Mileage vs Months Owned for Two Capacity Bar Losers

I took the data from the Wiki table of 2 bar losers for all entries that had both mileage and months owned and plotted them to see if there was any relationship. A least squares fit of the data gave a slope of -1568 and intercept of 39152. Correlation coefficient was only 0.17. Here is the graph, with the least squares fit line added manually:



It appears there is a weak correlation, such that every 1500 miles less driven gives one an extra month until losing 2 bars. This could be due to cycling losses, but an alternative explanation would be that those with higher mileage had their battery at a higher average SOC. Comments?

Stoaty said:

I took the data from the Wiki table of 2 bar losers for all entries that had both mileage and months owned and plotted them to see if there was any relationship. A least squares fit of the data gave a slope of -1568 and intercept of 39152. Correlation coefficient was only 0.17. Here is the graph, with the least squares fit line added manually:

It appears there is a weak correlation, such that every 1500 miles less driven gives one an extra month until losing 2 bars. This could be due to cycling losses, but an alternative explanation would be that those with higher mileage had their battery at a higher average SOC. Comments?

Click to expand...

Another Capacity bar thread???

Well, it is a neat graph, showing the two outliers, I commented on earlier:

Fun with scoring LEAF bars. If I got it right:

Take the list of two-bar losers that have complete entries.

Add months of ownership and thousands of miles driven.

The middle seven of the nine, all score between 29 and 36. A fairly tight grouping, IMO.

On the low side, the outlier #1, ev4me with 7,000 miles and 15 months, has a score of 22.*

Maybe someone should double check that second-party report, for accuracy, or special circumstances?

The outlier on the high side, is the one Texas LEAF, with a 39.5 score...

Click to expand...

http://www.mynissanleaf.com/viewtopic.php?f=27&t=9542" onclick="window.open(this.href);return false;

I don't think you can expect a correlation any stronger than this, unless you could control for what seems to be the dominant factor in battery capacity bar loss, the LEAFs ambient temperature history.

When the Pacific coast LEAFs begin to lose two bars, however far that is in the future, the correlation will almost certainly be much "weak"er.

Also, months since manufacture might be a better factor to use than months of ownership.

But you still have the problem of the likely huge "off the chart" variable, of whether the LEAF spent a few extra months in Long Beach, at 4 bars and 70 F during the post -tsunami delivery SNAFU in Spring 2011, or a few extra months "orphaned" and baking away in a Phoenix dealer's lot (kept at 100% charge?) over that Summer.

hoping not to draw a quiver-full of arrows in my ass, but how about the correlation between the number of cars with bars lost and the number of threads about it?

or the number of posts about it and the number of cars without bar loss after a year?

that is not to say that the work being done in AZ by a few owners is not valuable; just putting it into a larger framework.

edatoakrun said:

But you still have the problem of the likely huge "off the chart" variable, of whether the LEAF spent a few extra months in Long Beach, at 4 bars and 70 F during the post -tsunami delivery SNAFU in Spring 2011, or a few extra months "orphaned" and baking away in a Phoenix dealer's lot (kept at 100% charge?) over that Summer.

Click to expand...

I have wondered this since the first report of a "lost bar": how many (what percentage) of these cars were stuck in port that first Spring, and what condition were they stored in (most, if not all, had work done during this time--including paint, which requires heat, etc.). I don't personally know anyone who has one of these cars, which were all delivered a few months BEFORE mine--but it sure makes me curious.

Stanton said:

edatoakrun said:

But you still have the problem of the likely huge "off the chart" variable, of whether the LEAF spent a few extra months in Long Beach, at 4 bars and 70 F during the post -tsunami delivery SNAFU in Spring 2011, or a few extra months "orphaned" and baking away in a Phoenix dealer's lot (kept at 100% charge?) over that Summer.

Click to expand...

I have wondered this since the first report of a "lost bar": how many (what percentage) of these cars were stuck in port that first Spring, and what condition were they stored in (most, if not all, had work done during this time--including paint, which requires heat, etc.). I don't personally know anyone who has one of these cars, which were all delivered a few months BEFORE mine--but it sure makes me curious.

Click to expand...

Well, mine may be one.

My car has a March 2011 build date, and came with a tag in the owners manual pocket with a time stamp, only hours before the earthquake, IIRC. but I never found out when in the production/delivery process, that this card was issued.

So, I never have figured out whether my car sat in Japan, or in Long Beach, for the extra 5-6 weeks.

I have no significant loss of range yet, with 11,500 miles, under what may be slightly more "stressful" than average battery use. But if I do lose capacity in the future at an unusually fast rate, I will always wonder if it was because some Longshoreman charged it to 100%, just before it went into limbo.

Stoaty said:

I took the data from the Wiki table of 2 bar losers for all entries that had both mileage and months owned and plotted them to see if there was any relationship. A least squares fit of the data gave a slope of -1568 and intercept of 39152. Correlation coefficient was only 0.17. Here is the graph, with the least squares fit line added manually:

It appears there is a weak correlation, such that every 1500 miles less driven gives one an extra month until losing 2 bars. This could be due to cycling losses, but an alternative explanation would be that those with higher mileage had their battery at a higher average SOC. Comments?

Click to expand...

My take would be - the correlation is too weak to draw any conclusions.

evnow said:

My take would be - the correlation is too weak to draw any conclusions.

Click to expand...

Agreed. If you want to see an even weaker correlation (virtually none), here is a graph for one bar losers in Phoenix or Chandler, AZ. Correlation coefficient is .001:

Too small an "n" for any quantitative analysis, I'm afraid. A better pursuit might be a "qualitative study" of our 50 or so one and two-bar losers. In other words, interview people and get all that info into the Wiki, avg. temperature in their neighborhood, stored inside or outside, how long at the dealer outside, etc. We're really doing Nissan's job for them, I hope they are already doing this study.

JG