How often to "Cell Balance"? Weekly? Monthly? Depends?

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ldallan

Well-known member
Joined
Jan 16, 2017
Messages
52
Location
Colorado Springs, CO
>> http://www.mynissanleaf.com/viewtopic.php?f=9&t=26877#p541248
Point of fact, you should be charging to 100% roughly once a week, just don't let it sit for too long at a high state of charge - drive it.
I'm wondering what "best practice" is for how often to "cell balance" with relatively newer Leafs from 2015 to the present, including those with Lizard batteries.

I've read the lengthy thread "Cell Balancing 101" http://www.mynissanleaf.com/viewtopic.php?f=9&t=7360&hilit=cell+balance
but that's an older thread from 2011 - 2012.

I've been "cell balancing" my 2015 'S' about monthly. 40k miles, about 14,000 miles per year. 2 or 3+ L2 charges per week (typically to about 85% to 90%). 12 health bars remaining.

I use the ChargePoint L2 station and their smart-phone app for cell balancing. That app provides a handy graph of kW/hr rate delivered every 5 minutes.

I can observe the charge rate ...
  • being about 6.5+ kW/hr under 90% SOC
  • dropping off gradually to 3 kW/hr or less at about 95% SOC,
  • then zero charge rate at 100% SOC, I interpret that as "unbalanced 100% SOC"
  • then three or so cycles of low charging rate to no charging
  • I interpret that as "cell balancing".
  • That takes 3 to 4 hours, depending on the starting SOC.
 
Monthly is good, obviously, although I'd consider it near the infrequent end of good. Once a week is probably the most often anyone would need to do it. Thanks for the data points.
 
On a related question, it sounds to me like "cell balancing" is wasting energy. When the charge is "tapering," has anyone guesstimated what % of electricity actually gets stored in the battery? I was considering programming my EVSE to stop charging when the taper starts, except on days when I want to balance.
 
specialgreen said:
On a related question, it sounds to me like "cell balancing" is wasting energy. When the charge is "tapering," has anyone guesstimated what % of electricity actually gets stored in the battery? I was considering programming my EVSE to stop charging when the taper starts.

I'm inclined to disagree. During balancing the charger goes off and on repeatedly, letting the cells absorb power slowly. Do you mean the cooling system for the charger running? That may waste a little, but A: I'm not sure it still runs while balancing, and B: balancing doesn't take long, usually. So if you count any extra loss as "waste" then I guess it does. If you figure keeping the pack as efficient as possible to be worth the extra power, then I wouldn't call it a waste...
 
I thought that during balancing, some of the cells are "shunted". What does that mean? I imagined that some current for/from that cell was being discharged through a resistor to ground so that those cells don't overcharge. Reading https://aip.scitation.org/doi/pdf/10.1063/1.4958529

when [the shunt resistor] is in the ON condition, current flows through the shunt resistor, which converts some current into heat and reduces the current which flows to the negative terminal. Reduction of the output voltage is the result"
[sic]

I don't know if the shunted energy is a trivial amount, or if shunting happens outside of charging (maybe it happens all the time).
 
Generalizing, I would do it every time your expected needs will be 50% of your driving range. I found that frequent cell balancing worked on 24 kwh LEAFs but saw no evidence of anything on 30 (occasionally saw some but not enough variance to say anything really happened) and 40 kwh LEAFs. In fact, my 40 kwh LEAF seems to run on different parameters almost daily.

I charged to 100% 3 days of 5 and each one showed different SOC, Kwh available, and GID count. The variance was well beyond normal fluctuations and did an "up down up"
 
I honestly don't know how much power the shunt resistors 'waste,' but given the amount of time they run, the fact that they are only converting power from the highest-SOC cells to heat, etc, I'd say it isn't much. Some balancing algorithms just switch the higher cells out of the circuit, allowing the lower-SOC cells to 'catch up'.
 
LeftieBiker said:
Monthly is good, obviously, although I'd consider it near the infrequent end of good. Once a week is probably the most often anyone would need to do it.
Thanks for the feedback (and also thx to the other replies)
Thanks for the data points.
And thanks for your time and effort put into moderating the forum. It's MUCH appreciated. I've learned a lot from you (and others).
 
Shunt resisters are 430 ohm across each set of cells. When shunting they draw V^2/R power, or roughly 16/430 or 40 mW each. There are 96 sets of cells, so if all but one was active, total power would be about 3.5 watts.
 
WetEV said:
Shunt resisters are 430 ohm across each set of cells. When shunting they draw V^2/R power, or roughly 16/430 or 40 mW each. There are 96 sets of cells, so if all but one was active, total power would be about 3.5 watts.

You did that in your head, right? :shock:
 
flydiver said:
WetEV said:
Shunt resisters are 430 ohm across each set of cells. When shunting they draw V^2/R power, or roughly 16/430 or 40 mW each. There are 96 sets of cells, so if all but one was active, total power would be about 3.5 watts.

You did that in your head, right? :shock:

Actually no. I did use a calculator. I did check the results in my head.
 
LeftieBiker said:
I honestly don't know how much power the shunt resistors 'waste,' but given the amount of time they run, the fact that they are only converting power from the highest-SOC cells to heat, etc, I'd say it isn't much. Some balancing algorithms just switch the higher cells out of the circuit, allowing the lower-SOC cells to 'catch up'.
I think it would be possible to figure out from observing ChargePoint data,

ChargePoint provides the amount of kWh that have been delivered. If you were watching the ChargePoint app and noticed that value the first time that the SOC reached 100% and the rate of charge dropped to 0%, that would be the delivered kWh before cell balancing commenced.

Then you could make a note of the total kWh's delivered after the three+ separate times that power is shunted from the most charged cells to achieve balance.

The difference might be part of the 'waste overhead' from cell balancing. My W.A.G. (wild Allan guess) might be 0.3 to 0.5 kWh out of perhaps 13 kWh to go from a SOC of about 20% to 100% to 99% to 100% to 99% to 100% to 99% to 100%.

During the several hours of cell balancing, my observation was that about 1.0 kW/hr was being delivered about half that duration to bring the 'low cells' up to 100%. However, that wouldn't be 'wasted' as actual charge is being delivered.

Actual waste would be during the 'shunting' of cells that were at 100% when heat was brought about to draing cells to achieve balance.

I will try to take a closer look at the ChargePoint log that is recorded to my Andriod 7.1 device. I'm trying to figure out to get a 'screen shot' from my smart-phone ... Power-off + reduce-volume doesn't seem to work. Drat.

I suppose this info could also be derived from a Kill-A-Watt device at the wall-plug.

I don't have the LeafSpy app or ELM-327 connector, but my speculation is LeafSpy might provide reasonable numbers of how many kWh were delivered after the first time a SOC of 100% was achieved.
 
LeftieBiker said:
Monthly is good, obviously, although I'd consider it near the infrequent end of good. Once a week is probably the most often anyone would need to do it.
My concern about 'cell balancing' more than monthly ... especially weekly ... is that seems to be 3x more charges to 100% than I prefer to do to 'baby' our Leaf battery.

My typical practice when not doing cell-balancing has been to use the Leaf timer or the ChargePoint app with Level-2 to charge a bit high to about 85% to 90% SOC, and then start driving very shortly afterwards so that the battery is above 80% a very low amount of time. By the time I get home from the charger, the SOC is typically about 80%, depending on the errands I combine with the L2 public charging.

If I cell balance weekly, then 52 times a year the battery is at 99% to 100% for several hours. That seems like perhaps the disadvantage of doing 100% charges might out-weigh the advantage of cell balancing.

My speculation is that with weekly cell-balancing, I might slightly improve my usable range compared to monthly cell balancing, but at the 'price' of reducing my battery lifetime in the long term.

Or am I being overly concerned about 'baby'ing the battery'? I've still got 12 bars at 40,000 miles, and want to delay to loss of a bar for another year or three, if feasible.

Thanks for the data points.
And I would appreciate 'data points' from you and other experienced Leaf owners. I'm curious about how often you 'cell balance'? Weekly? Bi-weekly? Monthly? Less often? More often? Rarely? Before a long day when max range is at a premium?
 
First, I wrote that weekly was the most often I'd suggest balancing if needed - I wasn't suggesting it as best practice. Second, an indicated 100% SOC on a Leaf ranges from about 94% to about 96%, not actually 100%. Last, I do it fairly rarely because I don't drive much: I have yet to even approach 2,000 miles in almost a year. About once every two or three months now is it for me.
 
LeftieBiker said:
First, I wrote that weekly was the most often I'd suggest balancing if needed - I wasn't suggesting it as best practice.
Agree. I realized you weren't advocating weekly balancing. I did want to check if my inference of you suggesting about bi-weekly rather than monthly was closer to what you were in favor of for my moderate annual driving. And my apologies if my comments came across as argumentative or unappreciative of your feedback.

Second, an indicated 100% SOC on a Leaf ranges from about 94% to about 96%, not actually 100%.
Ah ... good point. I hadn't taken that into account. Is that the kind of info available from LeafSpy?

Last, I do it fairly rarely because I don't drive much: I have yet to even approach 2,000 miles in almost a year. About once every two or three months now is it for me.
Wow ... let me know if and when you put your 2018 Leaf up for sale. ;)

And again ... thx for your time and effort put into moderating and contributing your experience and expertise to this forum.
 
I wasn't miffed, just using an economical posting method. It hurts a bit for me to type. I'd say that for typical drivers, balancing every two or three weeks is fine. If the car's range isn't taxed, or it isn't driven much, less often is probably fine as well. And yes, LeafSpy is how we know what the actual (if any LeafSpy reading can be called that, since it's just an interpretation of BMS messages) SOC figures are.

I'll almost certainly be turning in my leased 2018 SL, but if you want to pay the residual, fees etc, that (you buying it) can happen too.
 
First rule of battery management; Make sure you have enough range to meet your needs.

Second Rule of Battery management; see Rule #1


If you are concerned about extra time at 100% charge...or 90% or whatever (its not an absolute. Its a sliding scale) and you have enough at 70% or whatever level to cover rule #1, then I suggest you don't balance at all.

Your cells balance all the time. The "only" advantage to top end balancing is getting a bit more range from a charge. Not charging to 100% is not going to result in a cumulative out of balance scenario.
 
The balance profile shown by LeafSpy does get more and more 'ragged' looking after a month, but the question remains of how much that really matters - except when maximum range is needed.
 
LeftieBiker said:
The balance profile shown by LeafSpy does get more and more 'ragged' looking after a month, but the question remains of how much that really matters - except when maximum range is needed.

True that and what if any of that unbalance is permanent? Cause I doubt it is. So any level of unbalanced can be rectified by anywhere from several days to maybe as much as a few weeks of full charge cycles if planning a trip. In my experience, it usually only takes 3-4.
 
Back
Top