Dumb**s alert: am I the first to drain it??

[LakeLeaf]

Do we know for sure if the car takes action based on the actual remaining state of charge, or on the projected range of the car?

You would hope that it is the former, but if so - you'd also think that at some point the "range to empty" algorithm would switch from projected miles to actual measured charge.

I'm hoping that Turbo's CAN work will give a definitive answer. From the posts here - it isn't really clear what exactly controls the last stages of life.

 

[mwalsh]

Other anecdotal evidence of cold weather doing a number on range, although he doesn't say under what conditions (speed, etc...) the errands were run:

I just picked up my leaf today. Reserved in April last year, ordered in August. It is so quiet and smooth.. We spent the afternoon running errands in 35 degree weather. Drove 30 miles had 45 miles left on the estimated range.

After I've driven 30 miles round these parts (so with warmer daytime temps), and at full highway speeds mind you, I've typically still got something like 60 miles apparent.

 

[evnow]

Today would be a good test in Seattle area - with temperatures around 15 now in my place (25 in Seattle).

 

[derkraut]

Today would be a good test in Seattle area - with temperatures around 15 now in my place (25 in Seattle).

15 degrees! Stay warm, bud.

 

[Ready2plugin]

Today would be a good test in Seattle area - with temperatures around 15 now in my place (25 in Seattle).

15 degrees! Stay warm, bud.

Yes, and it's moving straight down the coast, they are saying we may have measurable snow in San Francisco region tonight!

 

[Carcus]

One factor in accurate range estimation that I haven't seen mentioned yet .... wind.

If the first 1/2 of your drive is with a tailwind and then the drive home is with a headwind, this might cause your range estimation to be off.

While I don't have an electric car, I do obsess over mpg with a scangauge. I have seen mpg swings of 15% or more on a "going vs. coming" windy day commute.

If the two fellows who had the problem want to check their weather conditions (wind, temp, etc) at the time of draining, this can be done by checking the weather history at weather underground.

http://www.wunderground.com/

 

[garygid]

If one ignores the possibly-misleading Range estimates, and looks at just the SOC bars, there should have been a Low Battery warning, and Turtle Mode should have kicked in soon enough to give some safe distance, possibly much less if driving uphill.

I thought Turbo tested "inserting" SOC values and found that Low Battery and Turtle modes were triggered by SOC values, but maybe that is just another defect in my memory?

So, many LEAFs are reported to give good Battery Low warnings, and travel "miles" in turtle mode.

Two LEAFs have reported rather abrupt "endings".

Did the miles-remaining go to "---" and then the car give no other warnings?

Was there no Low Battery warning, visual or alarm?

Sounds like there is something strange (and NOT desirable) going on with these two cars, and Nissan should investigate.

Definitely let Nissan CS know about the car, and your "tow" experience.

 

[klapauzius]

There was a low battery warning at around 9 miles (~ 6 miles distance to my home). Turtle mode was just going a few 100 yards.
From what I read from the manual, it must have been the cold weather.

What is troubling is that apparently SOC is relative measure (i.e. 6 bars= 50% battery capacity) not an absolute one, e.g. 6 bars = 12 kWh. So this could mean
that 6 bars at the beginning of that 40 mile trip where not really 12 kWH (which should have been just about enough to make it), but maybe more like 8-10 kWh at 32 F. The range calculation does not takes this into account it seems, which is really really really bad.

 

[Jimmydreams]

What is troubling is that apparently SOC is relative measure (i.e. 6 bars= 50% battery capacity) not an absolute one, e.g. 6 bars = 12 kWh. So this could mean
that 6 bars at the beginning of that 40 mile trip where not really 12 kWH (which should have been just about enough to make it), but maybe more like 8-10 kWh at 32 F. The range calculation does not takes this into account it seems, which is really really really bad.

Therein lies the problem with using the energy bars instead of a true SOC indicator:

You have 6 bars showing...is that 6th bar 'fresh', i.e. did the 7th bar JUST go out and you have all of that 6th bar remaining, or is that 6th bar 'old' and about to go away after you drive only 100 yards? If each bar represents roughly 6 miles of driving, it makes a big difference what that 6th bar actually represents.

 

[smkettner]

Well the battery does have a temperature guage so I would try to assume Nissan would use that in predicting the remaining miles.
The system did seem to fail in these two circumstances.

Seems like if conditions are critical the driver should deduct at least 5 miles in poor conditions.

 

[Carcus]

Klapauzius,

A likely scenario could be that the Leaf's range calculations were originally based on lower speed suburban runaround at 48 deg. Then, the car was put into a higher speed (I5) commute at 32 deg. As the evening progressed -- high(er) speeds, thicker air and cold soaking took their toll.

In the most sophisticated of range forecasting, the Leaf would know (in no particular order):
1. Passenger load
2. Driver habits/ speed/ forecast speed
3. Traffic
4. Winds/ forecast winds
5. Temperature/ forecast temperature
6. Routing
7. Accessory load/ forecast accessory load
8. Elevation changes
9. SOC
10. Some other stuff I'm probably leaving out.

The nice thing is that all of this is entirely possible now, but it would require some additional input through the carwings system and from the driver.

The other type of less sophisticated "range forecasting" is that the driver will become more aware of all of these factors with experience and learn to provide their own mental adjustments to a more simplified dash display of the range.


/I'd imagine Nissan, Ford and the rest are working on this or have already got the programming done. They will probably want to sell you the upgrade(s). But hey, anyone with a computer knows how the "upgrade game" goes already.

 

[EVDRIVER]

Klapauzius,

The other type of less sophisticated "range forecasting" is that the driver will become more aware of all of these factors with experience and learn to provide their own mental adjustments to a more simplified dash display of the range.


/I'd imagine Nissan, Ford and the rest are working on this or have already got the programming done. They will probably want to sell you the upgrade(s). But hey, anyone with a computer knows how the "upgrade game" goes already.

It's called a proper SOC meter and experienced EV drivers have been using it to forecast range accurately for years, you just figured out what works and something Nissan was too stubborn to use because they thought they knew better. Believe me, they were told this long ago. All the other data points you mention plus many more need very good software, I know the pain in this because a close friend of mine is developing exactly this for a major auto maker for a well known vehicle that has electric drive, it makes the LEAF Carwings look like a primitive toy in range calculation. However, proper SOC goes a long way regardless

 

[turbo2ltr]

I haven't really been following this thread, but here's some input.

I have found the warnings and turtle mode are completely based on (what I am calling) SOC, which comes directly from the BMS. Indicated range and SOC bars have nothing to do with when the car shuts down, though the bars are directly related to SOC. I have "faked" SOC and put the car into turtle mode. The warnings and turtle always happen at the same SOC regardless of indicated range. Turtle mode is latching, so in order to get out of turtle mode, you have to shut the car down. The BMS does not concern itself with range. It doesn't generate it or even know it. It would have been really nice to see the actual SOC at the time of the shutdown.

Is SOC adjusted for temp? I don't know. I can't think of any way to test that without a really big refrigerator.

 

[Carcus]

Klapauzius,

The other type of less sophisticated "range forecasting" is that the driver will become more aware of all of these factors with experience and learn to provide their own mental adjustments to a more simplified dash display of the range.


/I'd imagine Nissan, Ford and the rest are working on this or have already got the programming done. They will probably want to sell you the upgrade(s). But hey, anyone with a computer knows how the "upgrade game" goes already.

It's called a proper SOC meter and experienced EV drivers have been using it to forecast range accurately for years, you just figured out what works and something Nissan was too stubborn to use because they thought they knew better. Believe me, they were told this long ago. All the other data points you mention plus many more need very good software, I know the pain in this because a close friend of mine is developing exactly this for a major auto maker for a well known vehicle that has electric drive, it makes the LEAF Carwings look like a primitive toy in range calculation. However, proper SOC goes a long way regardless

Well, perhaps the masked man can rig up a cable that'll take the CAN bus info to an iphone (maybe even a RAM mounted ipad), then ... your friend could write a program that displays a proper SOC on the apple device .... or better yet, he writes a program that does the whole shebang!

/but yeah, it'd be a lot better just to have it on the dash already. Pat Q. Public probably doesn't want the control panel of the Millenium Falcon in their car.

// BTW, my idea of a sophisticated "range planning tool":

Most of the time, the tool won't be required so it won't slow your startup sequence of the car down. But, on a day when you know you'll be driving a lot, all you have to do is speak the names of your destinations in order.... i.e. work, gym, grocery store, home, airport, mall, home. Then the planning program will run its forecast sequence and let you know how tight it thinks it's going to be, or if it thinks you'll need opportunity charging.

 

[evnow]

Is SOC adjusted for temp? I don't know. I can't think of any way to test that without a really big refrigerator.

Or ship your stuff to me to check it out

Just drove my Leaf home @ 25 deg.

 

[TomT]

It'll be cold enough in Southern California this weekend to check it out here!

Is SOC adjusted for temp? I don't know. I can't think of any way to test that without a really big refrigerator.

Or ship your stuff to me to check it out

Just drove my Leaf home @ 25 deg.

 

[SanDust]

Just drove my Leaf home @ 25 deg.

Congratulations!

 

[planet4ever]

There was a low battery warning at around 9 miles (~ 6 miles distance to my home). Turtle mode was just going a few 100 yards. From what I read from the manual, it must have been the cold weather.

Cold weather can, and does, happen almost anywhere. I consider a turtle mode that only takes you a few hundred yards to be not acceptable, no matter what the weather. Nissan needs to be told this, in the firmest terms. We can make guesses as to whether SOC is absolute or relative; whether battery power plummets at some temperature and SOC; whether the computer logic goes bonkers with certain inputs; we can make guesses about lots of things. But it's Nissan which must do the hard work here. We, and in particular, klapausius and Kelangst and anyone else who has a problem like this, need to keep pressing Nissan until they get to the bottom of the problem and fix it, whether it is hardware, software, or design.

We shouldn't, and can't, count on turtle mode to get us out of every jam. It may not get us home or to a charging station, but it's minimum role needs to be that it can get us to a safe parking space. I would put the minimum acceptable distance at 1 kilometer (5/8 mile or 1100 yards).

Ray

 

[klapauzius]

I did a few tests and I also told Nissan about this. They will come out to Seattle and check my car on Tuesday. But
here some information to consider:

Nissan told me that the 24 kWH battery actually uses just 20 kWH for driving. The rest is buffer to keep the Li-Ion healthy.
I tested this and found it to be accurate. Dividing actual miles driven by the average Miles/kWH one can get a reasonable good estimate of the actual energy used. I had the car full and was checking miles driven exactly when one of the SOC bar disappeared.
It turns out that the battery capacity is indeed pretty much 20 kWH.

This means using SOC and average miles/kWH can give you a pretty decent estimate of range.

Based on this I empirically found the following:

Heater on, freeway/city driving at 32 F gives 26.4 miles for 5 bars of SOC used up or ~ 3 miles/ kWH. Interestingly, this number, i.e. 3 miles/kWh shows also up on the dashboard. Based on this, I assume there would be 11.6 kWH left in the battery and using the current rate, i.e. 3 m/kWH, this would give me driving range of 35 miles for a total of ~60 miles.

The car however told me at 26.4 miles that I had 50 miles left. It is weird, since the car should have the same numbers available as me, i.e. the average miles/kwh and the current SOC.

35 miles vs. 50 is an unacceptable discrepancy, especially since the range estimator leads you to believe you can go much further than you actually can.

I would be very interested if others could repeat this calculation and see if this discrepancy is systematic?

 

[JRP3]

Nissan told me that the 24 kWH battery actually uses just 20 kWH for driving. The rest is buffer to keep the Li-Ion healthy.

Doesn't that contradict what others have found, and what Nissan has claimed? I thought 27kwh actual had been confirmed.
Regarding range information and turtle mode, I think a SOC numerical display should be mandatory, and I think turtle mode should be progressive, and start sooner than it seems to. My EV, (a home conversion), starts cutting back power gradually and progressively as you near empty, so you get early warning. I had one low SOC event and was able to drive about 3 miles as the performance gradually dropped till I was doing about 20 mph. Could have gone a little further but that got me home.