Fixing a low voltage cellpair in the HV battery without battery removal.

[Frankleaf14]

So I recently bought a 2018 SV with 106+k miles and discovered it had a low voltage cellpair which was causing a delta-V of around 250 up to 300mV. In practice, this has meant that the car was only usable from about 100%, down to 80% SOC. Driving it below that %, the car lost SOC rapidly on any hill climb then the SOC quickly rose again, going downhill. This meant it only had a maximum reliable range of about 30 miles. The car is showing 12 bars available and a range 'estimate' of 150 miles at 100% displayed SOC. Leafspy shows it has an SOH of around 83% but with cellpair #29 much lower than the cellpack average.

The conventional way to deal with this is to remove the HV battery and replace or recharge (depending on the problem) the offending module - easier said than done and/or an expensive garage repair. I didn't pay much for the car so have been trying other methods to resolve this. Since about the beginning of November, I've been leaving the car in READY mode 24/7, to allow the car's LBC/BMS do its weak cellpair balancing and this is producing good results. I've been noting values such as delta-V since 8Nov on a daily basis and have seen it decrease by about 66 mV in that time. This has produced a corresponding increase in my reliable driving range and the car SOC can now go to 68% before the craziness begins.
The delta-V is falling roughly between 2 and 3 mV per day with this method.

With the results so far, I expect(hope!) to see the battery behave something like a 'regular' 40kWh battery with 83% SOH, sometime in the late Jan/late Feb '25 time period. There's more I could say but this is my first post and it's already an essay. If anyone is interested in this topic, I'm prepared to discuss/answer their questions, otherwise I will just periodically post my progress.

 

[cornbinder89]

I'll be interested to watch how well you can get the cells to balance and how long it lasts.
Remember, something happened to allow that cell pair to go out of balance and nothing you are doing addresses the original cause.
I hope, you end up with a usable car that meets your needs, but an skeptical that the problem will be resolved long term as the root cause was never addressed.

 

[Gusset]

I've been leaving the car in READY mode 24/7, to allow the car's LBC/BMS do its weak cellpair balancing and this is producing good results. I've been noting values such as delta-V since 8Nov on a daily basis and have seen it decrease by about 66 mV in that time. This has produced a corresponding increase in my reliable driving range and the car SOC can now go to 68% before the craziness begins.

What does it mean to leave the car in READY mode 24/7? Are you putting it park, setting the parking brake, leaving the key fob in the car, and walking away without pressing the START button (that would normally turn it off)?

Interesting result. I'd be further interested in the hypothesis as to what the BMS is doing when left in this state for long periods to lead to the changes you're describing.

"Watch"ing this thread!

 

[Frankleaf14]

What does it mean to leave the car in READY mode 24/7? Are you putting it park, setting the parking brake, leaving the key fob in the car, and walking away without pressing the START button (that would normally turn it off)?

Interesting result. I'd be further interested in the hypothesis as to what the BMS is doing when left in this state for long periods to lead to the changes you're describing.

"Watch"ing this thread!

Yes, exactly that. I have the car in my garage, so just put it in READY mode (foot on brake, press start), then leave the key in the car and get on with my day! That way, the bms is now balancing 24 hours a day, instead of just when driving for an hour or so, then charging maybe an hour or so on an L2 charger, then OFF for about 22 hours as we usually do.
Now this will be a long process of several months, so it pays to keep the power consumption as low as possible, so I make sure there's nothing running in the car, like fan, lights etc. I've tried a bit harder by turning off the center display and dimming the drivers display and I pulled the fuses on the daytime running lamps. Leafspy now says it's consuming between 75W and occasionally peaks to 200W. That's a lot more than just the balancing which I've read, is about 40mW per cellpair when shunted, but the best I've been able to do for the moment.

 

[Frankleaf14]

I'll be interested to watch how well you can get the cells to balance and how long it lasts.
Remember, something happened to allow that cell pair to go out of balance and nothing you are doing addresses the original cause.
I hope, you end up with a usable car that meets your needs, but an skeptical that the problem will be resolved long term as the root cause was never addressed.

"Remember, something happened to allow that cell pair to go out of balance and nothing you are doing addresses the original cause."

Good point. Here's my thinking....when I got the car, I was using it within its limits (it's our second car and will be lightly used) and exploring the crazy behavior sub 80% SOC. Leafspy showed it only had one fast charge in 5 years (I guess the original owner was just checking it worked!). It has high miles but the SOH is good, so I don't think the battery has suffered much abuse. Now for the few months I've had it, the delta-V has been high but didn't get worse. Also, when I charge it, cellpair#29 moves up in lockstep with the other cells but maintains the same delta. That to me says it's capacity is fine but how did it get so far adrift from the others? From what I've seen, this drift has been happening slowly (perhaps over the 5 years?) due to...what? maybe a slightly higher internal resistance? So under heavy hillclimbs or fast acceleration, it was losing maybe just tens of microvolts more than the other cellpairs and just too much for the weak bms to deal with.
That's my theory, so my hope is that once I can reduce the delta enough, it will only drift back down very slowly. But that's another good question....what will I need to do to maintain a good balance? Will I have to 'rinse and repeat' this process, once a week? a month? a year? Hopefully the latter but it remains an open question. I've rambled enough.

 

[cornbinder89]

Not rambling and I am truly interested it what you can achieve with balance alone.
It will be a learning curve and I hope you'll continue to post (both good and bad) so we can all learn.
I was just trying to say, that I wouldn't get my hopes to far up that it will be cured.
The main thing with these car is expectations, if yours match what the car can do, then all is good.
Please continue to post the "good, bad and ugly" of your trial.

 

[RPaulThomas]

I've been leaving the car in READY mode 24/7, to allow the car's LBC/BMS do its weak cellpair balancing

Couple of questions please:

1) What SoC are you leaving it at?

2) Have you experimented with differing SoC, such as 100% or near the 'cliff'?

3) Have you tried leaving it with the power cable and READY mode both on?

 

[nlspace]

3) Have you tried leaving it with the power cable and READY mode both on?

Not possible to do this.

 

[Lucky]

Leafs 2018 and 2019 with this kind of battery state are usually qualified to replace the cells/packs. Not some but all of them. You can replace just a pack or two, but this job is labour intensive and costly just to realize after a couple of months, that the next cell sagged. I went quite deep in this and already worked on a car with around 60k miles. Since your car has 106k miles it is already bad, specially if it was driven hard and quick charged in between.
If you have time, check my video and the follow ups. Best wishes but it doesn't look good.

 

[MNLeaf2021]

Leafs 2018 and 2019 with this kind of battery state are usually qualified to replace the cells/packs. Not some but all of them. You can replace just a pack or two, but this job is labour intensive and costly just to realize after a couple of months, that the next cell sagged. I went quite deep in this and already worked on a car with around 60k miles. Since your car has 106k miles it is already bad, specially if it was driven hard and quick charged in between.
If you have time, check my video and the follow ups. Best wishes but it doesn't look good.

At end of video you state the battery is no good. Is there not a way to recover stability in low end of charge by balancing or is there just too many cells in bad condition that they would need to be replaced? I am just learning of this condition as my 2021 LEAF SV+ is getting very unstable once below 40% charge. Another question, is the LEAFSPY data of 40% remaining not valid? It is quite a bit off your 25% battery state in the car. When you got down to 0% on car percentage meter would the car actually shut off even though LEAFSPY shows there is adequate SOC? Thanks for showing this situation.

 

[Stanton]

I commend you for thinking "outside the box"...but in the end nothing will prevent the bad/weak cell pair from continuing to deteriorate. You may see some near-term improvement, but once you reach low SOC/temps (and especially a combination of both), you will experience the same irregular pack behavior; in addition, I'm not sure I would want to leave my car in "ready" mode for days/weeks(!) on end.
Sorry to rain on your parade, but I've been down this path before with my (replacement) 40kWh pack...and have had no issues after replacing the weak cell/module.

 

[Frankleaf14]

A lot of good posts here. I'll try one or two posts to cover their content - less work than individual replies

So my evolution to this point went like this....My first reaction was 'yikes, I need a new battery' to 'hang-on, this is just one low cellpair. I can just replace a module' to 'hang-on again, this cellpair is just a quarter volt adrift and stays there. I just need to charge it up to the others'. But even that last option is as difficult as the others in terms of the physical work required. So, I thought about a method to somehow run the bms balancing as long as possible to reduce the delta-V but the battery neither charging nor discharging. My first idea was to charge with a low power L1 charger and consume as much power as possible (lights, fan AC, heater etc etc). I persevered with this a few days then gave-up for 3 reasons. First, it was way too 'hands-on', constantly checking to turn the AC or heater back-on etc. Second, too much power consumption - it was over 24kWh when active for 24 hours. Third too stressful on all the car's accessories. Not good. Back to the drawing-board. Then I saw a post on the forum where the poster suggested the 24/7 READY idea in answer to someone with a similar delta-V to my battery, though he thought it impractical as it could take maybe a year to resolve. I decided to give it a go and to my surprise it seems to be pulling down the delta-V faster than expected.

Anyway, here is more detail on my particular approach. I am focused on improving the delta-V at the low end of battery charge, where the car SOC is in the teens% (but SPY SOC% says low 50s%) . I have taken to charging the SPY SOC to 61% then putting the car in READY and beginning the discharge back to the teens%. BTW when I started this approach about a month ago, the car SOC was only reaching up to the mid-30s% but has recently been as high as 57%.....almost in agreement with the SPY's 61%. I then note my delta-V after the discharge and repeat the process, charging back to SPY 61%. I like this approach because it's only consuming about 2kWh for the whole 24 hours in READY and I'm only doing about 2 charges per day, one in the late evening and the other around noon times. The charges only took about 20ish minutes but are now taking longer as the low cell improves. The approach is very stress-free for the car as you want nothing ON, plus the battery is really only being cycled up/down from about low 50s% to 61%.

This post is way too long, so I'll finish here. Will try to post tomorrow with my latest numbers which continue in the right direction. I'd also like to give my observations and thoughts on the bms balancing algorithm - at least what I'm seeing on this particular 2018 SV.

 

[Frankleaf14]

Leafs 2018 and 2019 with this kind of battery state are usually qualified to replace the cells/packs. Not some but all of them. You can replace just a pack or two, but this job is labour intensive and costly just to realize after a couple of months, that the next cell sagged. I went quite deep in this and already worked on a car with around 60k miles. Since your car has 106k miles it is already bad, specially if it was driven hard and quick charged in between.
If you have time, check my video and the follow ups. Best wishes but it doesn't look good.

I appreciate your reply and as always with these matters, YMMV. We need to make sure we are comparing apples to apples. I had a quick look at your video and can see at 4min 9sec, there are many problematic cellpairs in the rear pack (roughly #s 5 to 40 on your leafspy). I also see almost 1000 fast charges, compared to my single fast charge - so that might be the reason for your general weakness, though it's interesting how easily one can be fooled with the apparent good balance at the start of you video. To me, it looks like time for a new/recon battery. I hope you resolved jt.

 

[leafy greens]

Something that might also be worth a try is I have heard that they Ballance better at low and high state of charge, I Balanced my 40kwh pack with my evse set to 6 amps my theory was the longer it was charging at that high soc the more balancing would be effective, my pack wasn't too far out but it did level the field to 8mv

 

[Stanton]

Something that might also be worth a try is I have heard that they Balance better at low and high state of charge, I Balanced my 40kwh pack with my evse set to 6 amps my theory was the longer it was charging at that high soc the more balancing would be effective, my pack wasn't too far out but it did level the field to 8mv

You can do all the balancing you want, but it won't "fix" a weak/bad cell; eventually that module will have to be replaced.

 

[cornbinder89]

If, in fact it is a weak cell. It is also known that when replacing cells they must be charged to within a few mv of the other in the pack.
We all, including the OP suspect a bad cell, but he is willing to put in the time to try and see if he can bring it back more in line with the others through the time consuming balancing protocol.
Do I think it will succeed to a balanced pack? not likely, but as an experiment, if he is willing to put in the time, understands the chances of success, and can use the car as is anyway, I see no reason to discourage the attempt. Nothing but time to loose and everything to gain.

 

[leafy greens]

You can do all the balancing you want, but it won't "fix" a weak/bad cell; eventually that module will have to be replaced.

Absolutely, but for some it is worth trying to limp a bit more life out before a major repair

 

[Frankleaf14]

LATEST DELTA-V RESULTS

First, here's how I get my delta-V numbers. I've been recording the delta-V ( the big mV number on top right of leafspy) once the battery discharges to the high teens% on the car SOC, so staying just above first battery warning. This number varies up and down by a few mV when you watch it over a minute or two, so I note the high and low values as 'delta-V range' and I have noted these values since 8Nov. Also, since 14Nov, I've noted the delta-V of the low cellpair only, without including the high cellpair, so I get this by subtracting the low cellpair voltage (the left of the 3 voltages under the bar graph) from the cellpack average voltage (middle of the 3 voltages). This number also varies by several mV but I just record the smallest value.

So on 8Nov, my delta-V range was 230 to 237mV and on 14 Dec the range was 154 to 160mV. Like with like, that's a reduction of 76mV in 36 days - just over 2mV per day. The numbers for the delta-V of the low cellpair only, on 14Nov this was 196mV and 14Dec (30 day range) 136mV....right on 2mV per day.

All things remaining equal, if this reduction continues, my low cellpair delta-V will be 50mV in another 43 days, and around 20 or 30 mV from the lowest of the other cellpairs, so will that be good enough for close to normal behavior? .....watch this space!

 

[Chad E. Mo]

I'm following this with great interest.

My leafspy cell diff has jumped to ~140mV after a handful of deep depletions (to <8kWh) and L3 charges.

The concept of slower L2 charge rates to allow more balancing is intriguing, I'll also give that a try.

 

[nlspace]

At the ~40mW daily burn rate i estimate it will take about 376 days to balance the cells.

will that be good enough for close to normal behavior?

Unfortunately no because the weak cells will immediately diverge after the first full charge-discharge cycle. The balancing is only bringing down the healthy cells to the weak cells' level.