Out of juice - battery issue, UI issue, or user mistake?

[wk057]

Respectfully, I'm not misinformed about sub-zero reserve. I can personally attest that there IS a 10-20 reserve since we had that happen before in FL in November. Here's a pic where we, too have driven at highway speeds for miles at sub-zero range on the gauge. You can see the SC stop on the display.

This trip from Charlotte was entirely an elevation DECREASE.
It was set to rated not ideal miles. "Ideal range" is just where unicorns live.
Temperature was 85. Sunroof closed. No leadfooting. The graph showed plenty of miles at arrival, the whole way... except for the last 50 miles when it was like someone poked a hole in the gas tank.
Highway traffic cruise set at 70, which we've done for thousands of miles. Uses more power than 60 but not THAT much. 60 with glass closed yields best range on flat roads.
40MPH speed taper came in the last 30-40 miles, which contains the slow driving part of the trip and where the yellow went to red far faster than the odometer.
Pack in September 2014 charged to 266 miles. Once. Now only charges to 253-259.
Algorithms Alchemy and other magic aside, seems to me that the pack is failing but they saw no errors and say the pack is fine, even though it went to zero like there was a hole in the bottom of the car and we were towing a truck and it shut down hard at 4 miles. No reserve capacity.

So did you attempt to utilize such a "reserve" this time or not? Again, if the car cut off at or below 0 miles, then everything is fine.

At 40 MPH that's 1.5 minutes per mile. If the A/C is on that's using about 2-4kW, so could be using a full 100 Wh/mile at that speed. At 60 MPH the HVAC would account for up to ~67 Wh/mi since it's usage per mile is only spread over 1 minute. So while running the HVAC you would be seeing rated miles drop *faster* at 40 MPH than you would have at a higher speed. If the battery cooling kicked in (very possible) then this would happen even with the cabin A/C off. Driving 40 MPH is probably a big part of the problem.

There is no reserve. The "reserve" people have experienced is simply a miscalibration that underestimated the initial actual pack capacity vs real world measurements and allowed usage down to the true bottom of the SoC. Once that calibration is accurate (after a few brushes into the < 10% SoC range) then that "reserve" disappears and is just essentially shifted back to where it's supposed to be as part of the rated miles.

Also, by draining the pack down to cut off is horrible for degredation, especially if the pack sits for hours in the heat or cold at this level. The photo above shows the limiter at something like 70kW... that's ridiculous. That's an amperage limit of about 80%C, which by my estimates put's the pack voltage down in the 2.xV per cell area resting. That's super low for the Model S pack, which I've never seen below about 3.1V per cell resting on my own vehicles at near 0% on the meter. So, any range loss you experience from running the battery to the ground... well, that's to be expected and isn't anyone else's fault.

 

[deonb]

Where?

"Flat Rural parts" of Florida:
http://www.teslamotors.com/customer-stories/world-record-father-son-drive

Respectfully, I'm not misinformed about sub-zero reserve. I can personally attest that there IS a 10-20 reserve since we had that happen before in FL in November.

That's not a reserve, it's a lack of being able to do a reliable measurement of the SOC.

Tesla try not to "on average" have people run out of charge at 0mph, since that will mean half the people will run out of charge at > 0m. So rather, have people on average run out at about -10m, since then the far majority of people will be able to drive to 0m. You might have further capacity available, you might not.

 

[MorrisonHiker]

She had only 2 miles to the Supercharger

...

And for what it's worth, the Ranger service is no longer a flat $100. It's $3/mile each way. $1,000.

Too bad you couldn't just flatbed it the 2 miles to the SpC before it shut down. That would've been a lot cheaper.

 

[jaguar36]

It could have been weather related. A good headwind can put a real hurt on your range. That would also be consistent with the 40mph limit, if you're running into a 20mph headwind, I could see that being the max speed you could achieve running near the bottom.

How long after she pulled over did she try to start going again? Why did she pull over if she was going to keep trying? It could be that the pack was at the ideal temp and that gave it a bit more power, but once she stopped it fell out of the ideal range. Could also be that the A/C kept running while she was stopped and ate the rest of the juice.

 

[qwk]

There were a few versions of firmware that had a reserve, but that hasn't been the case for over a year. This story is akin to running out of gas. Gasoline gauges aren't accurate enough to predict down to the last mile, and neither are electric cars.

 

[wk057]

There were a few versions of firmware that had a reserve, but that hasn't been the case for over a year. This story is akin to running out of gas. Gasoline gauges aren't accurate enough to predict down to the last mile, and neither are electric cars.

There was an official reserve? As far I know this "reserve" nonsense has always been a calibration artifact from day 1.

 

[stevepeck1]

Appreciate all the comments and questions. We didn't bank on having miles when the meter showed 0. We simply watch the energy graph and trip computer which started with 70 miles more than we needed, leaving Charlotte. We used "range mode" and warm-side-of-comfortable AC temp. 40 MPH was all the car would do for the last sub-10% part of the trip. The car quit hard with 4 miles on the gauge and went to 0 in the couple hundred feet it took to find a place to pull over.

Perhaps worth mentioning is this, too: When the tow truck dropped us at the supercharger it was putting in a decent charge but not more than 315 mph. Probably not a big deal but thought I would mention that when alone at a supercharger, we normally get 350. There was no one else there.

Upon starting driving, the car would not accelerate any better than a Prius and took about a minute to get to 60 miles an hour. No errors on the screen during this slow thing. No valet mode. Notably, navigation was also off-line.

Concerned that it would shut down again if she stopped, she turned around and went back to the supercharger, rebooted the main screen, got out. locked. waited a few... and upon re-trying things were normal.

While I LOVE this car and have enough technical experience to accept the "guesstimate" nature of these batteries, the range on the gauge is entirely unreliable and considerably less than the 265 "advertised." Despite all the hype about ending range anxiety and all the modifications to the navigation system, (and please re-read my mad love for the car) we still have to put in about 40% more miles than we need for the trip to be assured that we will make it. The most critical consideration, the final 30 miles of stated range wasn't good for nearly that before it quit.

 

[yobigd20]

There was an official reserve?

nope, never.

As far I know this "reserve" nonsense has always been a calibration artifact from day 1.

correct.

 

[daxz]

evtripplanner (P85,21", 200#, Speed=1.0x) is showing 141 Rated Miles from Richburg, SC to Santee (~118 real miles from Santee) and the elevation changes 4418 ft up and 4874 ft down so a lot of hills. At speed=1.1x it shows 157 Rated Miles.

 

[qwk]

There was an official reserve? As far I know this "reserve" nonsense has always been a calibration artifact from day 1.

Yes. They took miles off of the top and hid them below zero. Then there was an outrage about zero being zero, and they got rid of it in FW 5.8.4.

 

[Max*]

Again, what did the trip tab of the energy consumption show? Did your green/yellow/red line go WAY below the estimated gray line? Was the estimated gray line set above 0% by the time you reach the SpC? etc.

This will help answer if it's an elevation issue, user error, or something else.

 

[3mp_kwh]

There was an official reserve? As far I know this "reserve" nonsense has always been a calibration artifact from day 1.

Maybe it is just people assuming that the SOC underneath where the car shuts down, is actually available for some extra miles.

IMO, the problem begins with 'Rated Miles' dropping faster than actual miles, almost as a rule. The significant other drives the car, expecting 1:1 miles, and here we are.

 

[yobigd20]

Yes. They took miles off of the top and hid them below zero. Then there was an outrage about zero being zero, and they got rid of it in FW 5.8.4.

where was this documented? where did Tesla confirm it's existence and removal?

 

[stevepeck1]

evtripplanner

is extremely enlightening. Upon entering the exact inputs, it says the car would use 156 Rated Miles, which bears out EXACTLY to the reality we experienced. A few miles short.

Craptastic! :crying:

It would be great if the in-car trip computer had this sort of accuracy.

 

[TexasEV]

Where? Sure wasn't done anywhere around here, and I'm also pretty sure that would be an average speed. The reason lowering speed on flat straight roads increases range is the lower wind resistance and lower C-rate draw from the battery overall as a result. In real world conditions, at 23 MPH you will still pull 40kW+ in some places, far more than wind resistance would cause normally. The ratio of the power increase due to climbing and the power increase due to wind resistance at a given speed is highly variable in the real world.

I wouldn't climb Black Mountain at 23 MPH in my Model S, that's for sure.

I don't recall where it was done, but he gave a presentation about it at TMC 2014.
Also here is the blog post from Tesla showing a computer model of the greatest range about 20 mph. Not that it's practical, but it shows the slower the better until you get below that speed:
Model S Efficiency and Range | Tesla Motors

 

[stopcrazypp]

On the whole subject of reserve, I recall updates removing that too (or at least changing it). I believe it used to be fairly consistent that you would get a reserve below zero, but now more often than not, there is no such reserve.

 

[wk057]

I think the whole mindset of updates modifying the "reserve" are about the same as every update changing how regen works. Basically, it doesn't, but people want to believe it, so... here we are.

The "reserve" has always just been a calibration artifact. I've had a model S pre-5.8 and it didn't have any miles "shift" from under zero to above zero. That would have resulted in a displayed range increase, which never actually happened.

There is no reserve. Never was.​

 

[mikeash]

While I LOVE this car and have enough technical experience to accept the "guesstimate" nature of these batteries, the range on the gauge is entirely unreliable and considerably less than the 265 "advertised."

I think this is the key. It's not a range gauge! It's a battery gauge that uses the highly unconventional unit of "miles" to display battery charge.

Seems like Tesla should change how this is displayed. There's some usefulness in the "rated range" number, but when it's displayed front and center it's way too easy to misunderstand what it means.

Did I imagine this, or was there some discussion about how the 7.0 UI revamp might just show a plain bar without the rated range number?

 

[Skotty]

From the description of the problem, it sounds like it was a legitimate issue at the time and not user error. Difficulty maintaining speed, poor acceleration, quickly dropping rated range, and especially that it seemed to get better after software reset (if I read that correctly). Is this a dual motor car? I wonder if it's possible to get one motor stuck in regen while the other is trying to drive. This would likely cause all of the described issues.

This may be an example of the disadvantage of a high tech car that has a lot of software and that doesn't rely as much on mechnical controls. It can suffer from software errors, and maybe even the occasional so-called blue screen of death. Tesla should build in a self diagnostic feature that can be triggered by an occupant whenever something unusual is occurring. Might help solve less frequent software glitches.

 

[TexasEV]

Even if it displays percent charge remaining, it's just an estimate as rated miles is an estimate.