Capacity Loss on 2011-2012 LEAFs

[RegGuheert]

I don't think that anyone expected to see batteries hitting EOL after 2.5 years in Phoenix, and I recall only one source using a 2-year life target. More importantly, based on Arrhenius law and the showing in Phoenix, one can expect first batteries hitting EOL after four years in SoCal, five years in the Bay Area and six years in Seattle. It would appear that very few will see 20% capacity loss in ten years, which part of the original advertising at market launch.

To be fair to Nissan, they DID tell all of us in our owner's manuals to expect to only have 80% capacity left after five years.

Frankly, I think we should acknowledge that the FIRST LEAFs hit EOL after only 1.5 years in Phoenix, not 2.5 years. That likely means that your timeline for when we should see first articles at EOL in different climates may need to be accelerated a bit.

That said, let's all recognize that we are talking about the edge of the bell-shaped curve here. I'm wondering what percent of the LEAFs in service in Phoenix reached EOL on their batteries after 2.5 years. Is it 5%, 10%, or more like 25%? I don't know. Let's assume 10% for discussion's sake. That implies about 10% of LEAFs in Seattle will reach battery EOL in Seattle in about six years.

So when can we expect the median LEAF owner to experience EOL in these various locations? While I do not think we yet know the 50th percentile point for Phoenix, yet, we likely have enough data from Phoenix first-bar loss data to start to get an idea of the standard deviation of the data. From that we should be able to estimate when the EOL median should occur for Phoenix. For this discussion, I'm going to guess that it is around 3.5 years. Based on that, we might expect the median EOLs in various locations to look like this:

Phoenix: 3.5 years
SoCal: 5 years
Bay Area: 7 years
Seattle: 9 years

Of course, that implies that about HALF of the LEAFs' batteries will last longer than this amount of time. Some will undoubtedly last much longer.

It seems that most Phoenix battery deaths will be covered by the new warranty and quite a few in SoCal, but very few in cooler climates will be covered since the early units will likely exceed the mileage limitation.

 

[planet4ever]

That said, let's all recognize that we are talking about the edge of the bell-shaped curve here.

Not to get all technical here, but isn't this more likely to be a Poisson distribution rather than a normal (Gaussian) distribution? For example, having a much longer tail on the back end than the front end.

Ray

 

[RegGuheert]

That said, let's all recognize that we are talking about the edge of the bell-shaped curve here.

Not to get all technical here, but isn't this more likely to be a Poisson distribution rather than a normal (Gaussian) distribution? For example, having a much longer tail on the back end than the front end.

Ray

Good point, Ray! Agreed!

 

[surfingslovak]

That said, let's all recognize that we are talking about the edge of the bell-shaped curve here.

Not to get all technical here, but isn't this more likely to be a Poisson distribution rather than a normal (Gaussian) distribution? For example, having a much longer tail on the back end than the front end.

It was not my intention to turn this into a lecture in statistics, I simply expressed an opinion. I intended that post to be thought-provoking. Why? Well, I believe that the packs are declining faster than anticipated in Phoenix. Remember the glidepath to 70% in five years Mark Perry mentioned last summer? I suspect that this is not going to happen, and inferring from the Arrhenius equation, we should see faster than anticipated declines in other locales as well.

I was particularly surprised to see a case of capacity bar loss in the Bay Area at 21K miles. This owner has a garage with Western exposure, much like Scott did, so there could be some similarity. Based on the modeling done last year, it would have been a stretch to think that anyone in that locale would qualify for the new capacity warranty. Now it looks like there might be several candidates. When you look at TomT's case, he could be a candidate also, even before his four-year lease is up. You are right, I don't have the numbers all lined up this time, but I have a sneaking feeling that something does not add up.

Yes, there was a pack replaced in late 2012 in Phoenix (Skywagon). However, that battery might not have reached EOL criteria as defined by the new warranty. I'm not sure if you are counting Scott's LEAF as the first EOL battery in Phoenix. Please recall that Scott was down three bars, and not four. How long it would have taken him to lose the forth bar? That's anyone's guess. Probably not very long, but his pack was not down four bars. We only saw that this summer. Scott also put substantially more miles on his LEAF than the average driver would.

I think the aging model Stoaty developed to reflect the data shared by a Nissan engineer with TickTock predicted that most packs will reach EOL after about five years in Phoenix. While it's premature to say much about that, I believe that we are seeing a faster decline, and mentioned that in another thread as well.

 

[RegGuheert]

Please recall that Scott was down three bars, and not four. We only saw that this summer.

According to the Wiki, Scott Yarosh lost his fourth bar on August 31, 2012 after only 14 months of ownership. You can see in the Wiki that another owner also lost their fourth bar on September 26, 2012.

Given that the vast majority of the degradation occurs during the summertime, you could say that Scott's car reached EOL after two years' worth of summers in Phoenix. You could equally say that many of the others we are seeing this year succumbed after three years' worth of summers.

I think the aging model Stoaty developed to reflect the data shared by a Nissan engineer with TickTock predicted that most packs will reach EOL after about five years in Phoenix. While it's premature to say much about that, I believe that we are seeing a faster decline, and mentioned that in another thread as well.

I have contended from the beginning of the work on that model that I was skeptical of the notion that calendar loss rates would decline as time went on. This was based on (old) papers which had tested calendar losses on other chemistries that showed accelerating calendar losses. Still, it is not clear whether this effect occurs only at very high aging rates because I have never seen calendar aging tests that were conducted for multiple years at lower temperatures. I think everyone is learning as things play out with the LEAF.

Really it is too bad that the BMS in the Volt apparently hides ongoing degradation. It would be interesting to have another set of data to compare against. (The cycling on the Tesla packs is so low and the chemistry is so different that the comparison is harder.)

 

[WetEV]

I think everyone is learning as things play out with the LEAF.

+1

I've always viewed the Leaf as a large scale experiment. We never can know how a lot of BEVs will work in the real world until we build and drive a lot of BEVs in the real world.

 

[mwalsh]

I've always viewed the Leaf as a large scale experiment. We never can know how a lot of BEVs will work in the real world until we build and drive a lot of BEVs in the real world.

I agree to an extent. BUT we trusted Nissan to have done their homework properly. In no small part because we saw evidence of them having done so. I guess if there is one thing I've learned from this, it's not to have a lot of faith in accelerated stress testing. :shock:

 

[mkjayakumar]

Can't wait to lose another bar, eh?

 

[TomT]

Well, I was there in just over two years...

To be fair to Nissan, they DID tell all of us in our owner's manuals to expect to only have 80% capacity left after five years.

 

[Stoaty]

I think the aging model Stoaty developed to reflect the data shared by a Nissan engineer with TickTock predicted that most packs will reach EOL after about five years in Phoenix. While it's premature to say much about that, I believe that we are seeing a faster decline, and mentioned that in another thread as well.

One big problem with the battery aging model is that it assumes Nissan has a good idea of how fast degradation would happen, which apparently is not the case. A couple of other interesting points:

1) The "newly developed battery" mentioned in the section of the Wiki on the Battery Aging Model found a doubling of capacity loss for every 6.85 degrees C. (not 10 degrees C.)
2) When the aging factors were calculated for the battery aging model, they had to be "scaled" to fit Nissan's curves. For example, Phoenix went from 1.8 to 1.5 when scaled (relative to LA Civic Center). It may be that the unscaled factors should have been used, and perhaps another version of the model should be created which uses the unscaled factors (if/when I have time and energy to do this)

 

[DTros]

Lost my 2nd capacity bar today at 27,435 miles - now down to 10. Close to 2 months and 2,504 miles from when I lost the first one.

 

[mksE55]

Lost my 2nd capacity bar today at 27,435 miles - now down to 10. Close to 2 months and 2,504 miles from when I lost the first one.

sorry to tell you but you will lose the 3rd bar in about 2-3 months more and another 3000 miles , it seems to decline slightly faster with 2-3, I am still waiting on the 4 bar to drop but by my mileage it will be soon.

 

[jhm614]

Lost my 2nd capacity bar today at 27,435 miles - now down to 10. Close to 2 months and 2,504 miles from when I lost the first one.

sorry to tell you but you will lose the 3rd bar in about 2-3 months more and another 3000 miles , it seems to decline slightly faster with 2-3, I am still waiting on the 4 bar to drop but by my mileage it will be soon.

Based on my experiences in the Dallas area, DTros probably won't lose bar 3 until late next Spring. (I lost bar 1 late spring 2012, bar 2 summer 2012, bar 3 late spring 2013, bar 4 this summer.) Of course, it may depend on how long the heat lasts this month...

j.

 

[mkjayakumar]

So all this talk about degradation initially being higher and then slowing down, is all just hog-wash..

 

[mksE55]

Lost my 2nd capacity bar today at 27,435 miles - now down to 10. Close to 2 months and 2,504 miles from when I lost the first one.

sorry to tell you but you will lose the 3rd bar in about 2-3 months more and another 3000 miles , it seems to decline slightly faster with 2-3, I am still waiting on the 4 bar to drop but by my mileage it will be soon.

Based on my experiences in the Dallas area, DTros probably won't lose bar 3 until late next Spring. (I lost bar 1 late spring 2012, bar 2 summer 2012, bar 3 late spring 2013, bar 4 this summer.) Of course, it may depend on how long the heat lasts this month...

j.

the more bars you lose the more you charge and usually the higher charge to discharge rate , all this will speed things along. if he was charging 80% to save battery, you loose this and increase battery degradation. seems faster for everyone. I dont know how you got such a gap from 2-3 everyone i see its significantly shorter. remember the first bar is 15% almost 3 times the next bars. but heat and charging and driving style do play a roll. good luck either way.I am tracking my Amp/hr loss and it seems to be falling at 1.5amp/hr loss per month. but in 3 weeks I can do a 2 month average.

 

[mksE55]

So all this talk about degradation initially being higher and then slowing down, is all just hog-wash..

it definitely is for the hot climate locations.

 

[jhm614]

seems faster for everyone. I dont know how you got such a gap from 2-3 everyone i see its significantly shorter. remember the first bar is 15% almost 3 times the next bars. but heat and charging and driving style do play a roll.

Check the wiki -- it looks like the gap between 2BL and 3BL is all over the place. Mark1313, Cyellen, DarkDave, TickTock, Turbo2ltr and jhm614 all had at least 10.5 months between 2BL & 3BL. Other folks only had 1 or 2 months. I think a big part of it depends on when you lose the first bar and how much seasonal heat is left... somebody needs to chart that using the wiki data.

 

[TomT]

Mine has been pretty linear so, yes, I'd have to say so...

So all this talk about degradation initially being higher and then slowing down, is all just hog-wash..

 

[highend]

Maybe there is possiblity (DIY) to allow air conditioning to cool batteries. Why Nissan didn't do that? Some air flow tunels to cool battery pack using onboard air conditioning. Should be cheap to adopt & efficient.
It could be switched on manualy or automaticaly using temperature sensors..... Of course air conditioning eats range (but not so much), anyway this could save battery life in hot climates.

 

[siai]

I would agree with the last poster about providing conditioned air to the battery. I have both a LEAF and a I-MiEV. On the I-MiEV the components to do this are installed on some of the vehicles. It is a easy fix to direct air from the A/C system into the battery when the system is running and the amount can be controlled by the driver. It adds very little loss in range because if it is hot out, you usually are using A/C anyway. On hot summer days I drive the I-MiEV and can drive with a traction battery enclosure temperature the same as the cabin temperature. Like the LEAF, if I am charging I can leave the climate control system operating and continue to cool the battery. Even when parked, turned off, the mass of the cooler pack will take longer to rise to a dangerous level then the mass of an already hot pack. And as soon as I start driving, I am starting to cool the batteries back down.