Plug In America Tesla Roadster Survey

[djp]

Thanks Tom! I agree completely with your caveats, but it's still nice to see there's no evidence of degradation from frequent range charges.

 

[hcsharp]

...
This chart seems to be saying that vehicles that are charged in range mode frequently are more likely to report a higher capacity.

There are two caveats to this conclusion: 1) there aren't a lot of data points for frequent range mode charging, and 2) in my experience, the car underreports battery capacity when it's been a while since the last low-to-full range mode charge, so the low values may not reflect actual battery pack capacity.

I'm not sure I understand the second caveat. I think you're suggesting the results for the less-frequently range charged cars might be skewed lower because of under-reported capacity? It's funny because I would have thought the opposite based on my own observations, but I don't have the data or experience that you do. My car usually loses capacity in steps, and then slowly recovers part of it. The steps down almost always occur after recharging from low to full, meaning it was over-reporting capacity right before the low-to-full charge. The longer I wait, the more it recovers to the point of possibly over-reporting again. I've seen the same pattern on every Roadster (which is only 3) that I've looked at. Maybe it's not representative of the fleet?

 

[djp]

I'm not sure I understand the second caveat. I think you're suggesting the results for the less-frequently range charged cars might be skewed lower because of under-reported capacity? It's funny because I would have thought the opposite based on my own observations, but I don't have the data or experience that you do. My car usually loses capacity in steps, and then slowly recovers part of it. The steps down almost always occur after recharging from low to full, meaning it was over-reporting capacity right before the low-to-full charge. The longer I wait, the more it recovers to the point of possibly over-reporting again. I've seen the same pattern on every Roadster (which is only 3) that I've looked at. Maybe it's not representative of the fleet?

My take on it was that battery packs which are infrequently range charged are more likely to be out of balance, which skews their capacity numbers lower (and the opposite for packs that are frequently range charged).

 

[Pantera Dude]

Great work Tom! I'm glad we have people in this group that are so informative for the rest of us.
Thanks,
Art

 

[tomsax]

I'm not sure I understand the second caveat. I think you're suggesting the results for the less-frequently range charged cars might be skewed lower because of under-reported capacity? It's funny because I would have thought the opposite based on my own observations, but I don't have the data or experience that you do. My car usually loses capacity in steps, and then slowly recovers part of it. The steps down almost always occur after recharging from low to full, meaning it was over-reporting capacity right before the low-to-full charge. The longer I wait, the more it recovers to the point of possibly over-reporting again. I've seen the same pattern on every Roadster (which is only 3) that I've looked at. Maybe it's not representative of the fleet?

Very interesting. I have not heard anyone report that behavior before.

My take on it was that battery packs which are infrequently range charged are more likely to be out of balance, which skews their capacity numbers lower (and the opposite for packs that are frequently range charged).

That's the idea, but there may be issues other than balance at play. If the car doesn't see how much energy goes in/out of the pack at the top and/or bottom of the charge, it may lose information or confidence about the pack's total energy capacity.

 

[thefortunes]

Very interesting. I have not heard anyone report that behavior before.


That's the idea, but there may be issues other than balance at play. If the car doesn't see how much energy goes in/out of the pack at the top and/or bottom of the charge, it may lose information or confidence about the pack's total energy capacity.

Although I have never done a full Range charge (I have added 10-20 miles on occasion when I thought I might need a little extra range) I also have seen CAC lost in steps, with subsequent recovery.

In my case, I somehow (as of this morning) actually have a higher CAC then when I bought the Roadster (used) about 12,500 miles ago.

 

[dhrivnak]

Granted this is just one datapoint but it has been a few months since a range charge so I decided to give it a try and do a range charge and let the car sit after the charge completed. There was active balancing 4 hours after the charge but my CAC remained steady at 143.75. It has been 144.5 for several years then ticked up to 145.5 over the summer before dropping to 143. It is likely just noise but will find out this Spring.

 

[hcsharp]

Very interesting. I have not heard anyone report that behavior before.

My take on it was that battery packs which are infrequently range charged are more likely to be out of balance, which skews their capacity numbers lower (and the opposite for packs that are frequently range charged).

That's the idea, but there may be issues other than balance at play. If the car doesn't see how much energy goes in/out of the pack at the top and/or bottom of the charge, it may lose information or confidence about the pack's total energy capacity.

Here's a recent TeslaGLoP plot of my CAC shortly before it started to decline. You can see at least 2 distinct steps followed by gradual increase or nearly level.

Here's the last plot I've seen of Doug_G's car (borrowed image from the TeslaGLoP thread) which also has 2 steps. Not sure what it's done since then. I should note that Doug initially thought the first step was due to a firmware update but he also mentioned that he took a long trip about the same time. I'm not sure what he finally concluded. I know I didn't see any change in CAC after the same update.

 

[Doug_G]

Didn't really come to any specific conclusion at the time, but why not take another look?

Here's the full lifetime of my car (up until it was garaged for the winter):

The first, larger drop happened between February 9 and 15, 2012. The change occurs immediately after a road trip where I drained the car 100% to 10%.

There's a second, but relatively tiny, drop between April 4 and 5, 2013. The second drop also occurred during a road trip, which again drained the car from a Range charge to about 10% SOC.

Here's a closeup of the first drop showing the SOC during my road trip:

I believe those are the only two times I ever drained the Roadster to 10%. I've been to 20% many times.

So I interpret these changes as a recalculation due to seeing a large swing in charge level. Also it suggests that the CAC was being overestimated until I did that.

 

[Slackjaw]

Standard mode jumped 165 to 187 miles

For the past year I have been driving the Roadster every day, but not far, typically 20 miles/day. I maintain the SOC between 80 and 120 miles and always charge in Standard mode at 240/10A. After letting it charge some time from 6 pm to midnight I stop the charge via OVMS.

Three weeks ago the car had its 36,000 mile service. As far as I know there was no work done on the PEM or ESS.

Two weeks ago I did a Standard charge which went to 165 ideal miles, my wife took the car and came back with maybe 30 ideal miles left.

Last weekend I did a Range charge to 100% (sorry, didn't note down the reported miles), a friend - let's call him "Bob" - drove the car to a show and back and we came back with about 50 miles range.

Last night I accidentally left the car charging in Standard mode until it stopped (usually I don't let it get to 100%, see above). The indicator then showed 187 miles range, which is about as much as I was getting 3 years and 38,000 miles ago.

I suspect this is some sort of odd effect of the charging sequence, the range charge, and possibly the fact that most of the driving in the last 500 miles has been done by people with a lighter foot than me; both Bob and my Wife get close to ideal range from the car whereas I get maybe 80% on a good day.

I took a quick look at the diagnostic screen and the CAC is showing 137.26 at 38,140 miles driven. This compares to 143.35 at 31,623 miles after the previous (2-year) service. So it can't really be holding its range quite so well. I might let the car charge up again tonight and see what the range is in the morning.

Anyway, I know "if it seems to good to be true, it probably is" but was a nice surprise to see the VDS this morning!

 

[slcasner]

The range after a full Standard mode charge in my 2008 Roadster began to decrease over the past 6-12 months where the usage pattern was a 13-mile round-trip commute with a Standard mode charge whenever the battery got down to 50-60%. Then since April we have taken some long trips that required Range mode charging and drives of 150 miles. As I expected, the predicted range increased:

Before: 168 standard mode IM, CAC 140.12

After: 179 standard mode IM, CAC 149.15

The reason is that the battery management system does not know where the bottom of the battery really is unless you get close to it.

 

[tomsax]

Roadster Survey Update for v3 Battery Upgrade

I've updated the Roadster Battery Survey so owners can report upgrading to the v3 battery pack. It would be great to get reports on the original battery packs just before the swap and then get numbers on the v3 pack after the swap.

There are some other improvements: the update form loads faster, it's easier to find the vehicle list and data download links on the results page, a few other things.

If you haven't reported your car before, or if it's been a while since your last report, it would be great to get your numbers. I'd like to dive into the Roadster data to see if anything has emerged since the report in 2013, and more current data would be a big help.

First Time Report
Update to Previous Report

 

[strider]

Thanks for the reminder Tom! I updated my report.

 

[thefortunes]

Updated my info also. Thanks Tom!

59k miles and charges to 180 standard ideal miles.

 

[AEdennis]

Guys. I just updated my information on the study (and the S as well.) We've got good participation from the Roadster community, but feel that we can do more from S folks. That being said, don't forget to give your update. I wrote a blog post (and Teslarati article) reminding folks that it's out there.

 

[ggr]

Updated for our 3.0 upgrade.

 

[AEdennis]

Updated for our 3.0 upgrade.

Awesome... Actually thought of you too (as I was writing my post.)