Bad cell pair on my 2012???

[Computerizer]

I've noticed (by driving) that my 2012 doesn't have the capacity it used to (last winter). Using the Leaf Spy app, I took this screenshot of the status of the cells and such:

(Full size: http://imgur.com/vbqTmhu)

Things to notice:

• This is after an 80% charge, not 100%
• One cell pair has lower voltage than the others. It is consistent every time I check my status with the app
• I have recently done a couple of deep cycles to try to make it better -- down as low as 15% and up to 100%
• After only 10 months, I'm at 93.71% capacity and 86.09% on the mysterious health-o-meter
• Only 13 DCQCs, and the temperature bars have never been over 7.

Any idea what might have happened here and how to make it better? I've always practiced good battery management, including mostly charging to 80% but every few weeks or so doing a 100% charge; only doing DCQC a few times and only when it's not hot; when left alone for 2 weeks, left it sitting at 50% and not plugged in.

Thanks,
Tyrel

 

[TomT]

I'd try a few more 100% overnight charges... If it stays like that then, yes, you may have a weak cell... 27 Mv is not really that much though... A dealer would have to do the more intensive test at low battery state to confirm the problem and get you a cell replacement...

 

[essaunders]

I have one of those on my 2012 too.


<edited to add a picture I found>

never been over 6 bars temp, no QC, 80% charge normal.

I'm not holding out much hope the dealer would find much. Does anyone have guidance on how much voltage delta would warrant action by Nissan? This is only 25mV from average to low.

 

[RegGuheert]

I'd try a few more 100% overnight charges... If it stays like that then, yes, you likely have a weak cell... A dealer would then have to do the more intensive test at low battery state to confirm the problem and get you a cell replacement...

+1

But I will note that this pack is top-balanced, meaning that you should not see too many differences when the pack is fully charged. My recommendation is to charge from fairly empty to full using L1, perhaps during the weekend. This allows the balancing to do its work. You should then see very similar cell-pair voltages. Then take screenshots at various SOCs (maybe every 10% or 20%) as you discharge the car and post them here. That should give us a good idea if this one cell simply has a lower capacity. It is certainly possible with two cells in parallel to have one of the two go bad and therefore to lose about half of the capacity of your pack.

 

[ebill3]

Have similar condition on my 2011. I'm not very concerned. The graph makes it look really bad, but in reality not much of a difference. In the OP's case, min to max difference is only 27 mv, my min to max difference is 31 mv.

 

[RegGuheert]

Have similar condition on my 2011. I'm not very concerned. The graph makes it look really bad, but in reality not much of a difference. In the OP's case, min to max difference is only 27 mv, my min to max difference is 31 mv.

Is your number when charged or discharged?

OP may typically charge to 80%, which can cause some imbalances, but I expect imbalances should be below about 20 mV at the top of the charge (100%) when fully balanced. L1 gives you the most balancing per charge cycle (in the 2011/2012s).

But we really won't know if there is a bad cell until the pack is first balanced and then discharged. Graphs at both ends are most useful.

 

[palmermd]

I've noticed (by driving) that my 2012 doesn't have the capacity it used to (last winter). Using the Leaf Spy app, I took this screenshot of the status of the cells and such:

Things to notice:

• This is after an 80% charge, not 100%
• One cell pair has lower voltage than the others. It is consistent every time I check my status with the app
• I have recently done a couple of deep cycles to try to make it better -- down as low as 15% and up to 100%
• After only 10 months, I'm at 93.71% capacity and 86.09% on the mysterious health-o-meter
• Only 13 DCQCs, and the temperature bars have never been over 7.

Any idea what might have happened here and how to make it better? I've always practiced good battery management, including mostly charging to 80% but every few weeks or so doing a 100% charge; only doing DCQC a few times and only when it's not hot; when left alone for 2 weeks, left it sitting at 50% and not plugged in.

Thanks,
Tyrel

You do appear to have one cell pair lower than the others, but it by no means out of specification.

http://www.mynissanleaf.com/viewtopic.php?p=263912#p263912" onclick="window.open(this.href);return false;

You are showing a 27mV spread, and it would have to be over 150mV at that voltage to be out of specification.

 

[RegGuheert]

You do appear to have one cell pair lower than the others, but it by no means out of specification.

http://www.mynissanleaf.com/viewtopic.php?p=263912#p263912" onclick="window.open(this.href);return false;

You are showing a 27mV spread, and it would have to be over 150mV at that voltage to be out of specification.

Specification is only for the condition when the cell-pair voltages are below 3.612V (IIRC), or around two charge bars or below.

 

[ebill3]

Have similar condition on my 2011. I'm not very concerned. The graph makes it look really bad, but in reality not much of a difference. In the OP's case, min to max difference is only 27 mv, my min to max difference is 31 mv.

Is your number when charged or discharged?

OP may typically charge to 80%, which can cause some imbalances, but I expect imbalances should be below about 20 mV at the top of the charge (100%) when fully balanced. L1 gives you the most balancing per charge cycle (in the 2011/2012s).

But we really won't know if there is a bad cell until the pack is first balanced and then discharged. Graphs at both ends are most useful.

My imbalance of 31 mv is at 80% charge. Next time I charge to 100%, we'll see what happens.

The business about L1 charging providing a better balance is news to me. Can you provide some evidence of that? There is a similar argument being made in a Tesla Model S thread, but it doesn't seem to be substantiated.

 

[RegGuheert]

The business about L1 charging providing a better balance is news to me. Can you provide some evidence of that? There is a similar argument being made in a Tesla Model S thread, but it doesn't seem to be substantiated.

It's mostly empirical. Several of us have noticed that if we have charged only to 80% for some time, then it takes multiple L2 charges to full to restore balance. An L1 charge to full generally can fully balance the pack in a single charge. The idea is that at L2 rates, the shunts which do the balancing can only bypass about 10% of the current going into the cell (assuming it doesn't taper the charge, which it actually does). At L1, the shunts can bypass about 30% of the current around a cell with no tapering, giving a better opportunity for balancing.

The question is how much tapering occurs in the 2011/2012. While I think it can taper down to L1 rates or even below, I think it ends up being a "too little too late" type of thing that results in at least one cell getting fully charged before balancing is complete.

It seems the 2013s have fully addressed this issue and it appears L2 can and does taper the charging enough to achieve full balancing each time the battery is fully charged. (With owners often complaining about the extra hour of charging that occurs at the end.)

 

[palmermd]

You do appear to have one cell pair lower than the others, but it by no means out of specification.

http://www.mynissanleaf.com/viewtopic.php?p=263912#p263912" onclick="window.open(this.href);return false;

You are showing a 27mV spread, and it would have to be over 150mV at that voltage to be out of specification.

Specification is only for the condition when the cell-pair voltages are below 3.612V (IIRC), or around two charge bars or below.

That is not correct. Follow the link and the manual and it indicates you can check the spread at any time. The tolerance widens as the pack is depleted, but it can be checked at any voltage/SOC.

 

[RegGuheert]

That is not correct. Follow the link and the manual and it indicates you can check the spread at any time. The tolerance widens as the pack is depleted, but it can be checked at any voltage/SOC.

Sure, you can measure the spread at any charge level. But the test of whether cell-voltage differences are out of specification is done when the minimum cell-pair voltage is 3.712V or below, per page EVB-65 of the April 2011 LEAF Service Manual. The only reason to look at the cell-voltage differences when the pack is fully charged is to see if the cells are balanced properly. In this case, I would say they could use a bit more balancing.

Looking at the cell-voltage pairs when a balanced pack is mostly discharged is how you detect if one or more cell-pairs is particularly low on capacity.

 

[palmermd]

Sure, you can measure the spread at any charge level. But the test of whether cell-voltage differences are out of specification is done when the minimum cell-pair voltage is 3.712V or below, per page EVB-65 of the April 2011 LEAF Service Manual.

OK, then your manual is different than mine. The Nov 2010 manual indicates you can check it any time, but the "best" place to check it is when it is at a low state of charge. I think we agree here, it is just a matter of semantics and which manual we choose to read. In any event, my pack is far worse than the one we are discussing. I've got a 35-45mV difference at 80% and it get up to about 50-60mV when discharged to near LBW. 27mV is not anything to worry about unless it grows to something over 150mV when it is discharged.

 

[smkettner]

I thought for best balance was to use L1 and no timers. Leave on charge 8 to 12 hours after it reaches 100%.
L1 would be slower and allow more time to do any balance.

Still might not help.

 

[RegGuheert]

Still might not help.

It should balance unless there is a significantly higher resistance in that one cell-pair. In that case, then that would seem to indicate an actual problem.

 

[dgpcolorado]

I also have a low cell-pair. In my car #48 has always been 20-30 mV below the average since I started tracking voltage in early July:


It gets more noticeable in the LBW and VLBW range.

Charging to 100% has no affect at all on the low cell-pair. If the specs allow 230 mV cell-pair variation at 360 Volts, I guess I shouldn't worry about it. Based on the average of 3.824 Volts in the picture above I'd guess that the pack was at 367 Volts, but I didn't record the measured number for that.

 

[RegGuheert]

Charging to 100% has no affect at all on the low cell-pair.

Even when the battery is still at 100%?

If so, then that cell must have a higher internal resistance than the others.

 

[Computerizer]

Thanks for the insights so far. First, I will attemptto discharge to LBW or so and then charge with L1 to 100 to see if thathe lps. I will show the graphs at each end. I will then grab a graph every 10% or so on a discharge to see how it changes. I can tell you first though just from observation that this cell pair is always the lowest at all charge levels.
I recently did a long drive and logged the data so I could pull that out let you see my cell voltages over the entirecour se of the drive. I will post that tomorrow if I remember.

Tyrel

 

[dgpcolorado]

Even when the battery is still at 100%?

If so, then that cell must have a higher internal resistance than the others.

Yes, presumably so.

 

[gbarry42]

We can never be quite sure what Nissan might do differently, but in common practice, the low cell means that you have to stop using the battery before the rest of it discharges to where it normally would. This would show up either as
(1) You reach LBW sooner than your energy reading (GIDs or whatever) would lead you to believe, or
(2) You reach LBW as predicted, but you don't get anywhere near the last ~10 miles to turtle.
That behavior, of course, is ultimately determined by someone's computer program in the controller.

But when that one weak cell reaches the official lower limit (~3.6 V) the controller will declare that you're done for the day.

If that cell has a higher resistance than normal, the voltage should not be significantly lower in an open circuit measurement (which we probably can't do), and not much different at idle. It should drop a lot more under load (look at it when climbing a hill), and ... it should be higher when being charged. In fact, take a look while under full regen--we don't get many 30kW charging opportunities while standing still!