Tuning the Battery Aging Model

[Stoaty]

The Ridgecrest data vehicle does not have the P3227 update and thus, should not be included.

Agreed.

P3227 update - No
Stoaty is not looking for non-P3227 update vehicles (where applicable).

True, but I did soften that to take data without P3227 just for interests sake. They won't be included in any analysis, though. I don't plan to include the 2013 models either, for two reasons:

1) They don't have a long enough time to be meaningful
2) Changes to the electrolyte for 2013 batteries may affect heat resistance

 

[DaveEV]

True, but I did soften that to take data without P3227 just for interests sake. They won't be included in any analysis, though.

Cool! I'll edit my post.

I don't plan to include the 2013 models either, for two reasons:

1) They don't have a long enough time to be meaningful
2) Changes to the electrolyte for 2013 batteries may affect heat resistance

I would be interesting to start gathering '13 data at least from 6 months+ simply to see if there is a difference or not...

 

[JeremyW]

Manufactured - 02/12
Delivered - 6/23/12
P3227 update - 7/13 (will try to find date and update post)
Location - Folsom, CA for 9 months, Pasadena, CA since March this year.
Miles/kwh - 3.9
Odometer - 18119
Capacity - 57.51
Date - 9/15/13
Parked in sun - 3 days/week (avg but varies substantially week to week)

Gids at 80% charge (this morning with the Brusa disabled)- 203. Last 100% charge two weeks ago showed 236.

 

[Stoaty]

I would be interesting to start gathering '13 data at least from 6 months+ simply to see if there is a difference or not...

Yes, I have two entries for 2013. If they (2013 owners) keep reporting and others chime in we should have a good idea in a year or two.

Jeremy W:

I used Van Nuys as the best proxy for Pasadena. Sacramento is a bit cooler, only made 1.4% difference in prediction when I put in that city (although it did add an extra year of predicted life if you had stayed in Sacramento).

Predicted Loss - 13.27%
Actual Loss - 13.19%
Years to EOL - 5.3 (sorry about that)
Mileage at EOL - 78,000

PS So far we have 13 readings from 2011-2012 Leafs that have had the P3227 update. I took the difference of (Actual Loss) - (Corrected Predicted Total Loss) and then took some readings on the differences:

Mean of differences - 2.07%
Standard deviation of differences - 1.92%
Median difference -2.4%

 

[91040]

Manufactured - 3/11
Delivered - 5/6/11
P3227 update - 8/12/13
Location - Sunland, CA
Miles/kwh - 4.8
Odometer - 54725
Capacity - 52.1323 Ahs
Date - 9/3/13
Parked in sun - 2 to 3 days/week (avg but varies substantially week to week)

 

[Stoaty]

Final update for the weekend for the Battery Aging Model. Version 0.95d is here:

https://dl.dropboxusercontent.com/u/48149991/Leaf%20Battery%20Degradation%20Model%20Version%20095d.ods" onclick="window.open(this.href);return false;

Changes:

--a few minor cosmetic changes
--Calibration tab - added a couple more reports (thanks!)
--Calibration tab - for Leafs with P3227 update, added mean, standard deviation and median of difference between predicted and actual capacity loss and the Predicted years to End of Life
--Simplified a couple of formulas to make them more understandable to someone (e.g., me) who looks at them in the future

Of interest is the End of Life prediction for the 14 Leafs we have data for so far:

Mean - 8.49 years
Standard deviation - 5.30 years (i.e. the numbers are all over the map depending on climate, amount of driving, etc.)
Median - 6.55 years

 

[bradbissell]

I'm more than happy to keep chiming in on my 2013. Glad to see I'm way under the curve. I only have a 2 yr lease, but I will keep updating up till I give it back.

 

[JeremyW]

Years to EOL - 5.3 (sorry about that)

Boooo!

Thanks for all of your hard work Stoaty. Depending on my work situation I may be back up in Sacramento sometime next year or so. We'll see. I don't think I'll be leveling out, and I predict my EOL will be around year 4. Well... hope more than predict. :roll:

 

[tivollix]

Here is my weekly battery capacity info...
https://dl.dropboxusercontent.com/u/6002074/battery_log.ods

 

[Armand]

Manufactured - 5/11
Delivered - 8/27/11
P3227 update - 7/12/13
Location - San Mateo, CA. 4 days a week in Milpitas during work hours parked in the sun
Miles/kwh - 4.2
Odometer - 31,723 miles
Capacity - 56.52 AHr
Date - 9/16/13
Parked in sun - 7 days/week at home and work.

 

[TickTock]

Manufactured - 4/11
Delivered - 5/31/11
P3227 update - 7/10/13
Location - Chandler, AZ during work hours
Miles/kwh - 4.4 (as reported on instrument cluster - weighted average over life)
Odometer - 31,056 miles
Capacity - 46.69 AHr
Date - 9/16/13
Parked in garage at home and in shade at work

 

[surfingslovak]

Parked in garage at home and in shade at work

Good to see this dataset! Did the swamp cooler have any impact in your estimation?

 

[Stoaty]

Armand:

Predicted Loss - 14.25%
Actual Loss - 14.69%

TickTock:

Predicted Loss - 24.60%
Actual Loss - 29.52%

Comment: the tweaks I made to the model over the weekend (using unscaled aging factor for climates hotter than L.A., correction to cycling loss as battery ages) have improved the predictions, but still can't account for the really hot areas.

 

[TickTock]

Parked in garage at home and in shade at work

Good to see this dataset! Did the swamp cooler have any impact in your estimation?

Almost impossible to tell. The rate of degradation does appear to be slowing, but cannot tell if that is the swamp cooler or normal tapering off as per Nissan prediction. Also, I got a slight bump up with the P3227 update.

 

[surfingslovak]

Almost impossible to tell. The rate of degradation does appear to be slowing, but cannot tell if that is the swamp cooler or normal tapering off as per Nissan prediction. Also, I got a slight bump up with the P3227 update.

Interesting, thank you. Any idea if the battery might be heating up noticeably if the LEAF is driven over hot pavement, which soaked the sun for some time?

 

[ttweed]

Manufactured - 03/11
Delivered - 4/1/11
P3227 update - 8/20/13
Location - La Jolla, CA
Miles/kwh - 4.3
Odometer - 35,499
Capacity - 52.99 AHr
Date - 9/12/13
Parked in sun - 7 days/week [actually, this means "parked outside" every day--the sun does not shine a majority of the days here, since we are in the Lo-Cal coastal microclimate that involves fog or overcast skies probably 70% of the time.]
Parked overnight in an uninsulated garage for charging. Never seen 7 temp bars (or 4), never QC'd (except for a 5-minute test to make sure the port worked OK).

TT

 

[TickTock]

Almost impossible to tell. The rate of degradation does appear to be slowing, but cannot tell if that is the swamp cooler or normal tapering off as per Nissan prediction. Also, I got a slight bump up with the P3227 update.

Interesting, thank you. Any idea if the battery might be heating up noticeably if the LEAF is driven over hot pavement, which soaked the sun for some time?

Not much. I haven't rigorously logged this, but the most I see is about 5 degrees increase. The drive home (~30 minutes) is too short to have much impact on the battery temperature. Remember this thread? This is the temperature as measured just outside the battery (probe was placed between the battery housing and the plastic undercarriage liner. 7 degrees climb on the drive home. The battery, presumably, is even more insulated as can be concluded based on how rapidly the temperature dropped back down once I had it parked in the garage (not air-conditioned at that point). In fact, I would submit that the battery temperature was 107 (compared to the start temp of 103) since it almost immediately dropped to 107 once I parked and then gradually cooled the rest of the way. So 4 degrees increase in battery temp.

 

[Devin]

Manufactured - 5/11
Delivered - 6/16/11
P3227 update - 7/23/13
Location - Van Nuys, CA
Miles/kwh - ~4.4
Odometer - 26,328mi
Capacity - 56.09
Date - 9/17/13
Parked in sun - 5 days/week (75% Burbank, 25% Woodland Hills)

Garage - Unventilated & uninsulated, exposed on South, East, and North sides (faces North)
Capacity bars indicated - 12
DC Charge Count - 18
AC Charge Count - 2,856

 

[Stoaty]

ttweed:

Predicted Loss - 16.63%
Actual Loss - 20.02%

Devin:

Predicted Loss - 15.89%
Actual Loss - 15.34%

Comment: Actual Loss minus Predicted Loss ranges between -0.66% and 3.87% for 15 reported readings. There are two that are >4% one from Dallas, the other from Phoenix. I decided not to include in the analysis the Leaf in Montreal that spent 14 months in California, since I don't know where exactly or how to adjust the aging factor.

Including the two outliers the 17 reported readings have the following statistics:

Mean of Actual Loss Minus Predicted Loss = 1.98%
Standard Deviation of Actual Loss Minus Predicted Loss = 1.95%

 

[Stoaty]

Here is version 0.95e of the Battery Aging Model:

https://dl.dropboxusercontent.com/u/48149991/Leaf%20Battery%20Degradation%20Model%20Version%20095e.ods" onclick="window.open(this.href);return false;

Changes:

Calibration tab - added more data from reports over the last several days
Calibration tab - changed Remaining Capacity column in table to Corrected Remaining Capacity, so that the calculations of predicted Gids at 80% and 100% charge for each entry is correct
Minor formatting changes