Phoenix Range Test Results, September 15, 2012

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Fabulous data, great work by Tony and a bunch of other testers. I did my own range test (not standardized), but I do have comparison data from 7 months earlier.

Start (Full Charge) - 255 Gids
Work (20.2 miles) - 195 Gids
Home (40.0 miles) - 152 Gids
Work (60.2 miles) - 81 Gids
Lunch trip (69.6 miles) - 69 Gids (increased to 75 Gids when car turned on 6 hours later)
Part way home (73.0 miles) - 64 Gids (22.7% - got 82 miles previously from full charge to 22.7% Gids)
Part way home (82.0 miles) - 49 Gids (LBW, 17.4% Gids)
Home ( 89.3 miles) - 44 Gids (15.6% Gids; long downhill descent to get home uses only 5 Gids for 7 miles)

Temperatures were a bit warmer for this test (70 degrees and 5 Temp bars to start, high at work around 90 degrees F.), driving conditions were quite similar, perhaps a bit more stop and go on the freeway than usual which may have hurt my range a bit.

Range to 64 Gids - 73 miles
Range to 64 Gids 7 months ago - 82 miles

Calculated range loss to 64 Gids compared to 7 months ago - 11%

Difference in percent Gids remaining at 82 miles: 22.7% - 17.4% = 5%

Predicted range from this data (if gids are linear and car stops at 4 gids): (90.7% - 4/281%) * 89.2 miles/(90.7%-15.6%) =

(90.7-1.4) * 89.2/75 = 89.3 percent * 1.189 miles/percent = 106.2 miles (granted, not bad since there is a 1450 foot elevation gain and loss during the round trip)

Before anyone jumps up and down about different temperatures, etc., I will be repeating the test in a few months when it is cooler to see if there is any difference. While the range loss to 64 Gids appears significant, the predicted range to turtle is almost exactly the same as that done 7 months ago.

Conclusion:

1) Without driving all the way to turtle it isn't possible to estimate the actual range, my extrapolation is only a guess.
2) It appears that Gids are highly unreliable as a predictor of how much range is left. The range doesn't appear to have changed much, but the Gids certainly have!
 
Interesting results. Better range than I would have thought for the degraded cars. It definitely sounds like there are in fact some instrumentation issues without the Leaf in addition to the capacity loss. The car that went extra far could have picked up a tailwind... That would be the easy explanation for the variance.

One minor suggestion on displaying your results table: I would encourage you to paste the results data into a google spreadsheet, freeze the header row and sort sheet by distance traveled or sort it by capacity bars. The random non linear order that the data is displayed in seems odd and makes it hard to process. Sorting by something would really help and would take 30 seconds to a minute to do and a few extra seconds to take a screen shot (Easy on a Mac of course, on a pc... :( sad face).
Seems like you are sorting by range now so that should be at the left and the GID mess should be to the right as it makes no sense.

Thanks for organizing the collection of all of this data for us Tony. Very interesting. Big thanks as well to everyone that participated in this real world leaf range study. Sad to hear about the scraped cars.
 
Given all the effort put into this, I am sorry to say that without range data from a new car, tested on the same route under similar conditions, the current results are not as meaningful as we would want them to be.

Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.

I am also not quite convinced that the gid count is meaningful. Certainly at the low and high end the gids loosely correlate to range, but e.g. in the middle (i.e. gids ranging from 70-75%, which covers 6 out of 12 cars), the correlation between gids and actual range drops to .58. If we include cars #2, #5, #6, #7, #8, #9 and #12 (covering gid percentage from 70-76%), the correlation drops even further to .36 at which point I am not sure that there is a meaningful correlation between gids and range.

Using them as a reference is probably not a good idea. As it is right now, we have nothing to compare these ranges to and we still don't know how much capacity loss the tested leafs have actually experienced.
 
There's an interesting thing to note here as well. Per the chart in the OP, for a new car there are an expected 9-24 miles (depending on efficiency) between LBW and turtle. The owners manual instructs you to find a charging station as soon as possible after LBW, preferable before VLBW. To me, this means that 10-15 of the miles you see on the post should not be used on a regular basis.

Thanks again, Tony!
 
klapauzius said:
Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.


you don't find it meaningful that one car went 60 miles and another 80 miles on the same route in the same conditions? Seems to me the car with 60 miles has a problem. About a 25% problem versus a car that is not even whole itself.
 
palmermd said:
klapauzius said:
Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.


you don't find it meaningful that one car went 60 miles and another 80 miles on the same route in the same conditions? Seems to me the car with 60 miles has a problem. About a 25% problem versus a car that is not even whole itself.

And that car has only 70% of the range autonomy of an 85 mile new LEAF (range of my new LEAF in April 2011 at 4.0 mi/kWh indicated). I recall having read that at 70% capacity, an EV pack is considered in the industry to be toast.
 
klapauzius said:
Given all the effort put into this, I am sorry to say that without range data from a new car, tested on the same route under similar conditions, the current results are not as meaningful as we would want them to be.

Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.
Agreed, but based on the data it looks like a new Leaf would have gone at least 84 miles in this test. If anything, it might have gone further, making the losses worse than shown in Tony's chart.
 
palmermd said:
klapauzius said:
Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.


you don't find it meaningful that one car went 60 miles and another 80 miles on the same route in the same conditions? Seems to me the car with 60 miles has a problem. About a 25% problem versus a car that is not even whole itself.
Yes, but that is only one out of 12. I do not doubt that probably all cars tested have reduced range, but the exact amount of reduction is still unknown. And it seems that neither the gidmeter nor Nissans capacity meter are reliable indicators.
 
Stoaty said:
klapauzius said:
Given all the effort put into this, I am sorry to say that without range data from a new car, tested on the same route under similar conditions, the current results are not as meaningful as we would want them to be.

Yes, Nissan wrote somewhere that a new car can go 84 miles (or 70 or whatever number), but unless you know how far it can go on the route you tested, we are still missing a reference.
Agreed, but based on the data it looks like a new Leaf would have gone at least 84 miles in this test. If anything, it might have gone further, making the losses worse than shown in Tony's chart.
Yes, it might even go 100 miles on that test-route, but without actually testing this under the same conditions, we dont know.
84 at 65 F with constant speed at 55?
Or
84 going downhill with rearwind at 100F?
 
So now what's next?

The user test in Phoenix is done, the conclusion is that there is clearly a battery problem, and also instrumentation problem as well.

It's been a couple of months since Nissan returned the Casa Grande test cars and they're still not sharing data with any of those owners and giving any conclusion of their own. You can only deduce that the data is not good for them, and that's why they don't want to share.

What should be done next?

More of the same test in Texas?

Release results to the media?

Organize class action lawsuit?

Surely, after Nissan's deafening silence for the last several months, we shouldn't expect that they will bother responding to the AZ Tempe test results at all, do we?
 
Volusiano said:
What should be done next?

More of the same test in Texas?

Release results to the media?

Organize class action lawsuit?

Surely, after Nissan's deafening silence for the last several months, we shouldn't expect that they will bother responding to the AZ Tempe test results at all, do we?
It would really nice to have a relatively new control car or two to run the same course. It sounds like that might happen in the near future. I don't think a repeat of the test in TX is needed. I'd imagine that some of the EV/alt energy media has already picked up on this beyond EVTV.

Some law firm already seemed to already be in the process of starting a class action lawsuit on the "battery cooling system". :roll:
 
Yeah, a control car is really important. We'd need to find both a car and a dealer that just fell off the turnip truck. The car would have to be fresh off of the dock, and the dealer would have to let us take it overnight and put 100 miles on it.

OR

Nissan can admit there's an issue and offer lease conversions, buybacks, or a capacity warranty.
 
azdre said:
Per the chart in the OP, for a new car there are an expected 9-24 miles (depending on efficiency) between LBW and turtle. The owners manual instructs you to find a charging station as soon as possible after LBW, preferable before VLBW. To me, this means that 10-15 of the miles you see on the post should not be used on a regular basis.
Bingo. My first thought as I was towing cars back to the chargers and hearing range numbers in the 70-mile neighborhood was, "Fu**! The conditions for this test are not revealing the severity of our issue!" But once I got some sleep and thought/talked it over, two things came to mind. 1, as my lovely wife stated, we are not 'supposed' to be driving our cars down to VLBW or turtle regularly, so those low end miles are essentially useless to us if we don't want to be blamed for any of the battery degradation. 2, this test did not use *any* climate control. You can still hit up the Nissan Leaf website and it will tell you that a cross-town Phoenix freeway commute of 70-miles is fine for the Nissan Leaf. And yet, when our car hit 15 months of ownership, there's no way in hell our car would do a 70-mile freeway trip with some light air conditioner use and not hit turtle or run completely dead. So is Nissan's website only telling people how suitable the car is for 3 months of ownership, perhaps? :roll:

Volusiano said:
What should be done next?

More of the same test in Texas?

Release results to the media?

Organize class action lawsuit?
Don't worry. We have two different news stations doing interviews/stories Wednesday, and I'm aware of a handful of folks working with lawyers right now.

Nissan really should start calling me back with a little more urgency. They multi-day delays in calling or emailing me just leave me more spare time to talk to news outlets and other people and organizations! :twisted:
 
A couple of thoughts come to mind:

VLBW to Turtle may be erratic in well worn to spent packs.. the first module that reaches the low voltage threshold will trigger that mode and voltage is dropping precipitously at that point.

Also, in a worn out pack internal resistance and thus voltages under load will also start to increase, making VLBW to turtle even shorter. I doubt the GID calculation accounts for that.

I dont have any suggestions short of testing the capacity of an actual cell by itself.
 
TonyWilliams said:
Even if none of the recharges were monitored, the m/kWh reports from these LEAFs should at least show whether they share a common error with the gid count reports.
Those numbers were ALL OVER THE PLACE. Not only is that data not needed, you can't prove much with bad data.
True - but at the very least - isn't sharing all the collected data potentially informative? - perhaps not from the perspective of showing that batteries have degraded and a significant amount of range has been lost it certainly has - but from the perspective of trying to identify other factors which may determine range and efficiency as well as provide data that may provide insight on how other cars may behave. Otherwise we might as well throw out your range chart which you built and others relied upon.

For example - I know that you make sure that the same OEM EP422 tires were on all cars - but amount of tread remaining can have an effect on efficiency - if you swapped the <10k mi tires on the 12-bar LEAFs with the 20k+ mi tires on the 9-bar LEAFs - I'm sure that mi/kWh numbers would get bounced around a bit. Another example would be alignment - slight differences in alignment can also affect efficiency and thus range. This is impossible to account for in this type of test unless you do some sort of coast-down testing.

Again - I'm not trying to "armchair quarterback" as some others here may be doing. I think the results and data you've provided speak for themselves. I'm just looking for a release of the other bits of data that you've recorded during the test which should be very informative for other purposes.

BTW - I touched on this earlier - what if the capacity meter only uses LBW-100% for it's percentages and not dead/turtle-100%? This might explain the higher than expected range.

Assuming 100% = 281 GID = 22.5 kWh
11 bar = 19.2 kWh or 85%
10 bar = 17.8 kWh or 80%
9 bar = 16.8 kWh or 75%
8 bar = 15.7 kWh or 70%

That helps somewhat. But still doesn't explain the wonder-car Blue534. Really want to know what makes that car tick. :)
 
drees said:
But still doesn't explain the wonder-car Blue534. Really want to know what makes that car tick. :)
Me, too! Unfortunately that VIN does not appear to be in the Wiki, so the identity of the owner is a true mystery. I don't know if LEAFfan (our range leader) participated in the test or not, but I've been wondering if he was the driver of Blue534!

Without outing the owner, perhaps Tony or someone else who knows the owner can collect information about how the car is used, garaged, charged, etc., and share it with us so that we can try to better understand best practices for preserving the batteries in our LEAFs!

Regarding the relatively poor showing of Tony's car, I believe that *might* be mostly due to poor cell balancing, as was previously suggested. It seems quite possible, perhaps even likely, that charging to 80% will lead to a very out-of-balance condition, particularly if the charging is done with the pack hot (because it is not isothermal) as Tony seems to have been doing. My experience with our LEAF is that we dont't get great good range if we charge to 80% many times and then charge to 100% once for a trip or a test. This had led me to think that our battery was fairly degraded. However, subsequent charges to 100% seemed to hold more charge.

Anyway, it would be interesting to someday test the impact of a couple of months of 80% charges and shallow discharges on a series of daily 100% range tests to determine if the range grows for a few days as the pack achieves better balance.
 
Many of the posts above brought up a lot of good points.

I understand why we can't trust the LEAF's instrumentation for the purpose of measuring true range from 100% to turtle. However, despite this, it would have been very easy to record the various LEAF instrumentation readings anyway while we were at it already, so that the test results can be looked at from many other different angles if so desired. I think it was suggested before the test took place in the other thread that this should be done because it would have cost nothing more to do it. So I don't know if that was done but those data not released. Or whether it was not done at all. If it were done as suggested but data not released, I hope the data can be released and shared.

For example, the few posts above an excellent point about how not too many people would dare to drive to turtle. So while LBW may not be accurate and may be all over the place, it's still an important indicator because people still have to rely on it (regardless of how inaccurate it is) to decide to stop driving and find a nearby charging station to fill up. So range to LBW and range to VLBW would have been just as useful to know in addition to range to turtle. If anything, at least to see how consistent (or erratic) it is between the test cars, and what to expect, or not to expect.

The energy efficiency reading (miles/kwh) may also be all over the place and unreliable and not good for range calculation. But at least if it had been recorded (and why not while we were at it already), let's share it, if anything, just to confirm that it's really all over the place and not reliable as suspected.
 
drees said:
But still doesn't explain the wonder-car Blue534. Really want to know what makes that car tick. :)
This is likely the instrumentation error Palmer spoke of. Keep in mind my car (white626) would have been 10 bars had it not been recalibrated/reset in Casa Grande and I would have been an out-lyer, too. Who knows, maybe I would have been down to 9 by now.
 
="TonyWilliams"

Those (m/kWh) numbers were ALL OVER THE PLACE. Not only is that data not needed, you can't prove much with bad data.

It is irresponsible for you to refuse to disclose this "bad data".

Withholding this information only puts your own credibility, and that of all the other test data, into question.

You seem to be saying that you only source for the 84 mile “new car demonstrated capacity” is the Nissan service bulletin, you posted on page one of this thread. As I read it, that states that the "estimated" LEAF range is 76 to 84 miles, at 4m/kWh.

So it looks to me like you may have misrepresented this as a source for an 84 mile range claim. I think any reasonable reading of this service bulletin indicates an estimated range of 80 miles, with a very large variation, of plus or minus 4 miles, or 5%. Neither this this source, nor your test, (or any other LEAF range test, AFAIK) appears to answer the important question of whether the individual range of new LEAFs, and/or their underlying battery capacities, actually varies by close to this much.

And so, it appears you may have misrepresented the range loss of all the LEAFs tested, by exaggerating the percentages of range lose by about that same 5%. And since you apparently did not collect any data that would allow us to try to estimate actual battery capacity of any of the test cars:

"TonyWilliams"

...This was a range test, not a battery capacity test...

You now seem to have no way to verify the test cars even matched the 4.0 m/kWh which Nissan specified as the rate required to achieve the indicated ~80 mile range.

So, the only thing I can conclude from the methodology and results of of this range test, its that eleven of the twelve LEAFs (~ten of which were selected for testing because the owners reported unusually high loss of range) have about 84% to 100% (with a large and unknown level of uncertainty) of what Nissan states (in your single source cited) is "estimated" new LEAF range.

Whether I personally considered the level of range reduction of any of the LEAFs greater than "promised' by Nissan, and unacceptable, would depend largely on the conditions of use these LEAFs experienced, which you have made no attempt to report.

Unfortunately, in the ~six months since the rate of battery loss has become such an obsession on MNL, very little has been accomplished in the way of collecting useful information as to what use factors lead to actual battery capacity loss, as opposed to the of capacity bar losses, or reduced gid counts, which the test seems to have confirmed to be quite inaccurate.

IMO, the most significant new information from this test seems to be that the hot Arizona climate has far less to do with actual capacity loss, than the capacity bar losses and gid count losses, once seemed to indicate.


TonyWilliams said:
edatoakrun said:
You are planning to monitor your recharge capacity after the range tests, I hope?

No, that's not range. This was a range test, not a battery capacity test. I'm not dumb enough to get into a spitting contest where they hold all the cards. Too many variables dilute the message, and I fear just introducing the Gidmeter is a stretch.

Even if none of the recharges were monitored, the m/kWh reports from these LEAFs should at least show whether they share a common error with the gid count reports.


Those numbers were ALL OVER THE PLACE. Not only is that data not needed, you can't prove much with bad data.


Does anyone still believe gids reflect an accurate and constant Wh value?



I thought it was a heck of a lot better than it clearly is.


What is the source for the 84 mile “new car demonstrated capacity”, as reported?


Nissan service bulletin, quoted in the piece.


Were multiple new LEAFs, or even a single one, ever actually tested under the same conditions, resulting in 84 miles of range?


As indicated, no car went 84 miles. My car (Black782) would certainly meet the 84 mile threshold a month or three ago, but as stated elsewhere, I only charge to 80% (except for the BC2BC trip) and really had no idea that this car could already be tanking. My previous LEAF went about a year before things starting falling apart.

So, mea culpa on me. I should have checked before bringing the car. I can check another car, if I find one, and if Nissan wants to keep this in the news longer, we can go back and forth. Absolutely fine with me. The fact remains, one car went X distance, and the others went X minus some significant number. It would be stupid of them to brag about 76-79 miles to dead, when they're advertising 100. I can make a new chart with the range compared to 100 miles if they want. Or we can get 5 or 10 cars new LEAFs and test only them. It just wouldn't end well for their argument, IMHO.

I have not played my whole hand on this, and I believe I'm ready for the hate fest (or they could pull their heads out of their 6 anytime and do right in Phoenix).


I don't think Nissan ever gave range estimates near this value. So, without documentation of this as the actual capacity and range common to new LEAFs, the calculated percentages of loss could appear to be exaggerated.


Again, Nissan's own data. But, sure, if they want to say 84 miles is an exaggerated range at 62mph, we can do that.


On the PR front, I suspect Nissan may actually be fairly pleased that this range test, of a group of LEAFs, so heavily weighted toward those perceived to have the very worst capacity loss, would seem to show that that all but one car (with 29 k miles) have come fairly close to their own highway range estimate:


Yes, it is better than initially perceived, but I'm not sure it's wise to start bragging publically about the "good" losses in 12-18 months... or far less miles. Nobody considers 29,000 miles "high mileage", and it is still under warranty.


There are an infinite numbers of variables which will effect your actual range, but these numbers give you a good idea of what to expect based on your own personal driving habits...


I didn't quite follow the reason for that whole angle. We used one set of variables... That was kind of the point.
 
My LEAF was part of the test. RED429. Thanks Tony for helping PHX LEAF owners collect and use more data to make our decisions on owning a LEAF!! I was really impressed that a bunch of total strangers came together for a few hours and worked coorperatively and collegially to complete a job with a high degree of complexity! It really was amazing!!

This is my take on the numbers:
This test has been run on very few samples, not enough for "real" statistical analysis, but let's look at the Correlations anyway!! They do indicated areas of further study!!

We might want to know whether number of BARS are related to Miles to Turtle. The statistical technique for determining the degree to which two variables are related (i.e., the degree to which they co-vary) is, not surprisingly, called correlation.
This correlation ranges from –1.00 to +1.00. A correlation of 1.00, whether it’s positive or negative, is a perfect correlation. It means that as scores on one of the two variables increase or decrease, the scores on the other variable increase or decrease by the same magnitude.

Thus, a correlation of .8 or .9 is regarded as a high correlation, i.e., there is a very close relationship between the data on one of the variables with the data on the other. And correlations of .2 or .3 are regarded as low correlations, i.e., there is some relationship between the two variables, but it’s a weak one. Knowing the data on one variable wouldn’t allow you to predict their data on the other variable very well.

Correlation Bars to Turtle Miles .797
Correlation GID to Turtle Miles .787
Correlation Odometer Miles to Turtle Miles .789

With this LEAF test all Correlations are high.
The difference between the Correlations is very low, probably not statistically significant since the test has very few cars in the test.

The Correlations would read this way for each variable:

The more BARS you have the farther you go to Turtle.
The more GID you have the farther you to to Turtle.
The more miles are on your odometer, the fewer miles you go to Turtle.

One of the Correlations I would like data on is, Days from Manufacture date to Miles to Turtle!!
That would tell me if the age of the battery is related to Miles to Turtle!

I pondered BLUE534. The car has a high ODOMETER reading and was able to go alot of miles to Turtle.
The ODOMETER reading indicates the car being driven farther each day than I drive! My commute is 25 miles a day.
If BLUE534 drives farther each day, it is recharging from a more depleted battery each day.
It reminds me of when we bought our laptops or phones and we were reminded to "run the battery down" before we recharged so it didn't get a "memory". I know this battery chemistry is not supposed to have "memory", but that one car's data reminds me of that type of situation.

I really wish we had a couple of new LEAFs too!! That would determine the loss of range better that just saying "84".
 
Back
Top