Capacity Loss on 2011-2012 LEAFs

[vegastar]

My gids at 80% charge are clearly going up now that the weather has cooled:

Other thing that I noticed is that this gain comes mostly from the low voltage gids. For example, yesterday I got 59gids@360V, but in September I had 42 gids at the same voltage level.

In energy from the wall I'm not finding any recovery from start voltage to end voltage of charge, so it is a problem with the instrumentation.

 

[Joeviocoe]

There are clearly instrumentation accuracy issues with the Leaf so the two most reliable methods of measuring battery capacity loss are 1) a range test at constant speed all the way to turtle or shutdown and 2) measuring charging energy at the wall to charge from turtle or shutdown to 100% charge. LBW and VLBW seem to be moving targets--more and more range is available after these alarms as time goes on and odometer readings increase.

It is disappointing to read disparaging remarks about the Phoenix range test. The folks who organized this test (especially Tony, the test coordinator) put a lot of effort into making the tests consistent between vehicles and into gathering meaningful data from each vehicle. As one who participated in the test (not involved with the planning), I want to express my appreciation for their efforts. I believe the test results are valid and I don't see a way to gather more accurate data without having access to the positive and negative battery terminals and using laboratory-grade instruments during a range test.

Gerry

I give an A for effort to Tony and all involved. But your "two most reliable methods of measuring battery capacity loss" are wrong. I understand why it is financially not practical to test so many Leafs on a Dyno running the EPA's 5 cycle drive map test.... but THAT is the most reliable method. There are unavoidable errors when conducting road tests. And range can easily fluctuate by 10% or more just by driving style differences.

It appears that Tony was meticulous enough to get 'reasonable' results. But a Dyno would be 'definitive'.... especially since A DYNO running the EPA test is exactly what Nissan promised to Leaf buyers with the Monroney Sticker range. Not Nissan's road tests, but Nissan conducting a certified Dyno map to EPA specifications.

What Tony did is quite admirable... and does well to convince fellow Leaf owners of the problem, especially since Nissan won't share any of their test data. But Nissan does not have to take Tony's road tests seriously if they don't want to.

 

[TonyWilliams]

I give an A for effort to Tony and all involved. But your "two most reliable methods of measuring battery capacity loss" are wrong. I understand why it is financially not practical to test so many Leafs on a Dyno running the EPA's 5 cycle drive map test.... but THAT is the most reliable method. There are unavoidable errors when conducting road tests. And range can easily fluctuate by 10% or more just by driving style differences.

I've got to say, this "driving style difference" is troubling that any person who read the paper on our range test would conclude.

What style was used with the cruise control locked on to 64mpg indicated, each car over the same route, that could cause a 10% error? It isn't a human style, that's for sure.

It appears that Tony was meticulous enough to get 'reasonable' results. But a Dyno would be 'definitive'.... especially since A DYNO running the EPA test is exactly what Nissan promised to Leaf buyers with the Monroney Sticker range. Not Nissan's road tests, but Nissan conducting a certified Dyno map to EPA specifications.

We weren't trying to duplicate the 73 mile EPA range that your Dyno test might more easily compare. But, sure, if you're going to pay for the equipment, we will dyno test as many cars as you can afford !!!

Honestly, you're not the first person to propose this, and there are many posts about it. I agree, great idea, but the Mark Larsens and Russ Finleys of the world would find fault. Hence, it doesn't prove anything more to the folks this matters to; it just costs more.

For a court situation, absolutely, we want all the tests, including Nissan's.

What Tony did is quite admirable... and does well to convince fellow Leaf owners of the problem, especially since Nissan won't share any of their test data. But Nissan does not have to take Tony's road tests seriously if they don't want to.

Oh, let's see, we're up to six of twelve cars tested are no longer owned (mine was traded in, though!). I'd say Nissan listened, whether they care to dispute the results, or not (by the way, they haven't).

 

[TonyWilliams]

We drove a brand new LEAF (RedXXX) on Nov 4, 2012 in Phoenix with only 138 miles on the odometer (and a recent production date) to run the exact course and parameters in similar weather as the Sept 15, 2012 LEAF range autonomy demonstration. This was a shortcoming of our original test, with no "control" car that was capable of producing 84 miles of range autonomy at 4 miles/kWh.

While I suspect both Yanquentino and Russ Finley will have any number of exceptions with this, the results didn't surprise me, or those who are intimately affected or impacted by these issues.


83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle

Start time: 12:58 pm, November 4, 2012
Start battery stored energy: 265 Gids / 94.3%
Start pack volts: 393.5 (4.1 per cell average)
Start SOC: 91.4%
Start GOM: 103
Start temperature: 6 bar segments
Economy: 0 miles/kWh (reset)

End of test battery stored energy: 21 Gids / 7.47%
End time: 2:21 pm

TEST COMPLETE. The car was driven an additional 4 miles to a charger when these readings were recorded:

Gids: 11 / 3.9% remaining
Pack volts: 317.5 (3.3 volt average per cell)
SOC: 4.3%
GOM: "---" (normal for "Very Low Battery")
Battery temperature: 7 bar segments
Economy: 4.3 miles/kWh

Time (MST) - Temp - Dew Point- Hum- Press - Visibility - Wind - Wind/Gust -- Precip -- Conditions
12:51 PM 84.0 °F 30.9 °F 15% 30.00 in 10.0 mi WNW 3.5 mph - N/A Partly Cloudy
1:51 PM 88.0 °F 30.0 °F 12% 29.97 in 10.0 mi Calm Calm - N/A Partly Cloudy
2:51 PM 89.1 °F 28.9 °F 12% 29.96 in 10.0 mi Calm Calm - N/A Partly Cloudy

 

[cwerdna]

I give an A for effort to Tony and all involved. But your "two most reliable methods of measuring battery capacity loss" are wrong. I understand why it is financially not practical to test so many Leafs on a Dyno running the EPA's 5 cycle drive map test.... but THAT is the most reliable method. There are unavoidable errors when conducting road tests. And range can easily fluctuate by 10% or more just by driving style differences.

I've got to say, this "driving style difference" is troubling that any person who read the paper on our range test would conclude.

What style was used with the cruise control locked on to 64mpg indicated, each car over the same route, that could cause a 10% error? It isn't a human style, that's for sure.

It appears that Tony was meticulous enough to get 'reasonable' results. But a Dyno would be 'definitive'.... especially since A DYNO running the EPA test is exactly what Nissan promised to Leaf buyers with the Monroney Sticker range. Not Nissan's road tests, but Nissan conducting a certified Dyno map to EPA specifications.

We weren't trying to duplicate the 73 mile EPA range that your Dyno test might more easily compare. But, sure, if you're going to pay for the equipment, we will dyno test as many cars as you can afford !!!

Besides the amount of dyno time and cost needed to perform the 5 or 2 cycle test (from http://www.mynissanleaf.com/viewtopic.php?f=24&t=8885" onclick="window.open(this.href);return false;, I'm unclear if the 73 EPA miles derived from 2 or 5 cycles), there is a HELL of a lot more room for error for someone who's never performed the test before than cruise control at a fixed speed. See the graphs at http://www.fueleconomy.gov/feg/fe_test_schedules.shtml" onclick="window.open(this.href);return false; for the speeds at a given time. One can read http://www.caranddriver.com/features/the-truth-about-epa-city-highway-mpg-estimates" onclick="window.open(this.href);return false; to learn more about how the tests are conducted (this was written WAY before the Leaf). To quote from there:

Every last detail of an EPA fuel-economy test has specific rules—there’s even a set volume of air that a fan must blow under the car’s raised hood. “Driving” the test by matching the red line is tricky and takes a sensitive foot. __________________________________________________________________________________

After a vehicle is strapped down on a dyno, the staff punches in coefficients that allow the dyno rolls to simulate real-world factors, such as wind and road friction.

One of the EPA’s six drivers is behind the wheel of the test car. With an average of 20 years’ experience, they have extremely fine-tuned throttle and brake-control skills. They “drive” by following a precise red line of speed versus time that’s displayed on a monitor hanging just in front of the windshield. Using the gas pedal and the brakes, the driver attempts to match the red line with the car’s wheel speed, which is shown in white. We got into a test car and tried it, and indeed, it’s very difficult to maintain the speed of the tests, particularly when it dithers in the single digits and a brush of the throttle can send the white line careening off-course. If the speed deviates from the test cycle by more than 2 mph, the results are thrown out...

http://blogs.cars.com/kickingtires/2012/11/hyundai-kia-mileage-mishap-how-it-happened.html" onclick="window.open(this.href);return false; and http://www.freep.com/article/20121103/BUSINESS01/121103009/How-the-Hyundai-Kia-mileage-rating-mistakes-happened?odyssey=nav|head" onclick="window.open(this.href);return false; talk about how Hyundai's ICEV test results ended up getting inflated.

I just stumbled across https://hyundaimpginfo.com/overview/new-testing-process" onclick="window.open(this.href);return false; and https://hyundaimpginfo.com/resources/details/coastdown-facts" onclick="window.open(this.href);return false;. There's too much to quote/repost here.

 

[RegGuheert]

83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle

And now we know...the rest of the story.

TEST COMPLETE.

Thanks, Tony, for going the extra mile to wrap this up neatly.

 

[downeykp]

Nice work Tony! There are many of us out here in MNL land who appreciate all you have done to provide some clarity to a problem that Nissan created.
I hope we have not lost you as a contributor now that you are a Rav4 owner.
Mahalo

83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle

And now we know...the rest of the story.

TEST COMPLETE.

Thanks, Tony, for going the extra mile to wrap this up neatly.

 

[spooka]

We drove a brand new LEAF (RedXXX) on Nov 4, 2012 in Phoenix with only 138 miles on the odometer (and a recent production date) to run the exact course and parameters in similar weather as the Sept 15, 2012 LEAF range autonomy demonstration. This was a shortcoming of our original test, with no "control" car that was capable of producing 84 miles of range autonomy at 4 miles/kWh.

83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle

Looks like This is Much Ado About SOMETHING
Time to rewrite this page (to at least adjust the graphs to 84 miles per the service bulletin)
http://www.casteyanqui.com/ev/capacity_kerfuffle/index.html


Hopefully data will triumph over opinion.

 

[Yanquetino]

Looks like This is Much Ado About SOMETHING
Time to rewrite this page (to at least adjust the graphs to 84 miles per the service bulletin)
http://www.casteyanqui.com/ev/capacity_kerfuffle/index.html

Hopefully data will triumph over opinion.

The achieved range is admirably toward the high end, but 83.2 miles is still within Nissan's 76-to-84 range at 4 miles-per-kWh in a new Leaf. Those parameters are not "opinion": they are Nissan's own projected estimates, and likely what the automaker would cite to defend their claim that "The cars and the battery packs are behaving as we expected."

 

[Stoaty]

83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle

And now we know...the rest of the story.

TEST COMPLETE.

Thanks, Tony, for going the extra mile to wrap this up neatly.

I will add this test result to the Wiki this weekend; it's the final nail in the "Nothing wrong here" coffin... unless of course you bet on the wrong horse (see post above).

 

[surfingslovak]

Those parameters are not "opinion": they are Nissan's own projected estimates, and likely what the automaker would cite to defend their claim that "The cars and the battery packs are behaving as we expected."

And the endless argument continues :roll:

There is an easy solution to all this, Nissan could publish the AESC cycle life spec sheets recorded at different ambient temperatures. They have undoubtedly tested this extensively over the years. But they won't say anything. The same thing applies to usable battery pack capacity, its price, and other fundamental metrics.

When owners complain about declining range, it's either their own fault, a mysterious software bug or a gauge error. Or a combination of all three.

Meanwhile, the manufacturer drags its feet the best it can, and our official liaison to Nissan does not see a problem. He even goes as far as recommending wet nurses to the most vocal and impatient of those that are dissatisfied. Given this situation, what will a longtime evangelist and EV activist of your stature do? Write an article attacking owners, and defending the manufacturer hoping that "it might help repair at least modicum of the damage".

We don't have to agree, obviously, and everyone is entitled to their opinion. It would be good to realize though, that situations like this one should be avoided in the future. That might take a change of attitude by everyone involved.

Start battery stored energy: 265 Gids

83.2 miles driven (with 21 Gids / 7.47% remaining)
88.7 miles calculated range to turtle
GOM: "---" (normal for "Very Low Battery")
Gids: 11 / 3.9% remaining
Economy: 4.3 miles/kWh

Many thanks for doing this due diligence, especially to the driver in Phoenix. While this test is fresh in your memory, please recall that Jay did a similar run in April, two months after taking possession of his Leaf. He drove on level freeway in 80 F weather with no wind, and 65 mph indicated on the dash. His results are a mirror image of what was measured last weekend in Phoenix.

Here are Jay's key metrics:

GIDs on Jun 03 came out to 262

LBW at 76 miles with 8 miles left in the GoM
I would have seen VLB at 83 miles
Projected total range 89.9 miles (76 + 3.3 x 4.2)
4.2 miles/KWH

Click to open

 

[JeremyW]

Meanwhile, the manufacturer drags its feet the best it can, and our official liaison to Nissan does not see a problem. He even goes as far as recommending wet nurses to the most vocal and impatient of those that are dissatisfied. Given this situation, what will a long-time evangelist and EV activist of your stature do? Write an article attacking owners, and defending the manufacturer in the hopes that "it might help repair at least of modicum of the damage".

My way to "fight back" is to run my own experiment. I believe one can extend the life of these batteries significantly by adding more buffer to the top and bottom of the pack. Obviously this limits the range of the vehicle, making the 100 mile claim a more painful joke. Lucky, for around town driving I need a range of about 20 miles. This means I can actively keep the average state of charge low. I rarely charge to 100% or even 80%. I normally stay between 7 and 3 bars.

As far as dealing with high ambient heat, I vented my garage and routed fresh air around the underside of the car during the summer. On especially hot summer days I tried to stay as low as 3 or 4 charge bars. I also have shaded parking at work most of the time. There's not much else one can do too much for high ambient temps. Well, maybe invest in a portable air conditioner. :lol:

I hope that my efforts will raise awareness to what the one can do to extend the life of a lithium traction pack in an EV. Nissan is seriously lacking in this area. :|

 

[DaveinOlyWA]

Meanwhile, the manufacturer drags its feet the best it can, and our official liaison to Nissan does not see a problem. He even goes as far as recommending wet nurses to the most vocal and impatient of those that are dissatisfied. Given this situation, what will a long-time evangelist and EV activist of your stature do? Write an article attacking owners, and defending the manufacturer in the hopes that "it might help repair at least of modicum of the damage".

My way to "fight back" is to run my own experiment. I believe one can extend the life of these batteries significantly by adding more buffer to the top and bottom of the pack. Obviously this limits the range of the vehicle, making the 100 mile claim a more painful joke. Lucky, for around town driving I need a range of about 20 miles. This means I can actively keep the average state of charge low. I rarely charge to 100% or even 80%. I normally stay between 7 and 3 bars.

As far as dealing with high ambient heat, I vented my garage and routed fresh air around the underside of the car during the summer. On especially hot summer days I tried to stay as low as 3 or 4 charge bars. I also have shaded parking at work most of the time. There's not much else one can do too much for high ambient temps. Well, maybe invest in a portable air conditioner. :lol:

I hope that my efforts will raise awareness to what the one can do to extend the life of a lithium traction pack in an EV. Nissan is seriously lacking in this area. :|

Jeremy; sounds like you have a valid experiment going. how many miles have you traveled? have you posted your stats on the plug in america survey?

http://www.pluginamerica.org/surveys/batteries/leaf/" onclick="window.open(this.href);return false;

 

[JeremyW]

Jeremy; sounds like you have a valid experiment going. how many miles have you traveled? have you posted your stats on the plug in america survey?

Yes I have. I've got ~5,000 miles on it. Surfingslovac estimated 6-8% loss. I got the car with 271 gids. Hope to get a better gid count tomorrow night! I think that loss will drop a bit with the cooler weather we have now. :?

 

[DaveinOlyWA]

Jeremy; sounds like you have a valid experiment going. how many miles have you traveled? have you posted your stats on the plug in america survey?

Yes I have. I've got ~5,000 miles on it. Surfingslovac estimated 6-8% loss. I got the car with 271 gids. Hope to get a better gid count tomorrow night! I think that loss will drop a bit with the cooler weather we have now. :?

you have 5,000 miles and already have 6-8%?? oh man! sorry to hear that. you will gain something back on cooler weather but no where near that much. maybe half?

i gained back about 2-3 % but didnt matter because i have to deal with the 10-20% loss from cold weather. all i can say is if you get cold weather, get a heated jacket. used it last night and it works pretty good!

 

[turbo2ltr]

I understand why it is financially not practical to test so many Leafs on a Dyno running the EPA's 5 cycle drive map test.... but THAT is the most reliable method. There are unavoidable errors when conducting road tests.

Actually a constant rate discharge test directly on the battery using a load bank, eliminating the drivetrain variables completely would be the most reliable...

 

[surfingslovak]

Actually a constant rate discharge test directly on the battery using a load bank, eliminating the drivetrain variables completely would be the most reliable...

Yes, and I believe that's exactly what Nissan did with the seven Leafs we have heard of in Casa Grande last July. The load was presumably steady 5 kW, which would be likely representative of urban driving without freeway use. Some of these results have reportedly been shared with individual owners and they found their way onto this board.

The results from the road test organized by Tony match very closely with the known capacity for two cars tested at Casa Grande we have the numbers for: Red500 (Azdre/opossum) tested at 85% by Nissan, and was at 82.5% capacity during the range test. White 626 (Ticktock) tested at 87% by Nissan, and 87.5% during the range test.

All of this was discussed before, repeatedly, and it's on the Wiki as well, thanks to Stoaty's tireless efforts.

 

[JeremyW]

you have 5,000 miles and already have 6-8%?? oh man! sorry to hear that. you will gain something back on cooler weather but no where near that much. maybe half?

Yeah. I know. I think I got the car with ~3% loss. Luckly I don't need much. If we get a few quick chargers around here I won't care at all! I think I lossed about 3% over summer with a good amount of ~100 degree days. I'll do a turtle to 100% charge soon and see where I stand.

 

[spooka]

The achieved range is admirably toward the high end, but 83.2 miles is still within Nissan's 76-to-84 range at 4 miles-per-kWh in a new Leaf. Those parameters are not "opinion": they are Nissan's own projected estimates."

Yanquentino, this is not an attack on you and I hope you don't perceive it in that manner. I believe that you love the Leaf and want to see EVs thrive as much as I do. We both approach the battery issue from different ends of the spectrum and the following is my argument for my concerns about the data you present on your site.

Correct, the service bulletin states a range of 76 to 84 miles. The "opinion" enters the arena of discussion when a number is arbitrarily chosen from this range and as a result, incorrect conclusions are deduced from it. The brand new Leaf that just ran the Tempe Twelve course at the speed Nissan states is needed to obtain 4Miles/KW made it 83.2 miles with 21 GIDs remaining in the pack. It is reasonable to extrapolate that it would have made 88.7 miles to turtle, based on the average of 2.9 GIDs/mile obtained at the tests end point of 83.2 miles. This is observed data that demonstrates Nissan's 84 mile upper limit in the service bulletin was conservative and that all Leafs in the test can now reasonably be compared to the upper end of the range using 84 miles to turtle vs the 80 miles you arbitrarily chose. I hope your web article will reflect this change as others have used your presented data as factual in media stories which only perpetuates the distorted data in your article. I am publicly asking you to address this and base the data on your site on 84 miles. The data obtained from this Leaf proves that Nissan thankfully does not sell new battery packs with useable capacities below 21KW (or 24KW total). When you pick a happy medium of 20KW (80 miles) you are implying that Nissan is selling new Leafs to the public in an already damaged or defective state. The issue is what happens to the pack after use, not before delivery. Your whole assumption in your article is based on the "before" aspect and leaving your article as would (IMO) show a bias on your part now that there is hard data to the contrary.

Thanks for considering this.

 

[Weatherman]

The data obtained from this Leaf proves that Nissan thankfully does not sell new battery packs with useable capacities below 21KW (or 24KW total).

I wouldn't quite go this far. There's, at least, anecdotal evidence to suggest a fair amount of variability in the usable capacity of new battery packs. True, a vast majority should be around the 21kWh range, but my first LEAF had perceptibly less than that to start. That's why I sold it.