Capacity Loss on 2011-2012 LEAFs

[dvlax40]

Just let me know rick.


update, after multidays with battery over 100 (currently 8:30pm and battery still shows 106 after 3 hours in a 90 degree garage)

my battery capacity is down to 95.73% AHr= 63.42

 

[TomT]

IMHO, to assume that you can accurately measure the battery capacity to that level of precision with the Android App method is simply seeing something that is not there... If you can actually get to within a few percent, I'd be surprised.

my battery capacity is down to 95.73%

 

[Valdemar]

Today I drove 73.9 miles to LBW from 100% charge with 5.8m/kwh according to the dashboard. It was above 70F today, more like 85F during daytime, 6 temps bars, given the car was in the shade while parked I doubt the battery warmed up more than 75F. Using Tony's method I arrived at the new number for the remaining capacity of 78%. At 78% I should be at 10 capacity bars, or at least 11. Yet I still have all 12 bars showing. Subjectively it feels like I lost about 10-12% of range since new, which is in agreement with the capacity bar reading. Sounds like the calculations that Tony used produce an overly pessimistic number?

 

[Boomer23]

I drove my 2011 LEAF down to turtle for the first time since July of 2012 to see how much range I actually have left.

LEAF was delivered March 30, 2011. I have 23,816 miles on it. I still have 12 capacity bars. The Android/ELM app shows 56.8 Amp hours and 85.76 CAP.

When I do these range tests, I make sure that I have an indicated 4.1 mi/kWh on the dash at the end of the test.

Today I got 69.9 miles total range from full charge at indicated 4.1 mi/kWh. Turtle came at 5 Gids on an original Gidmeter.

July 2012 the same test yielded 75 miles.

In April 2011, when the car was new, I got 86.5 miles at an indicated 4.0 mi/kWh.

I'm calculating a loss of 19.2% of range from the 86.5 miles when new to 69.9 miles now. (Ignoring the discrepancy between 4.0 mi/kWh indicated on the 2011 new battery test and the 4.1 on the subsequent tests.)

Note: I did put Michelin Primacy MXV tires on the car this April, replacing worn out Ecopias. I suppose this could have affected the range measurement, but only if the Michelins are of different diameter to the Ecopias. I don't have data on this. If anyone does, please comment.

 

[Valdemar]

^^^ So, if anything, your battery capacity meter is too optimistic which is the exact opposite of what Nissan is saying.

 

[Boomer23]

^^^ So, if anything, your battery capacity meter is too optimistic which is the exact opposite of what Nissan is saying.

Yup.

 

[Stoaty]

I'm calculating a loss of 19.2% of range from the 86.5 miles when new to 69.9 miles now. (Ignoring the discrepancy between 4.0 mi/kWh indicated on the 2011 new battery test and the 4.1 on the subsequent tests.)

Note: I did put Michelin Primacy MXV tires on the car this April, replacing worn out Ecopias. I suppose this could have affected the range measurement, but only if the Michelins are of different diameter to the Ecopias. I don't have data on this. If anyone does, please comment.

There are reports of the Michelin tires decreasing efficiency (and thus range) by 8-10%, so I would say your test is not valid. You are assuming the miles/kwh on the dash is accurate, which may not be true.

 

[dvlax40]

IMHO, to assume that you can accurately measure the battery capacity to that level of precision with the Android App method is simply seeing something that is not there... If you can actually get to within a few percent, I'd be surprised.

my battery capacity is down to 95.73%

I understand that there is a possibility of inaccuracy. im keeping track more for a reference.

but i have to say, if the data is being pulled directly from the can-bus with no interpolation or interpretation, it should be pretty damn accurate.

 

[TomT]

When I first got my MXV4s, I experienced a range drop of about 4%. Now that they are worn in, it is more like 2%. BTW, new tires will always be less efficient than old ones...

There are reports of the Michelin tires decreasing efficiency (and thus range) by 8-10%, so I would say your test is not valid. You are assuming the miles/kwh on the dash is accurate, which may not be true.

 

[TomT]

There is a difference between accuracy and precision. The data you are reading has high precision but relatively low accuracy...

but i have to say, if the data is being pulled directly from the can-bus with no interpolation or interpretation, it should be pretty damn accurate.

 

[TonyWilliams]

Yes I guess I should call, I had already asked for a buy back before. I'm over 2 years now, but I have proof I brought this up at around 12K when I lost the first bar. I'm home sick today with the flu so probably not much going to happen in the next day or two. I do have a CAN sniffer available at work, I just need to get the OBD2 plug and figure out how to use it, anyone know where to get one locally?

dvlax40, I may take you up on that offer in a couple days.

I don't know what a "CAN sniffer" is, but the Turbo3 app for Android devices is the only easy way to get the AMP HOUR data we need. Time is of the essence.

 

[dvlax40]

There is a difference between accuracy and precision. The data you are reading has high precision but relatively low accuracy...

but i have to say, if the data is being pulled directly from the can-bus with no interpolation or interpretation, it should be pretty damn accurate.

what are some causes of the low accuracy? I would hope that the can-bus is feeding information to the ecu that was at least 90%+ accurate. if this was an ICE a discrepancy like that would cause it to run terribly

 

[TomT]

Likely cost more than anything else. Nissan did not use instrumentation that is particularly accurate in some cases. Their Hall Effect Coulomb counter is one such example... Some things, such as voltage, are easy to inexpensively measure very accurately, but others are not...

what are some causes of the low accuracy? I would hope that the can-bus is feeding information to the ecu that was at least 90%+ accurate. if this was an ICE a discrepancy like that would cause it to run terribly

 

[dvlax40]

got it... poor form on nissan as the technology for accurate capacity measuring is not terribly expensive

 

[WetEV]

my battery capacity is down to 95.73%

Measuring battery capacity isn't easy.

Basically you are measuring the current vs voltage curve of the battery to predict how much current will be needed to fully discharge the battery.

The current measurement is made with a Hall effect sensor that isn't very accurate. Yes, this could be done better for more money. This is a Leaf, not a Tesla.

The voltage measurement should be fairly accurate, but is trying to measure fairly small changes in voltage.

The quality of the measurements can be improved by various cross checks, and by integrating over time. But there are real limits to how accurate this measurement can be made.

I suspect (and have no information to back this up) that a full charge to LBW cycle would have improved accuracy, and that daily driving from 80% to 50% would have reduced accuracy.

In My Never Humble Opinion, 5% accuracy would be fairly good. 1% would be very excellent.

 

[Boomer23]

I'm calculating a loss of 19.2% of range from the 86.5 miles when new to 69.9 miles now. (Ignoring the discrepancy between 4.0 mi/kWh indicated on the 2011 new battery test and the 4.1 on the subsequent tests.)

Note: I did put Michelin Primacy MXV tires on the car this April, replacing worn out Ecopias. I suppose this could have affected the range measurement, but only if the Michelins are of different diameter to the Ecopias. I don't have data on this. If anyone does, please comment.

There are reports of the Michelin tires decreasing efficiency (and thus range) by 8-10%, so I would say your test is not valid. You are assuming the miles/kwh on the dash is accurate, which may not be true.

OK, point taken, though I tend to believe the dash energy economy display because the results of this test match well with many informal calculations that I've done using Gids on many trips down to LBW and below. And if I posit that the display is at least precise, then the change in tires would not have an effect on the data as long as I keep the dash energy economy reading constant. Unless the new tires are a different diameter than the Ecopias were when they were new, which would affect the odometer reading.

But here is some additional data of charging energy required to recharge to full from turtle using TED mounted on my breaker box:

April 2011 Turtle event with new battery: 26.1 kWh required.
June 2013 Turtle event: 20.2 kWh required.

That's a 23% drop in energy required for refill.

Or if you like, we could calculate theoretical miles per each charge and compare with measured miles.

April 2011: 26.1 kWh minus 15% charging inefficiency = 22.2 kWh at the battery X 4.0 mi/kWh = 88.8 theoretical miles. Measured miles = 86.5

June 2013: 20.7 kWh minus 15% = 17.6 kWh at the battery X 4.1 mi/kWh = 72.2 theoretical miles. Measured miles = 69.9

I'm inclined to believe my data, but I don't expect anyone to consider it rigorous.

EDITED: I took another look at my TED readout and I see that the power used for full recharge last night was 20.7, not 20.2. I must have misread it this morning. I just wanted to be as precise as possible. The overall conclusion still holds, I think.

 

[Stoaty]

But here is some additional data of charging energy required to recharge to full from turtle using TED mounted on my breaker box:

April 2011 Turtle event with new battery: 26.1 kWh required.
June 2013 Turtle event: 20.2 kWh required.

That's a 23% drop in energy required for refill.

That data I find pretty convincing, although I would wonder whether battery pack temps were similar for the two events.

 

[SanDust]

I'm inclined to believe my data, but I don't expect anyone to consider it rigorous.

I think it's just mental rounding which is leading to the skepticism. People look at 19% and think 20% which seems a little steep. If you said 16.7% they mentally take that to 15% and think it was about right.

I got my Leaf at the same time you did, have a few more miles on it than you do, and get about the same range as you do. I know my range is reduced but couldn't really say it's been 20% because I never measured it when I first got it. Could be but I didn't really have a good feel for the range until the summer of 2011, at which time the baseline range was probably lower than when I first got it.

Leaf owners at work are reporting losing a bar. I suspect I'll lose one by the end of the summer. When I last checked almost a year ago I was at 87%, and the reports are that you lose the bar at 82-83%. I haven't paid much attention to the bar loss because it just tells you that your range has been reduced, which you already know.

 

[Valdemar]

Reduced range is not a big deal until your daily commute is impossible on a single charge and you have no convenient places to charge such as at work. I'm currently driving 50-70 miles a day and considering the rate of degradation up to this point I estimate that my Leaf will (conveniently) work for my needs for about 2 more years if not more. I'm debating if I should dump the car (I bought) now before there are visible signs of capacity loss or push it until the loss of range gets in the way. The problem is finding a replacement for the Leaf considering my rather long commute and a the hit I would take by selling it now. Any other car, including a new Leaf, will be more expensive even if I had to get a new battery in the future (pending battery replacement cost announcement). I guess I could get a small ICE car like a Honda Fit or the like, but I don't want to go that route. With that I'm leaning towards keeping the car and hope that I will either have access to charging at work in the future or my commute will be shorter.

 

[DaveEV]

But here is some additional data of charging energy required to recharge to full from turtle using TED mounted on my breaker box:

April 2011 Turtle event with new battery: 26.1 kWh required.
June 2013 Turtle event: 20.2 kWh required.

That's a 23% drop in energy required for refill.

That data I find pretty convincing, although I would wonder whether battery pack temps were similar for the two events.

Yep, it also closely correlates with the dash numbers, too:

April 2011: 86.5 mi / 4.0 mi/kWh = 21.625 kWh
June 2013: 69.9 mi / 4.1 mi/kWh = 17.050 kWh

17.050 kWh / 21.625 kWh = 78.9%
20.7 kWh / 26.1 kWh = 79.3%

+- a couple percent is all one can ask for given the precision of the mi/kWh gauge which at around 4 mi/kWh is only accurate to +-2.5%.

So yeah, Boomer's car is definitely down at least 20% in capacity compared to new and 21% is a good estimate.