 50% off new battery capacity.

SOH 1 34%

SOH 2 72%

Range on meter showing 38km, but I know that I can make 75km without worrying.

After 3/4 hours of charging and rest of the night being plugged, BMS does what it can to balance but in the beginning of the trip this are the figures:

Cell 1 4.08V (worst cell)

Cell 2 3.95V (next to the worst cell)

Cell 96 3.95V (as bad as no 2)

Cells 3-95 3.90 (state of all “normal” cells)

My guess is that no 1 goes up to 4.2V and charging stops, and balancing until morning gets above figures.

After my mentioned trip this are the figures:

Cell 1 2.80V (worst cell)

Cell 2 3.51V (next to the worst cell)

Cell 96 3.57V (as bad as no 2)

Cells 3-95 3.65 (state of all “normal” cells)

After charging to 100% at work and leaving plugged for less time than in the night (probably between 6-8 hours) BMS managed to balance them to the similar values like in the morning before my trip.

My guess is that parallel with balancing redistribution goes around from higher voltage cells to the lower voltage. I know this is happening because after my car being parked whit-out charging, values are changing, specially no 1 gets some energy from no 2 and it’s voltage raises.

Now, if I disconnect no 1 completely (replace it with a connector/cable) and it’s voltage sensor cables connect to no 3 cell, charging will last longer because the worst cell now will be no 2 which is not drastically different to all others, and more kWh will be again stored in to whole battery.

To the question of what will happen with the balancing, I presume that BMS will have less work to do by not seeing bad cell anymore but two times values of no 3.

But main reason for doing this is to avoid having Turtle mode every time when no 1 goes under 2.5V which is a pain in the neck for me.

This is all my “theory” and hope, to be seen in practice soon.

I’m “lucky” that I have “only” 1 bad cell and this might work in my case.

 

[JejuSoul]

Elbilmek the Norwegian company seen fixing a Nissan Leaf above have stated :

Battery for Kia Soul, cannot be repaired, unfortunately... We can change the battery to used, the price about 60 000 NOK including VAT. The capacity will be about 90%. We offer a 1-year warranty or 40,000 km.

There is also a company in Hungary that is offering 44kWh replacement batteries.
Link to Facebook in Hungarian : e-Service
I know nothing about this company so until someone who has had this replacement done reviews it I wouldn't recommend using them.

 

[EVlite]

I'm really interested to see how you get on with this. I have same problem with 2014 Soul, #2 cell is worst on mine but a couple of others not great.
Best of luck and please keep posting your findings.

 

[Locuturus]

I wonder if it's possible to use the cells from classic Ioniq - supposedly they have more cobalt in them so they degrade less?

 

[JejuSoul]

There is still considerable debate on whether replacing single cells or modules in a degraded pack will work. (For more than a few months at least.) I now tend to believe it isn't a good option. The best bet if you are out of warranty is to find a complete pack from a scrapped car.

I recently saw this discussion about a Rich Rebuilds video.

Jason Hughes, who has performed the same repair before, has said that the new modules will result in an imbalance every time at some point. He says that unless the owner leaves the car alone for weeks at a time, it will likely not last a year.

Rich disagrees and claims he has one working for the past two years:

In terms of the repair. We’ve done quite a few ourselves. Using both methods. We’ve found that the customers typically prefer to spend the extra to replace the whole brick and attempt to keep as much range as possible. Our first one was 2 years ago so far so good.

 

[Dondahaka]

So Koyairvs did You manage cell replacemend? Dimensions from #38 post are correct to original cells in KIA Soul? I bought Kia Soul after accident, but it looks that battery is faulty too. Cell number 96 has big differences that others when charge/discharge. Other cells keeps differences in 0.04V.

From founded informations on this forum I assume that cell no 96 is in this brick on top?

 

[Locuturus]

@JejuSoul - hypothetically, given that manufacturers seem to make lifepo4 battery cells in pouch form as well - assuming we find something with similar voltages (or something to adjust voltage/amperages to whatever the car demands), and all the regular sensors are put back in - would it be possible to get better efficiency (prevent rapidgate+coldgate) on the car - maybe increase total capacity?

A lot of suppositions, but i recall seeing a video of someone replacing the cells in a Nissan Leaf with CATL cells/modules...

best thing would be to have a kia soul with a dead battery but everything else working, and tinker with it I suppose. The engineering hours would be insane...