Low cell-pair voltage, CVLI, pack capacity

[GroundHog]

4.0V / cell is a standard 80% charge, not full charge. Not the best for long term storage, but it's what Nissan recommends.

Even at $4000 it's the cheapest we've seen 48 modules sold for yet.

ahagge - not likely that swapping out only a couple modules would do you any good - so far modules in the pack tend to degrade pretty evenly.

Is this really true? I have a 2013 that Leaf Spy shows ONE module that is low upon discharge, and it is this one module that forces the car to turtle every time. With Leaf Spy, on a full charge, I always start at 284 gids (101% gids), 97% SOC, 22 kWh available, and that module is level with all the rest. But at turtle, I'm at 6 GIDS (2% gids) and around 12% SOC, 20 kWh used. This seems to indicate that the one bad module is making me lose about 10% of my potential range. Of course once you get a bad module, it should get progressively worse due to temperature. As the module drains down, the voltage decreases, the internal resistance increases, and the amount of heat generated increases, making this module hotter than the rest of the pack. Thus leading to a greater rate of capacity loss in the future. (or at least I think this is all true).

I would think that in my leaf, replacing that one module would give me a 10% greater range.

I've been posting this a few times, wondering if anyone else has a leaf with a single low module and what effects they see. But I'm starting to wonder if I'm the only one. To my knowledge there are no error codes registered from a single low module, so unless you have LeafSpy or LeafDD, you would never know...except that your leaf just doesn't go as far as it should. And 10% is pretty hard to measure with certainty

Oh and on that storage at 80%. How many of you see brand new Leafs a the dealer charged up to 100%???? Nissan doesn't even follow Nissan's own advice.

 

[dgpcolorado]

Is this really true? I have a 2013 that Leaf Spy shows ONE module that is low upon discharge, and it is this one module that forces the car to turtle every time. With Leaf Spy, on a full charge, I always start at 284 gids (101% gids), 97% SOC, 22 kWh available, and that module is level with all the rest. But at turtle, I'm at 6 GIDS (2% gids) and around 12% SOC, 20 kWh used. This seems to indicate that the one bad module is making me lose about 10% of my potential range. Of course once you get a bad module, it should get progressively worse due to temperature. As the module drains down, the voltage decreases, the internal resistance increases, and the amount of heat generated increases, making this module hotter than the rest of the pack. Thus leading to a greater rate of capacity loss in the future. (or at least I think this is all true).

I would think that in my leaf, replacing that one module would give me a 10% greater range.

I've been posting this a few times, wondering if anyone else has a leaf with a single low module and what effects they see. But I'm starting to wonder if I'm the only one. To my knowledge there are no error codes registered from a single low module, so unless you have LeafSpy or LeafDD, you would never know...except that your leaf just doesn't go as far as it should. And 10% is pretty hard to measure with certainty

Oh and on that storage at 80%. How many of you see brand new Leafs a the dealer charged up to 100%???? Nissan doesn't even follow Nissan's own advice.

I have long had one cell-pair (#48) that is always low when discharging, typically two to three hundredths of a volt. I asked about it and there were no codes set. Turns out that it is well within Nissan's specs for voltage variation (someone posted them from the service manual but I don't recall where). I try not to take the car much below VLBW so I don't find out if it shuts me down early. But even at 20 Gids it isn't anywhere near 3.0 Volts, so I am not close. I would never turtle my car on purpose, why stress the cells unnecessarily? (I bought my LEAF so the usual argument that "it's leased, who cares if I trash the battery?" doesn't apply. I care. A lot.)

While I don't doubt that your low module is the one that causes turtle I don't think that you are anywhere near the 10% loss of capacity that you suggest. How are you measuring that 20 kWh used between 284 Gids and 6 Gids? Although a Gid is nominally 80 Watt•hours, in use it is somewhat less than that IIRC.

 

[GroundHog]

With Leaf Spy, on a full charge, I always start at 284 gids (101% gids), 97% SOC, 22 kWh available, and that module is level with all the rest. But at turtle, I'm at 6 GIDS (2% gids) and around 12% SOC, 20 kWh used. This seems to indicate that the one bad module is making me lose about 10% of my potential range.

I have long had one cell-pair (#48) that is always low when discharging, typically two to three hundredths of a volt. I asked about it and there were no codes set. Turns out that it is well within Nissan's specs for voltage variation (someone posted them from the service manual but I don't recall where).

My cell variation at 20 Gids is around 100-300 mV, much more than you report. The test from the service manual is that the (avg-low) > 2.5 * (high-avg) BUT that this CVLI test is only run when error code P33E6 is reported. My cell variation is well outside the range of Nissan's Specs, but I also don't get the P33E6 error code, so I don't know what that means. I posted about this at
http://www.mynissanleaf.com/viewtopic.php?f=30&t=16070" onclick="window.open(this.href);return false;

I try not to take the car much below VLBW so I don't find out if it shuts me down early. But even at 20 Gids it isn't anywhere near 3.0 Volts, so I am not close. I would never turtle my car on purpose, why stress the cells unnecessarily?

When you get low, is your %GIDS and %SOC about the same? For a normally degraded battery, I'd expect %GIDS to be larger than %SOC, and for a new battery, I'd expect the same. When I get low (LBW and VLBW), my %SOC is about 10 higher than my %GIDS (26% vs. 16% at LBW and 20% vs. 8% at VLBW). Turtle is not necessarily when the lowest cell reaches 3.0 V. When I get turtle, my bad cell is usually around 2.7 V, while the average is 3.1 V. At that point, my cell variation is huge (nearly 500 mV).

While I don't doubt that your low module is the one that causes turtle I don't think that you are anywhere near the 10% loss of capacity that you suggest. How are you measuring that 20 kWh used between 284 Gids and 6 Gids? Although a Gid is nominally 80 Watt•hours, in use it is somewhat less than that IIRC.

I measure this two ways and I get very close the same number over many trials. I drive at a steady speed on the highway with cruise control (50-60 mph), and usually no climate control. I use the number of miles travelled divided by the miles/kWh as reported on the dash (reset each time of course). For example 74 miles driven at 3.7 miles/kWh --> 20 kWh. This is very close to the 20.1 kWh that LeafSpy reports that I have used with 0.3 kWh left in the battery at turtle. Whereas at the start of the trip, LeafSpy reported that I had 22.0 kWh.

 

[dgpcolorado]

...When you get low, is your %GIDS and %SOC about the same? For a normally degraded battery, I'd expect %GIDS to be larger than %SOC, and for a new battery, I'd expect the same. When I get low (LBW and VLBW), my %SOC is about 10 higher than my %GIDS (26% vs. 16% at LBW and 20% vs. 8% at VLBW). Turtle is not necessarily when the lowest cell reaches 3.0 V. When I get turtle, my bad cell is usually around 2.7 V, while the average is 3.1 V. At that point, my cell variation is huge (nearly 500 mV).

I never pay attention to %Gids and only occasionally look at %SOC. I find plain old Gids much more useful to measure whether or not I can make it home, and by what route. So I have no opinion on whether %Gids is a meaningful measurement when compared to %SOC.

As for the cell-pair voltage variation at low SOC, this has long been reported, for example here:
http://www.mynissanleaf.com/viewtopic.php?f=44&t=12561&start=107" onclick="window.open(this.href);return false;
I think that the voltage measurements become unstable at low SOC when under load. But I'll leave that to the EEs here to explain why that might be the case. My own measurements of my low cell-pair are at higher SOC levels; I don't recall what it was at below VLBW. Last year at 54 Gids it was about 7 hundredths of a Volt lower than average. I think that the variation has declined a bit as the battery pack has been aging (but it might be that I'm so used to the low cell-pair that I just don't pay attention to it anymore).

I measure this two ways and I get very close the same number over many trials. I drive at a steady speed on the highway with cruise control (50-60 mph), and usually no climate control. I use the number of miles travelled divided by the miles/kWh as reported on the dash (reset each time of course). For example 74 miles driven at 3.7 miles/kWh --> 20 kWh. This is very close to the 20.1 kWh that LeafSpy reports that I have used with 0.3 kWh left in the battery at turtle. Whereas at the start of the trip, LeafSpy reported that I had 22.0 kWh.

I think that you are placing too much emphasis on the precision of the measurements, especially the rather coarse dash mileage number. The reason I asked is that I was wondering if you were measuring it from the "wall" as I do. Please understand that the kWh numbers — and Gids, for that matter — reported by LeafSpy are extrapolations from measurements that have some variability and are subject to some measurement error. I expect that you've seen the number of Gids change when you park the car, turn it off, then turn it back on. The reason for this, as I understand it, is that the resting voltage of the cells differs from the voltage when under load and it is difficult to come up with a precise number that accurately describes the usable energy left.

That said, you may well be right that something is off on your car. Just be aware that the measurements you are using to try to discern this are subject to considerable variation.

 

[stjohnh]

...When you get low, is your %GIDS and %SOC about the same? For a normally degraded battery, I'd expect %GIDS to be larger than %SOC, and for a new battery, I'd expect the same. When I get low (LBW and VLBW), my %SOC is about 10 higher than my %GIDS (26% vs. 16% at LBW and 20% vs. 8% at VLBW). Turtle is not necessarily when the lowest cell reaches 3.0 V. When I get turtle, my bad cell is usually around 2.7 V, while the average is 3.1 V. At that point, my cell variation is huge (nearly 500 mV).

I never pay attention to %Gids and only occasionally look at %SOC. I find plain old Gids much more useful to measure whether or not I can make it home, and by what route. So I have no opinion on whether %Gids is a meaningful measurement when compared to %SOC.

As for the cell-pair voltage variation at low SOC, this has long been reported, for example here:
http://www.mynissanleaf.com/viewtopic.php?f=44&t=12561&start=107" onclick="window.open(this.href);return false;
I think that the voltage measurements become unstable at low SOC when under load. But I'll leave that to the EEs here to explain why that might be the case. My own measurements of my low cell-pair are at higher SOC levels; I don't recall what it was at below VLBW. Last year at 54 Gids it was about 7 hundredths of a Volt lower than average. I think that the variation has declined a bit as the battery pack has been aging (but it might be that I'm so used to the low cell-pair that I just don't pay attention to it anymore).

I measure this two ways and I get very close the same number over many trials. I drive at a steady speed on the highway with cruise control (50-60 mph), and usually no climate control. I use the number of miles travelled divided by the miles/kWh as reported on the dash (reset each time of course). For example 74 miles driven at 3.7 miles/kWh --> 20 kWh. This is very close to the 20.1 kWh that LeafSpy reports that I have used with 0.3 kWh left in the battery at turtle. Whereas at the start of the trip, LeafSpy reported that I had 22.0 kWh.

I think that you are placing too much emphasis on the precision of the measurements, especially the rather coarse dash mileage number. The reason I asked is that I was wondering if you were measuring it from the "wall" as I do. Please understand that the kWh numbers — and Gids, for that matter — reported by LeafSpy are extrapolations from measurements that have some variability and are subject to some measurement error. I expect that you've seen the number of Gids change when you park the car, turn it off, then turn it back on. The reason for this, as I understand it, is that the resting voltage of the cells differs from the voltage when under load and it is difficult to come up with a precise number that accurately describes the usable energy left.

That said, you may well be right that something is off on your car. Just be aware that the measurements you are using to try to discern this are subject to considerable variation.

...When you get low, is your %GIDS and %SOC about the same? For a normally degraded battery, I'd expect %GIDS to be larger than %SOC, and for a new battery, I'd expect the same. When I get low (LBW and VLBW), my %SOC is about 10 higher than my %GIDS (26% vs. 16% at LBW and 20% vs. 8% at VLBW). Turtle is not necessarily when the lowest cell reaches 3.0 V. When I get turtle, my bad cell is usually around 2.7 V, while the average is 3.1 V. At that point, my cell variation is huge (nearly 500 mV).

I never pay attention to %Gids and only occasionally look at %SOC. I find plain old Gids much more useful to measure whether or not I can make it home, and by what route. So I have no opinion on whether %Gids is a meaningful measurement when compared to %SOC.

As for the cell-pair voltage variation at low SOC, this has long been reported, for example here:
http://www.mynissanleaf.com/viewtopic.php?f=44&t=12561&start=107" onclick="window.open(this.href);return false;
I think that the voltage measurements become unstable at low SOC when under load. But I'll leave that to the EEs here to explain why that might be the case. My own measurements of my low cell-pair are at higher SOC levels; I don't recall what it was at below VLBW. Last year at 54 Gids it was about 7 hundredths of a Volt lower than average. I think that the variation has declined a bit as the battery pack has been aging (but it might be that I'm so used to the low cell-pair that I just don't pay attention to it anymore).

I measure this two ways and I get very close the same number over many trials. I drive at a steady speed on the highway with cruise control (50-60 mph), and usually no climate control. I use the number of miles travelled divided by the miles/kWh as reported on the dash (reset each time of course). For example 74 miles driven at 3.7 miles/kWh --> 20 kWh. This is very close to the 20.1 kWh that LeafSpy reports that I have used with 0.3 kWh left in the battery at turtle. Whereas at the start of the trip, LeafSpy reported that I had 22.0 kWh.

I think that you are placing too much emphasis on the precision of the measurements, especially the rather coarse dash mileage number. The reason I asked is that I was wondering if you were measuring it from the "wall" as I do. Please understand that the kWh numbers — and Gids, for that matter — reported by LeafSpy are extrapolations from measurements that have some variability and are subject to some measurement error. I expect that you've seen the number of Gids change when you park the car, turn it off, then turn it back on. The reason for this, as I understand it, is that the resting voltage of the cells differs from the voltage when under load and it is difficult to come up with a precise number that accurately describes the usable energy left.

That said, you may well be right that something is off on your car. Just be aware that the measurements you are using to try to discern this are subject to considerable variation.

There aren't too many people who have reported cell voltages at turtle, and even fewer that have reported the lowest cell pair. I've done this a couple of times, and I think turtle does occur at or about 3.0 v for lowest cell pair, however, how this voltage is arrived at is unclear. The voltage does drop quite a bit immediately on acceleration at very low voltages, and turtle does not occur at the instant the lowest cell voltage is 3.0, it is likely that the lowest cell voltage has to be below 3.0 for 5 or 10 or 60 seconds, or something like that.

Also remember that all of the numbers we are dealing with are those reported by the Nissan computer, they are not direct measurements in the sense of putting a multimeter on the battery terminals. All have some type of programming done to arrive at the numbers. The most obvious are the Gids, capacities, and KWH remaining. It is clear that these values are very nonlinear (value vs miles remaining at constant load), compared to what would occur if there were a direct measurement. Not only that, it is clear that different year LEAFs are programmed differently in the way they report these values. Nissan has not confirmed (to my knowledge) that ANY of these values mean what we think they do.

 

[dgpcolorado]

...Also remember that all of the numbers we are dealing with are those reported by the Nissan computer, they are not direct measurements in the sense of putting a multimeter on the battery terminals. All have some type of programming done to arrive at the numbers. The most obvious are the Gids, capacities, and KWH remaining. It is clear that these values are very nonlinear (value vs miles remaining at constant load), compared to what would occur if there were a direct measurement. Not only that, it is clear that different year LEAFs are programmed differently in the way they report these values. Nissan has not confirmed (to my knowledge) that ANY of these values mean what we think they do.

Well said. We have a tendency to view these numbers as being more precise than they really are. And the chemistry of batteries isn't as linear or as predictable as we would prefer!

 

[GroundHog]

I think that the voltage measurements become unstable at low SOC when under load. But I'll leave that to the EEs here to explain why that might be the case.

This is why I take the measurements that I 'quote' only when there is no load, like sitting on the side of the road for a few minutes to make sure LeafSpy has updated a few times. Yesterday, I got the battery down to 317V to check the specs against what the manual states. It says that at 317V and no load, the cells are ~3200 to ~3400mV. At 317V (7 GIDS), mine measured 3064 to 3334 mV (avg=3295). I let the car sit for several update cycles before recording that data. How far out of specs does the battery need to get before it registers an error code...if the programming would EVER register an error code that is.

I use the number of miles travelled divided by the miles/kWh as reported on the dash (reset each time of course). For example 74 miles driven at 3.7 miles/kWh --> 20 kWh. This is very close to the 20.1 kWh that LeafSpy reports that I have used with 0.3 kWh left in the battery at turtle. Whereas at the start of the trip, LeafSpy reported that I had 22.0 kWh.

I think that you are placing too much emphasis on the precision of the measurements, especially the rather coarse dash mileage number. The reason I asked is that I was wondering if you were measuring it from the "wall" as I do. Please understand that the kWh numbers — and Gids, for that matter — reported by LeafSpy are extrapolations from measurements that have some variability and are subject to some measurement error.

I think you and stjohn are right about the kWh data from LeafSpy. But I have started to read kWh at the wall now that I no longer need to preheat the car everyday :evil: . What I find now is that the wall charger is also indicating that my battery capacity is low (low charge going in, low charge coming out, and yet 284 GIDS on a full charge!!).
%start---%end---miles---m/kWh by dash----kWh by miles---kWh by wall
80%-----20%----46.7------4.3--------------10.9-----------12
80%------T-----59.5------3.9--------------15.3-----------17
80%-----17%----40.2------3.6--------------11.2-----------13

The wall numbers seem to imply a usable battery capacity of 19 kWh or so. That seems pretty bad for a leaf that reports 284 GIDS. Does anyone think I am asking too much of the numbers the Leaf gives me? I know my wall meter is guaranteed accurate. All of these numbers do have some display errors like +-1 kWh for the wall meter, but that averages out in the long run.

That said, you may well be right that something is off on your car. Just be aware that the measurements you are using to try to discern this are subject to considerable variation.

My delima, and why I am seeking advice and help understanding the numbers available to us, is now to convince the dealer that my capacity is low and that the bad cell is likely the culprit.
I suppose the real issue here is that our batteries have a warranty for 100,000 miles but not for capacity except for the new '9-bar warranty'. But nearly everything that goes wrong with a battery will affect capacity in some way. So what exactly is that battery warranty good for?

 

[dgpcolorado]

Part of the problem with using Gid meters to try to discern the battery capacity is that they were designed for 2011/2012 models. The batteries/software of 2013 and, presumably, 2014 models do not give readings that mesh with what we know about the older cars. This adds additional uncertainty with trying to use LeafSpy to measure what's going on with your battery. So far as I am aware, the puzzle of why newer LEAFs give odd readings hasn't been solved (Chemistry? Software tweaks? Both?). This makes it awfully hard to draw conclusions from your measurements. The extrapolations of capacity by the meter just might not work for 2013 models.

As for the warranty factors, Nissan does have figures in the service manual that define whether or not a cell-pair is out of spec. Whether these numbers have changed for 2013 models, I don't know. If a cell-pair is unusually low in voltage compared to the others I believe that it will set an error code and it will be replaced under warranty. For my 2012 I was surprised how wide the voltage range was. My always low cell-pair wasn't remotely close to the limits. So I just learned to ignore it. If you take a look at the limits in the 2013 service manual you might find the same thing.

 

[DaveEV]

GroundHog, what are your charging / driving habits?

 

[GroundHog]

GroundHog, what are your charging / driving habits?

Driving:
I drive about 1000 miles per month. Most trips are 10-20 miles, with lots of dad-taxi service going on. I try to take the flat routes with few stops and average 25-50 mph. Due to the issues with this car, I have done more highway driving just to test the vehicle out. So I've added around 500 miles of driving at constant speeds of 55-70 mph for testing.

Charging:
No access to QC, so forget that. Always level 2 charging. Charge to 80% on timer. 100% only on some weekends, when I expect to drive far OR if I'm out of town in the gas-car, my wife charges to 100% each day, still on the timer though. Charging is always done in off-peak early morning hours, except I may plug it in during the day if I'm driving alot that day.

 

[DaveEV]

One thing I found is that 100% charging induces a lot more balancing activity than you see on 80%. I also have a cell-pair that tends to read low (it reads 4.08V on a 100% charge when the average is 4.10V - a couple others will be at 4.09V but the rest all around 4.095-4.105V), but when charging to 80% only a handful of cell-pairs normally see balancing. When charging to 100%, probably 80% of the shunts are activated. The balancing activity often continues past the next day - some times the next day after a 100% charge, I'll see the pack shunting the same cells.

If you'd like to give the car all the opportunity possible to balance the pack, I would try charging to 100% on end-timer. Set the end time to your expected departure time, in my experience this will finish about 20-50 minutes before the end time.

Then drive the car down to 60% or whatever level you think you'll need more charge - if that's daily or every other day - fine, just discharge it down under 80% before charging it back up.

Do that for two weeks and see how the pack has balanced. To help track progress, grab a screenshot of the cell-pair voltages after every 100% charge.

That should help bring the low cell-pair back up.

 

[gbarry42]

I can't quite figure out all the data. The table with the three lines all show the car charged to 80%. Then you say it took only 17kWh. But that's a reasonable figure for 80% if the battery is new.

You also report 284 GID, which can only be reached by a 100% charge. So I know you do that, too. Since the algorithm is top-balancing, all of the cells can be made to reach their full voltage regardless of capacity. If the GID value is based on the top end, then lack of capacity would result in higher than expected GID value at turtle.

Bad cells are rare, so it's hard for anyone to relate to that. There was one person that got a call from Nissan to bring the LEAF in for a cell replacement, and I don't think they ever saw an error code. Well, that's what "should" happen. Maybe if you keep thrashing it, they will pick it up. Assuming every car still has Carwings and still reports what it knows.

 

[dgpcolorado]

...That should help bring the low cell-pair back up.

Interesting suggestion. As it happens my low cell-pair has finally disappeared into the pack, so to speak.

Perhaps not coincidentally, the battery capacity has started to decline again now that it has warmed up above 12ºC much of the time (14ºC to 17ºC today). The capacity was rock solid most of the fall and winter while the battery was cool.

 

[GroundHog]

I can't quite figure out all the data. The table with the three lines all show the car charged to 80%. Then you say it took only 17kWh. But that's a reasonable figure for 80% if the battery is new.

17.5kWh of usable energy is about right for an 80% charge. But the 17kWh I get is measured using a utility grade meter mounted on my wall feeding only my charger which also measures the charging losses which are about 15%. Thus by measuring the 17kWh, I surely have a battery with 15% less capacity or so. I have to thank dgpcolorado for the idea to measure this. Since I have been measuring the kWh at the wall, I get data that is consistent with that 17kWh (metered at the wall) for 80% of a usable charge EVERYTIME. That is the number of kWh metered at the wall is 10-15% higher than the number of kWh by miles (like I posted before). And the kWh by the wall meter are never high enough to account for a full capacity battery and charging losses. This really indicates a problem in my mind.

Bad cells are rare

I don't know that we know that. What seems to be the case is that since no error codes are registered, and the loss from a bad cell is about 10-15% of the total capacity, you would have to be VERY diligent in keeping data to know for sure that you have a bad cell. You are more likely to just be upset with your car's range. If you complain, Nissan will just discount your situation as "driving conditions vary". I don't know that many of the 100,000 Leaf drivers out there have LeafSpy or LeafDD. I suspect that less than a few hundred drivers have them. So it is quite possible that there are perhaps 100-1000 bad cells out there (0.1% to 1%) and their owners are unaware.

 

[GroundHog]

One thing I found is that 100% charging induces a lot more balancing activity than you see on 80%. I also have a cell-pair that tends to read low (it reads 4.08V on a 100% charge when the average is 4.10V - a couple others will be at 4.09V but the rest all around 4.095-4.105V), but when charging to 80% only a handful of cell-pairs normally see balancing. When charging to 100%, probably 80% of the shunts are activated. The balancing activity often continues past the next day - some times the next day after a 100% charge, I'll see the pack shunting the same cells.

If you'd like to give the car all the opportunity possible to balance the pack, I would try charging to 100% on end-timer. Set the end time to your expected departure time, in my experience this will finish about 20-50 minutes before the end time.

Then drive the car down to 60% or whatever level you think you'll need more charge - if that's daily or every other day - fine, just discharge it down under 80% before charging it back up.

Do that for two weeks and see how the pack has balanced. To help track progress, grab a screenshot of the cell-pair voltages after every 100% charge.

That should help bring the low cell-pair back up.

This is sort of what the dealer told me. So I did go through quite a few 100% charge cycles with lots of time for balancing. It didn't really change anything. My bad cell is usually the lowest when fully charged by about the amount yours is, but then as it discharges, the bad cell gets out of balance again. PLUS when the 5-star report is done (I had 3 stars!!), they say that I am charging to 100% too much and that explains why "my battery has lost capacity". I stopped trying to get it to balance at 100%, and now my 5-star report is indeed 5 stars.
Also I think the battery is more balanced at 80% anyway (deltaV of 25 mV at 100%, deltaV of 20mV at 80%).