Discuss data from the LEAF Battery app, and Comparisons

[mwalsh]

Comparing it to where I lost my segments from a capacity percentage POV:

Loss of bar 12 - 81.61%
Loss of bar 11 - 78.63%
Loss of bar 10 - 71.90%

Capacity at 66.22% as of today, below the published threshold for bar 9 loss. Though, as you can see, my car does trend towards losing a bar below the published threshold. AHr reading is 43.44AHr. Gids are down to the mid-170s.

As always, full data can be seen here:

http://www.mynissanleaf.com/viewtopic.php?f=44&t=12789&start=567

 

[fooljoe]

How is it "OK" that actual bar losses like mwalsh has seen are so far off from Nissan's published values? I'm bamboozled by the fact that we've seen bar losses reported over such a wide range of Ah and SOH readings and that these readings are always below what Nissan says they should be. Clearly there must be some consistent internal measure that determines when capacity bars are dropped, but if it's not Ah or SOH, then what could it be?

If it hasn't happened already, at some point we're bound to see someone whose capacity fell below the 66.25% published threshold in time but didn't lose the 9th bar in time and so got denied a replacement. What then? I'd expect Nissan would say these numbers pulled from LeafSpy and the like (their own numbers) are unreliable, but if that's the case, what is the reliable number?

mwalsh or anyone else, have you done actual range tests, driving to turtle and recording miles driven and miles/kWh and compared the measured kWh consumed as a fraction of usable kWh when new to your reported SOH value? It seems we need to do something to establish how accurate our available capacity measures are in preparation for the eventuality of a disputed warranty claim.

 

[mwalsh]

How is it "OK" that actual bar losses like mwalsh has seen are so far off from Nissan's published values? I'm bamboozled by the fact that we've seen bar losses reported over such a wide range of Ah and SOH readings and that these readings are always below what Nissan says they should be. Clearly there must be some consistent internal measure that determines when capacity bars are dropped, but if it's not Ah or SOH, then what could it be?

If it hasn't happened already, at some point we're bound to see someone whose capacity fell below the 66.25% published threshold in time but didn't lose the 9th bar in time and so got denied a replacement. What then? I'd expect Nissan would say these numbers pulled from LeafSpy and the like (their own numbers) are unreliable, but if that's the case, what is the reliable number?

mwalsh or anyone else, have you done actual range tests, driving to turtle and recording miles driven and miles/kWh and compared the measured kWh consumed as a fraction of usable kWh when new to your reported SOH value? It seems we need to do something to establish how accurate our available capacity measures are in preparation for the eventuality of a disputed warranty claim.

Well, Nissan can (and likely will) always suggest that the numbers from LEAFSpy and other data loggers are not entirely accurate, and that might well be the case - I would imagine the boffins who reverse engineered the LEAF's CanBus did have to make some (well) educated guesses.

And for sure Nissan will only accept their own metric when it comes to warranty replacement, and that metric is 8 bars. No ifs, ands, or buts about it. So it doesn't really matter if other data is presented that suggests a car should have lost a bar when it hasn't, Nissan isn't going to acknowledge it. Ever.

As to why cars are loosing bars all over the place (though almost all loose bars within the anticipated ranges for both AHr and Capacity) it could be that the numbers from whatever data stream Nissan in using to generate the capacity bars are being averaged over a period of time before a LEAF's systems decide that it's time to stop displaying a bar.

 

[fooljoe]

it could be that the numbers from whatever data stream Nissan in using to generate the capacity bars are being averaged over a period of time before a LEAF's systems decide that it's time to stop displaying a bar.

Yeah, I remember reading somewhere in here a while back that something like this is likely the case - but it was probably just speculation. Now that we know that you've crossed the threshold and when, we'll just have to track how long the "waiting period" is that's imposed by whatever algorithm Nissan uses. I'm about 1-2 months behind you (currently 44.4 Ah), so I'm following your progress intently! Unfortunately, the available data probably don't support measuring this waiting period for significant numbers of drivers - people are much more likely to notice what the capacity was at the time of bar drop rather than at what time they crossed the magical 66.25% threshold.

If this time-averaging or waiting period or whatever is the case, I still think there's an issue with the integrity of Nissan's publicized SOH vs. capacity bars mapping - the table says "value at which bar drops", not "value at which bar drops after a month" or whatever time. That waiting period could and likely will cost at least somebody a $5k+ battery! Anyway, here's to hoping that at least our bars drop without getting into such a nailbiter situation!

 

[Turbo3]

Where are people getting the Capacity number from? As far as I know, LeafSpy does not provide such a number. How is it read from the Leaf (which CAN message)? Or is it calculated from some other numbers that are available?

 

[mwalsh]

Where are people getting the Capacity number from? As far as I know, LeafSpy does not provide such a number. How is it read from the Leaf (which CAN message)? Or is it calculated from some other numbers that are available?

The number in parentheses in the old Leaf Spy; what I believe became SOH in LEAFSpy Pro. Am I mistaken that this is capacity? It sure looks like capacity to me!

 

[Turbo3]

Where are people getting the Capacity number from? As far as I know, LeafSpy does not provide such a number. How is it read from the Leaf (which CAN message)? Or is it calculated from some other numbers that are available?

The number in parentheses in the old Leaf Spy; what I believe became SOH in LEAFSpy Pro. Am I mistaken that this is capacity? It sure looks like capacity to me!

If you can give me the old version number I can have a look at the old code and see how it was generated but it would not be SOH which is an integer number. Most likely a calculated value based on something read from the Leaf divided by a fixed constant.

 

[mwalsh]

If you can give me the old version number I can have a look at the old code and see how it was generated but it would not be SOH which is an integer number. Most likely a calculated value based on something read from the Leaf divided by a fixed constant.

0.26n9. I use it over LEAFSpy Pro most of the time, even though I have both on my phone.

 

[TimLee]

Déjà vu.

I believe there was near identical discussion six months to a year ago.

 

[Turbo3]

Where are people getting the Capacity number from? As far as I know, LeafSpy does not provide such a number. How is it read from the Leaf (which CAN message)? Or is it calculated from some other numbers that are available?

The number in parentheses in the old Leaf Spy; what I believe became SOH in LEAFSpy Pro. Am I mistaken that this is capacity? It sure looks like capacity to me!

The number in parentheses was the AHr rating (shown on the top) divided by 65.6 or whatever max value you set on the setting screen if you did not use the default value.

65.6 was the max AHr of a new 2011 battery. Newer Leaf batteries had a higher initial AHr rating so you would need to increase this max value in settings.

This may or may not be what Nissan is calling capacity.

 

[dhanson865]

I've been trying to figure out in terms of AHr precisely where I might loose bar 9. Some time ago, and I really don't think this came from published data, I surmised that the following would hold true:

Loss of bar 12 - between 54AHr and 55.5AHr
Loss of bar 11 - between 50AHr and 51.5AHr
Loss of bar 10 - between 46AHr and 47.5AHr
Loss of bar 9 - between 42AHr and 43.5AHr

So if Ed's 4ahr bands are correct it'd be

Losses are triggered at
56 AHr (12 down to 11)
52 AHr (11 down to 10)
48 AHr (10 down to 9)
44 AHr (9 down to 8)

with a disclaimer that you may see it drop 2.x AHr past that if in extremely hot conditions.

or if you prefer bands maybe something like

Loss of bar 12 - between 53.75 AHr and 56 AHr
Loss of bar 11 - between 49.75 AHr and 52 AHr
Loss of bar 10 - between 45.75 AHr and 48 AHr
Loss of bar 9 - between 41.75 AHr and 44 AHr

I'm at 52.7x AHr so I'll drop bar 11 before next summer. I'll be sure to post my log data when the loss happens.

edit: fixed typo of 42 where I meant to have 44.

 

[mwalsh]

So if Ed's 4ahr bands are correct it'd be

Losses are triggered at
56 AHr (12 down to 11)
52 AHr (11 down to 10)
48 AHr (10 down to 9)
42 AHr (9 down to 8)

with a disclaimer that you may see it drop 2.x AHr past that if in extremely hot conditions.

or if you prefer bands maybe something like

Loss of bar 12 - between 53.75 AHr and 56 AHr
Loss of bar 11 - between 49.75 AHr and 52 AHr
Loss of bar 10 - between 45.75 AHr and 48 AHr
Loss of bar 9 - between 41.75 AHr and 44 AHr

We already have a set of numbers that work quite well with the data available to suport them. I don't personally see it as being necessary to revise them.

This may or may not be what Nissan is calling capacity.

My numbers for 10 and 11 segment turn-off are reasonably close to the published ones, so I'm willing to say that it's capacity or a close facsimile. Though I can't explain why my 12th bar turned off at such a low number compared to the published one.

 

[TimLee]

... 65.6 was the max AHr of a new 2011 battery. ...

A lot of people stated 66.25 as the 2011 new value.

But there weren't many people with brand new 2011 LEAF that had device to read the value.
And Nissan never published an official value of what they are using in their capacity bar determination or the time duration they are using before dropping a capacity bar.

So difficult to be definitive about % capacity.

 

[mwalsh]

I've been trying to figure out in terms of AHr precisely where I might loose bar 9. Some time ago, and I really don't think this came from published data, I surmised that the following would hold true:

Loss of bar 12 - between 54AHr and 55.5AHr
Loss of bar 11 - between 50AHr and 51.5AHr
Loss of bar 10 - between 46AHr and 47.5AHr
Loss of bar 9 - between 42AHr and 43.5AHr

Now a quick look through the a Capacity Loss WiKi doesn't find a great many bar loss reports complete with AHr readings, but I think the ones that are there validate this hypothesis pretty well:

Loss of bar 12 - between 53.92AHr and 55.76AHr
Loss of bar 11 - 51.6AHr (only one reported with AHr data)
Loss of bar 10 - between 46.9AHr and 47.5AHr
Loss of bar 9 - between 41.77AHr and 43.56AHr

I've been asked privately why I think my numbers for bar loss are pretty much spot-on, particularly when it comes to bar 9, and I replied that it was an educated guess (from some time ago, as it happens) that I felt was starting to be pretty well supported by the available data.

We also had private discussion on the AHr spread for the loss of bar 9 and, since I'd previously only provided the lowest and highest reported numbers at loss, I decided to at least find all the easier ones (from the Wiki) so you can see that it is literally all over the place within the spread I'd suggested.

So your takeaway should probably be to not see the highest reported loss and think, "I'm getting pretty close to 43.56AHr, I'll probably loose bar 9 at any time!", although it is good to see many more disappear above 43AHr than I had thought - in my mind I had the majority losing bar 9 between 42AHr and 43AHr and this sampling, small though it may be, does not appear to support that.

Edit: It is now even between those reporting loss above 43AHrs and those reporting loss below 43AHrs, with pretty much all the recent bar losses being at under 43AHr (well under in some instances).

I'll add more if and when I find them with deeper digging:

41.77AHr brettcgb
41.79AHr evoforce
42.10AHr sasun
42.15AHr theoldchum
42.17AHr mx5racer
42.22AHr pjospeh
42.34AHr jrtozer
42.35AHr 1932highboy
42.36AHr tombobca
42.39AHr stanton
42.53AHr pchilds
42.55AHr speedski97
42.61AHr 91040
42.65AHr valdemar
42.66AHr schamberlin
42.75AHr sksingh
42.75AHr fooljoe
42.83AHr baronvonshush *approximate - recorded about a week after actual bar loss
42.84AHr memoryisram
43.04AHr patricioEV
43.11AHr mwalsh *noted at conclusion of charging. 5 miles of driving later, the reading (true number?) was 42.76AHr
43.14AHr darkdave
43.16AHr silverleaf
43.19AHr ahagge
43.20AHr clarke
43.27AHr highdesertdriver
43.28AHr opencar
43.32AHr pipceil
43.39AHr jeremyz
43.34AHr tucsonleaf
43.41AHr ajay (Bay area resident!)
43.43AHr occ
43.47AHr tomt
43.51AHr gbshaun
43.56AHr jpvleaf

 

[mwalsh]

Last night I decided to sit down and see where I might end up, AHr wise, based on the losses during the past two summers.

From 6/22/13-8/26/13 I lost 1.53AHr
From 6/24/14-8/21/14 I lost a massive 2.24AHr
From 6/25/15-8/27/15 I have only lost 1.32AHr

Possible reasons for less loss in 2015 than in previous years:

P3227 update?
Not as many miles - only 500 total vs. ~1000 monthly
Cooler summer so far?

In previous years between late June and early October I had seen 2.49AHrs of loss (as of 10/8/13) and 3.66AHr of loss (as of 9/21/14). If 2015's loss continues as it has, that would be around 85% of 2013s. At that I am probably looking at a little over 2Ahrs between late June and early October, which should put me at around 42.5AHr. Since I've generally only taken one reading month, I don't know exactly when in October seasonal losses start to abate, but certainly by the time I've taken the 10/21 reading a tick upward is already occurring.

So loosing bar 9 above 42.5AHr some time before mid-October is a good bet. Below...not so much.

 

[mwalsh]

I've been trying to figure out in terms of AHr precisely where I might loose bar 9. Some time ago, and I really don't think this came from published data, I surmised that the following would hold true:

Loss of bar 12 - between 54AHr and 55.5AHr
Loss of bar 11 - between 50AHr and 51.5AHr
Loss of bar 10 - between 46AHr and 47.5AHr
Loss of bar 9 - between 42AHr and 43.5AHr

<Snip>

My own losses were:

Loss of bar 12 - 54.07AHr
Loss of bar 11 - 51.58AHr
Loss of bar 10 - 47.17AHr

So you can see that I lost bar 12 on the low side; but bars 11 and 10 both on the high side. I'm therefore predicting (or perhaps it's just wishful thinking) that, when the time comes, I'll also loose bar 9 on the high side, particularly as this may happen (if it's going to happen at all while the warranty is still valid) during the hotter months of a Southern California fall. Or at least that's the hope.

So...not going to loose bar 9 as high as I would have liked - 43.14AHr as of today and the darn thing is still there.

 

[kolmstead]

We know that when a single variable like charge level controls an indicator that the Leaf is totally predictable. Everybody sees LBW at 49 GIDS and VLBW at 24, and it's the same number every time.

So we suspect that the wide variability in ninth bar loss is caused by several variables interacting. My money is still on Ahr, temperature and time since some trigger capacity was reached. The result is a bell-shaped curve just as your data shows. Oh, and there is also the likelihood that P3227 diddled with the variables and limits.

Thanks for posting the numbers, Mike. Quick histogram seems to show max reports around 43.3, so you should be in the sweet spot right now. My battery was at 43.76 Ahr this morning.

So what we know is that Andy Palmer was doing the same thing to us that Mark did before... lying. Andy said that Nissan would replace the battery if it dropped below 70% capacity. In reality, the qualifying level is 64 to 66%. So that excludes four to six percent of the folks who were promised new batteries.

-Karl

 

[mwalsh]

So...not going to loose bar 9 as high as I would have liked - 43.14AHr as of today and the darn thing is still there.

Update:

100 degrees peak ambient indicated by my car's temperature gauge for the last two days, and I saw 7 battery temperature bars today for the first time for a long time (maybe ever?).

Battery temps right now, after being home and parked for 7 hours, are still in the 95-97 degree range across the four sensors.

42.99Ahrs.

Still at 9 bars.

 

[fooljoe]

Right behind you at 43.77 as of this morning - down .4 Ah in just about a week. My theory is that we've probably already crossed the magic number (maybe it's 44 Ah - who knows?) and are now in some sort of fixed waiting period. Perhaps the reason for the wide variation in reported capacity at time of bar loss is the rate of degradation at the time the threshold is crossed: Maybe those who lost the bar in the 43s were in a period of locally slow degradation, while those in the 42s were in a period of locally rapid degradation. I suppose if we knew the dates of occurrence for those who've reported dropping the 9th bar we could put this theory to the test - those who lost it in the summer should have lower reported capacity than those who lost it in the winter.

Anyway, if the theory is correct, that means that given the very high heat of late (I saw 7 heat bars for the first time the last two days as well) we'll probably not end up losing bar 9 until the capacity reads in the low 42s.

 

[mwalsh]

Maybe those who lost the bar in the 43s were in a period of locally slow degradation, while those in the 42s were in a period of locally rapid degradation. I suppose if we knew the dates of occurrence for those who've reported dropping the 9th bar we could put this theory to the test - those who lost it in the summer should have lower reported capacity than those who lost it in the winter.

That's actually a pretty good theory, and would go a long way towards explaining why every single recent bar 9 loss has been below 43AHrs.