Ex Leaf-owner club (Why we gave up on Nissan Leaf)

[Valdemar]

I'm keeping mine for the time being as despite the reduced range I can still manage my 60-70 mile commute just fine and will be able to for a while thanks to now and soon be available DCQC spots along my typical route and charging stations at work, if run low. Just saw 50k miles yesterday, can still do 65 miles at 4.6m/kwh with 5 miles remaining on the GOM. The whole battery fiasco is very disappointing and I'd probably switch to another EV if I could make the numbers work. But no, it will likely cost me more than just driving the Leaf until I no longer can. Not too mention there are not too many great choices when it comes to EVs, or even PHEV. The Volt is good but I feel claustrophobic in it. The RAV4 EV many people are jumping to, sure, but in addition to boring looks and last generation body I don't have enough reasons to justify dropping more than $18000 on it over 3 years and give it away in the end when I will probably be able to get a new battery for my Leaf for around $10000, when (and if) it is going to be available for sale and have an option to use QC when I need it. The i3? Maybe, just because it is cool.

 

[RegGuheert]

I recommend reading the key conclusions on page 11 at the very least.

That survey contains no data which can shed light on the question of whether capacity degradation slows down as the battery degrades. On top of that, the results were written in December 2012 when we had even less data on how es battery degrades.

Let's go through this again: The BEST way to determine whether capacity loss slows, accelerates or stays constant, is to track capacity losses for INDIVIDUAL VEHICLES. But capacity loss varies dramatically with temperature, so we need to be able to capture data which is recorded under similar temperature conditions. Unfortunately, this eliminates nearly ALL of the data we have available to us.

But we do have one data point which sheds like on this question: cyellen's LEAF bar losses

Bar 1: June 8, 2012: 10200 miles
Bar 2: August 20, 2012: 11558 miles
Bar 3: July 31, 2013: 17500 miles
Bar 4: October 4, 2013: 18942 miles

The hottest month in Phoenix is July and 2012 was hotter than 2013.

Bar 1 to Bar 2: 72 days and ~1350 miles
Bar 3 to Bar 4: 65 days and ~1450 miles

So, for this particular Nissan LEAF, the fourth capacity bar was lost seven days FASTER than the second bar. This is true, even though the temperatures were cooler on 2013 when the fourth bar was lost. OTOH this LEAF was driven 7.5% farther between the bars 3 and 4 than between bars 1 and 2.

No, this data is not perfect and it would be great to have more points. But it IS real data and it directly contradicts unsupported statements such as this one:

Two bar loss or 80% means that your battery is about half way in time and cycles to 70%, a normal EOL percentage.

This statement implies that capacity loss between 80% and 70% occurs at HALF the rate that it did between 100% and 80%. But the data we have for this car indicates the following:

100% to 78.75%: 16.5 months and 11558 miles = 1.28%/month and 1.84%/1000 miles
78.75% to 66.25%: 12.5 months and 7384 miles = 1%/month and 1.69%/1000 miles

This latest calculation is not as interesting as the earlier one showing no slowing of degradation since this result contains seasonality which affects short-lived LEAF batteries in Pheonix much more significantly than it affects those of us who go through many seasons before losing our first bar. Still, even this distorted result indicates that hoping for capacity loss rates to drop by half after losing the first 20% will result in severe disappointment. Most likely, the time and miles between the losses of bars 3 and 4 will closely match those between bars 1 and 2, assuming similar temperature and driving profiles.

Feel free to dig up more high-quality data like Stoaty has collected from cyellen with bars two and four lost under very similar conditions. More is certainly better. But don't expect those results to be much different than this first data point.

 

[Valdemar]

I think generalizing a single case located in an extreme climate is just plain wrong.

 

[RegGuheert]

I think generalizing a single case located in an extreme climate is just plain wrong.

I find making a case on NO data is much more wrong.

That said, there is some support in the literature for the idea that very high calendar loss rates result in those rates growing faster as the cell degrades. It is certainly possible that is what is happening here and that losses in cooler climates will taper off, but we do not have that data, yet.

But, again, the literature indicates that many chemistries have linear calendar capacity loss rates. Expecting it to degrade much more slowly than linear is just plain wrong unless losses are dominated by cycling losses. But it appears very few early LEAFs will have more cycling loss than calendar loss.

 

[RegGuheert]

Here is a second data point: JasonT's LEAF bar losses

Bar 1: July 14, 2012: 25,500 miles
Bar 2: September 30, 2012: 32,500 miles
Bar 3: May 21, 2013: 50,600 miles
Bar 4: July 14, 2013: 53,600 miles

The hottest month in Tuscon is July, so having data ending and starting in the middle of July should be roughly equivalent conditions on either side.

Bar 1 to Bar 2: 75 days and ~7000 miles
Bar 3 to Bar 4: 54 days and ~3000 miles

So, for this particular Nissan LEAF, the fourth capacity bar was lost 21 days FASTER than the second bar. This is true even though this LEAF was driven 57% FEWER miles between bars 3 and 4 than between bars 1 and 2.

This data point clearly indicates accelerating capacity loss as this LEAF battery degrades.

 

[Valdemar]

For all I know both of these cars may have had bad batteries from the same batch.

 

[RegGuheert]

For all I know both of these cars may have had bad batteries from the same batch.

Could be. VIN numbers are 613 and 683. It could also be true that many/most/all batches of 2011/2012 LEAFs perform similarly.

That doesn't change the fact that the only data we have so far indicates that the rate of LEAF battery loss does NOT slow down as the battery degrades.

One thing to note about these two cars is that the miles driven were vastly different, yet the battery degradation was extremely similar. That gives an idea of just how dominant calendar losses are in that climate.

 

[WetEV]

This data point clearly indicates accelerating capacity loss as this LEAF battery degrades.

No, this is cherry picking. A "data point" is an anecdote.

I can do that too:

dsh bought a car with a one bar loss. The next two went very quickly, both in August 2012. Fourth bar loss was Feb 2014. A whole summer with no bar loss. In Phoenix, AZ.

But I don't think that this clearly indicates anything. A single data point, with very many undefined variables, shows nothing clearly.

 

[tomsax]

That survey contains no data which can shed light on the question of whether capacity degradation slows down as the battery degrades. On top of that, the results were written in December 2012 when we had even less data on how es battery degrades.

The survey does exactly what you want: it allows owners to enter data on their vehicles as often as they like to build a history of their vehicle. You can view the most recent record of every vehicle or drill down, look at the history of any vehicle, or download every record.

Although I haven't written up results recently, the full data set is available for download so anyone can analyze it.

 

[RegGuheert]

The survey does exactly what you want: it allows owners to enter data on their vehicles as often as they like to build a history of their vehicle.

First of all, I appreciate you collecting all of this data! But I don't think it answers this question currently. You posted yesterday that you just recently started collecting data from the instrumentation that some have. Given that, you may be just now starting to collect the kind of data which will be needed to answer this type of question in the future.

Although I haven't written up results recently, the full data set is available for download so anyone can analyze it.

Thanks! When I have a chance I will have another look to see if there is something useful there that could answer this specific question.

The best source of information could be the NREL (?) testing that was done with four LEAFs in Arizona, but, unfortunately, their testing spans several seasons so it is not useful to determine if the aging rates

 

[tomsax]

First of all, I appreciate you collecting all of this data! But I don't think it answers this question currently. You posted yesterday that you just recently started collecting data from the instrumentation that some have. Given that, you may be just now starting to collect the kind of data which will be needed to answer this type of question in the future.

Since I started the survey in October 2012, I've been collecting 100% and 80% Gid values.

Also since the very beginning, I've been asking people who have lost capacity bars to enter the date and odometer from when they lost each bar, exactly the type of data you asked for in your post. Unfortunately, not many people have filled out that data, but I can't force people to enter data.

I added the pack Ah value almost a year ago. The recent add was just SOH and Hx. From what I've read, SOH is just a more convenient substitute for 100% Gids and I haven't heard a good explanation for what Hx is, but I added it in case it turns out to be interesting.

So, I've been asking for pretty rich data from the very beginning and adding data as it has become available to the owners. Please stop claiming the survey data is useless until you've actually looked at it.

 

[RegGuheert]

dsh bought a car with a one bar loss. The next two went very quickly, both in August 2012. Fourth bar loss was Feb 2014. A whole summer with no bar loss. In Phoenix, AZ.

But I don't think that this clearly indicates anything. A single data point, with very many undefined variables, shows nothing clearly.

dsh had the P3227 reprogram done between the loss of bars 3 and 4 on July 14, 2013. Likely his experience indicates that the BMS firmware was reading extremely pessimistically before it was reprogrammed.

And I will grant that the old BMS results could be a real source of bad results if they are doing something (like hiding a bunch of capacity below VLBW). Unfortunately, if that is the case then we will not really be able to make ANY judgment about capacity rate changes on ANY pre-2013 LEAFs.

 

[RegGuheert]

Also since the very beginning, I've been asking people who have lost capacity bars to enter the date and odometer from when they lost each bar, exactly the type of data you asked for in your post.

There are no columns in the spreadsheet for this information, but I see a few notes that contain that information. I don't think you all for bar loss information for any vehicle. As you see, I have two such data points from the wiki here.

Please stop claiming the survey data is useless until you've actually looked at it.

I'm not claiming it is useless. As I said, perhaps it will contain enough information to answer this question in the future.

 

[hill]

Could be. VIN numbers are 613 and 683. It could also be true that many/most/all batches of 2011/2012 LEAFs perform similarly.

That doesn't change the fact that the only data we have so far indicates that the rate of LEAF battery loss does NOT slow down as the battery degrades.

One thing to note about these two cars is that the miles driven were vastly different, yet the battery degradation was extremely similar. That gives an idea of just how dominant calendar losses are in that climate.

Using just mnl members, we're talking a small sampling (btw, @40k+ miles, I'm still at 12 bars, with vin 659). Plus, don't you need to factor QC frequency besides high temps?
.

 

[tomsax]

You said this:

That survey contains no data which can shed light on the question of whether capacity degradation slows down as the battery degrades.

Then you said this:

Given that, you may be just now starting to collect the kind of data which will be needed to answer this type of question in the future.

Both statements are wrong and dismissive of the survey I designed to answer exactly these types of questions. I've been working on these questions for a year and a half, facilitating the collection of the largest publicly available body of data on the subject.

Have you contributed your data to the survey?

 

[RegGuheert]

You said this:

That survey contains no data which can shed light on the question of whether capacity degradation slows down as the battery degrades.

Then you said this:

Given that, you may be just now starting to collect the kind of data which will be needed to answer this type of question in the future.

Both statements are wrong and dismissive of the survey I designed to answer exactly these types of questions.

As I have stated previously and you quoted, your survey does not have the information needed to answer the question at hand.

I've been working on these questions for a year and a half, facilitating the collection of the largest publicly available body of data on the subject.

And I just thanked you for that in the above post.

Have you contributed your data to the survey?

Yes. How is that relevant?

 

[tomsax]

As I have stated previously and you quoted, your survey does not have the information needed to answer the question at hand.

I don't understand how you can say that without having looked at the data. I suspect analysis of the data could indeed shed light on that issue. I know you think the only way to answer the question is to track single vehicles, and that there is data in the survey to allow that, but I think other approaches could also work.

Getting more people to contribute their experience would certainly help. I plan to do a push to get more data, perhaps mid April, after I finish up some other work so I can spend more time on this survey. If you want to help get the word out, that would be great.

 

[RegGuheert]

I don't understand how you can say that without having looked at the data.

I have looked at the data. From my post above:

There are no columns in the spreadsheet for this information, but I see a few notes that contain that information. I don't think you have all four bar loss information for any vehicle.

I suspect analysis of the data could indeed shed light on that issue. I know you think the only way to answer the question is to track single vehicles, and that there is data in the survey to allow that, but I think other approaches could also work.

Perhaps, but please note that LEAFs are delivered from the factory with differing battery capacities. So if we ONLY look at remaining capacity after a time or some amount of miles, then we do not know how much capacity was lost for a given vehicle or for the group. Tracking a single vehicle and collecting capacity data at multiple points in time eliminates many factors which cannot be eliminated within a fleet.

Getting more people to contribute their experience would certainly help. I plan to do a push to get more data, perhaps mid April, after I finish up some other work so I can spend more time on this survey. If you want to help get the word out, that would be great.

Thanks again for collecting this data. I agree that it has many uses. And now that many of us have capacity meters to collect data, it will become more and more useful.

FYI, I've just added another data point to your database (the first from me which includes battery capacity).

 

[dhanson865]

As I have stated previously and you quoted, your survey does not have the information needed to answer the question at hand.

I don't understand how you can say that without having looked at the data. I suspect analysis of the data could indeed shed light on that issue. I know you think the only way to answer the question is to track single vehicles, and that there is data in the survey to allow that, but I think other approaches could also work.

Getting more people to contribute their experience would certainly help. I plan to do a push to get more data, perhaps mid April, after I finish up some other work so I can spend more time on this survey. If you want to help get the word out, that would be great.

the entry containing

01/29/2014 13:51 2013 United States 10/2013 12/2013 t MI Bloomfield Hills 48304 1554

contains a value of 3 for bars remaining but the user states the car performs like new / little to no battery loss. I'm not sure why that one isn't a 12 for bars remaining.

 

[RegGuheert]

Plus, don't you need to factor QC frequency besides high temps?

I would say that is included in the calculation. If you are QCing your LEAF more often because your battery is degraded and that causes it to degrade more quickly, then it still is degrading more quickly. In fact, that effect is very similar to the fact that as the battery capacity drops, each trip of the same length results in larger and larger cycles on the battery. (That was Reason #4 in my list of 10 reasons why I think capacity losses will not slow as many expect. I should now add to the list that there are now two data points in the wiki indicating flat-to-increasing rates of degradation for LEAF batteries.)