Capacity Loss on 2011-2012 LEAFs

[azdre]

TickTock compared wall, driving range, gids, bars, carwings, etcetera

wall - this is power in but if power used isn't the full range this doesn't tell us everything - see driving range

driving range - this could be limited by a software bug that shuts down the car before the battery is fully depleted. I'm not saying that is the case but it's the only thing I can see in the list that everything else relates to and OrientExpress is saying we should consider software issues.

gids, bars, carwings - these all go back to whatever software manages the bottom limit on driving range.

I think what Orient Express is saying so caustically is that short of pulling the battery pack out and testing it independent of the leaf's ECUs you are dealing with indirect data and can't be sure it is trustworthy.

But I also have to ask more to be able to defend TickTocks side here. How did you determine driving range? Did you drive until the turle mode stopped working and the leaf was dead and log odometer miles? Or are we talking some less extreme method of measuring drop in driving range?

I'm not in favor of his style of delivery but I think it is worth considering how many computers are inside a single car and how any possible firmware/software/signalling/measuring error could make things complicated.

Agreed. Knowing how much energy the batteries *are* storing does not necessarily tell us how much energy they still *might* be capable of storing, independently of the car's charging and reporting systems.

At the end of the day, the only "measurement" that really matters is that the distance the car can travel has been reduced to the point where the car is no longer useful as a primary commuter, and we are just lucky that we have a Prius and a motorcycle in the garage to augment our travel needs.

I tried an 80% charge today. 4.5 miles/kWh got me ~41 miles to LBW. I won't be doing that again. 85% my a$$.

[vrwl]

Well, add me to the list. I lost a capacity tick this morning... 22,000 miles, 2 QCs, normally charge to 80 percent, 17 months since delivery.

Tom, your car has been added to the 1 Bar Lost table in the Wiki. Can you tell us the Mfg Date of your car? (metal plate inside driver-side door opening) And once you've reported the bar loss to Nissan, please give us your case number along with the date it was reported to Nissan. Thanks and I'm sorry about your car.

[TickTock]

We have confirmed the loss through many means: power from the wall, driving range, gids, bars, carwings, etc. All have consistent results.

The car s/w has nothing to do with the accuracy of my wall power meter nor the number of miles I drive. This was the point I was trying to make - we aren't just basing this on one source of information - it has been confirmed through several different means. Some of which have no dependence on the car s/w or interpretation of OBD messages.

The issue with the idea of multiple means to base your assumptions is that they are all using the same base data to present their results.

I have not seen any other sources of base data except the data that is presented on the can-bus that is in-turn used by the Gid meters, charger, and systems that determine how much energy is available to drive n miles. Until someone can recreate this source data independently of the source data that is being used now, then a SW bug cannot be ruled out, nor can it be assured that the source data being used for all of the hypothesis floating around is accurate either.

Can anyone say with certainty that a data register that might have something to do with ambient temperature does not have corrupt data in it based on a combination of conditions that the affected cars experienced? No, because no one here knows how that source data is created. No one knows for certain the condition of the batteries in an affected pack are, all they can do is try to interpret un-verified data.

This is why a SW bug may very well be the issue.

OK. Maybe I misunderstood your original statement. I thought you were saying our measurements of the reduction in the amount of charge the battery can take (when charge by the car) or the driving range may be wrong due to a s/w bug. This is what has been verified multiple ways and there really is no disputing it. If, however, you are saying that maybe the car is experiencing a reduction in the available charge due to a bug limiting the charge capability then I concur that has not been conclusively disproven.

[TickTock]

[edit: deleted weird repeat post]

[vrwl]

I live in Plano TX and just lost my second bar yesterday after 13.5 months of ownership and 16,500 miles. I lost the first bar at about 12 months and 15,000 miles. Reported to Nissan with Case# 913-3854

fattmerris, your car has been added to the 1 Bar Lost list AND the 2 Bars Lost list in the Wiki. If you would, please provide us with the Mfg Date for your car (metal plate inside the driver-side door opening) along with the date you reported your losses to Nissan. Thanks and I'm sorry to hear about your car.

[vrwl]

PC to Nissan to report the battery loss , seemed pleasant enough, an they are aware of the issue in warmer climates. No news at this time and I have an official case # 9129683. I Will wait and see. Hard to think I would lose it this fast with the short driving and really have only had mine for just 3-4 months of summer use. I bought it in the winter 10/11. mfr date 5/11, put about 8000 miles myself with 2800 coming from dealer. no QC ever , and did the 100% charge in the winter when it was cooler, 80% in summer, 11,000 total miles I would say is early for loss. I see some after 15,000 that is what I was shooting for.

The wiki has been updated with the additional info you provided.

[TickTock]

Following up on OrientExpress's comment, I was wondering what sort of bug could cause this. I plotted two charging profile's - one taken last October and the other last month. X axis is the charge from the wall (not charge into the battery) starting from gids=6 to wherever the 100% charge stopped, Y axis is the battery voltage throughout the charge (as reported by canbus). At first glance, it does appear that if the battery voltage sensor is reading about 5V high from last October, that could result in the behavior we are seeing. Shift the red (new) curve down 5V and it looks like it will track the blue (old) curve fairly well and would allow further charging. However, at low charge levels, the curves cross suggesting that the sensor doesn't, at least, have a static offset. It appears it takes the same amount of charge to get to the knee, but the voltage when it reaches there is higher. Maybe someone who understands more about battery chemistry can comment about whether we would expect the voltage of the knee to increase with age.

[DaveEV]

Following up on OrientExpress's comment, I was wondering what sort of bug could cause this. I plotted two charging profile's - one taken last October and the other last month.
...
It appears it takes the same amount of charge to get to the knee, but the voltage when it reaches there is higher. Maybe someone who understands more about battery chemistry can comment about whether we would expect the voltage of the knee to increase with age.

That is awesome data right there. It leaves very little to dispute that you've lost about 4 kWh from the wall over the 7 months between charges. Considering that on a new car it takes 24-25 kWh from the wall to charge, you are close to 6 kWh down in capacity.

As far as what might cause the voltage knee to rise like that when charging - my best guess is that at least some of it is due to ambient temperature. Hotter temps should see higher voltages.

It could also mean higher internal resistance. But then I'd expect the voltage to be flatter as charge rate drops.

I do have to wonder if this is part of the reason why L1 charging might hide capacity loss somewhat - the reduction in charging power gives the battery more time to absorb the charge before hitting voltage limits.

[TonyWilliams]

TonyWilliams"]While everybody is pontificating about temperature adjustments to available capacity, please address why I had 15-18% loss of RANGE / AUTONOMY in 75F temperatures?...

How many times did you run your LEAF past VLBW?

To "turtle"?

Until shut-down?

="TonyWilliams"....And a similar numbers of performance reduction from other LEAFs in other "hospitable" climates...?

I don't doubt some other "moderate climate LEAFs were as "close to" losing a bar as yours was, before you sold it, but I doubt that you actually are claiming your LEAF was typical, for all LEAFs with similar temperature histories, charge cycles, and time from manufacture.

I'm confident you know my car's history well enough to answer that for yourself? So, is that the problem with my battery (and not mentioned in the owner's manual)? TomT's car probably most closely matches my weather pattern, without the turtle events, yet already lost a bar? How do you reconcile that?

Actually, his battery is obviously slightly worse than mine was (from last week, when I last owned it). I suspect you could write the public Nissan summary to this whole debacle; the owners all killed their own batteries, which are normal anyway.

[TonyWilliams]

At the end of the day, the only "measurement" that really matters is that the distance the car can travel has been reduced to the point where the car is no longer useful as a primary commuter

This is where Nissan will lose in the end game. They carefully write themselves out of battery capacity culpability, BUT THEY HAVEN'T DONE THE SAME FOR >>> LOSS <<< OF RANGE / AUTONOMY.

I don't give an F what the battery is doing, and neither will a jury. The car was bought and went X miles; less than a year later it goes X - 30% in the exact same conditions. That's the question to be answered.