Range Failure

[juliusa]

Passing a supercharger because it says you will have 12% remaining at your destination is a completely separate choice where the previous "5%" input is irrelevant stale data.

Seems the 5% was the correct data, and the reported observation of 12% remaining was the problem.

That is why Boeing is changing their software that previously gave pilots only 40 seconds to switch off a system that relied on one sensor to force the plane into a dive, believing (incorrectly) that the plane was about to stall. The system will now reportedly require two sensors (the second sensor used to be an option that those two planes did not have), and after putting the plane into a dive twice it will automatically switch off.

The OP did absolutely nothing wrong. He did not charge enough to reach his destination, passed by a supercharger thinking the car had enough charge to reach his destination, and then decided he needed more charge and stopped to charge. I find absolutely nothing wrong with that.

The instrumentation in my Teslas behaves very differently. I have never observed the big changes in % charge remaining on arrival in such short periods of time, and get constant nags from the nav, which is such a conservative worry wart, if I will arrive with a low SOC.

There are many possible explanations for the scenario posted. The lack of repeatablity, the lack of other similar experiences, and the lack of a "black box" to provide more precise data, all lead me to the conclusion this is not a systemic issue to be worried about at this time. My opinion would change if any of those things change. They did not ground the 737 until there were two incidents deemed to be related.

 

[MXn00b]

It's interesting that coming back to the TMC after a couple months, i see the same subset of users who are either love their tesla, or complain about their tesla. Like you don't see a owner in both categories, its usually just one or another.

You must not have caught up on my posts, either

I was expecting a rational discussion. Some people here seem to want to blame me for everything including kidnapping the Lindbergh baby. Sorry if you feel I am picking on you.
Even if I could see on the graph that the consumption increased, what would that tell me other than what I already know, the consumption increased? I still would not know why.

It could tell you that your current consumption has increased or your expected consumption increased. That's the real question: Did the estimate change because it estimated that you had less battery than expected or that consumption to your destination was estimated to be higher given different data? That is why your Wh/mi is so important to know.

It's very frustrating to expend more energy than expected, but we can't expect the car to understand better than us. For some things, yes, I would love to nerd out on Trip Planner inputs to customize speed, temperature, etc. because there are known inputs it does obscure. On my -3% expected arrival, I was pretty sure I could make it given elevation, environment, and my control of speed. I would like to know exactly how close I cut it, though. I think I slowed down enough to make it with reported 2%.

My estimate was definitely less than 5%, which has been shown to be a magic number. That's why people keep bringing it up. Think of it like 0% on your old iPhone. Yeah, it just died, but you can turn it back on and Candy Crush for a few more minutes before it dies again. To @gnuarm, I'm not saying you're wrong to leave at 5% when explicitly telling the Nav to ignore charging stops. I'm just trying to provide possibilities and help you plan trips better than Blanche Dubois. My guess is you could have charged 10 more minutes to get to 6% and this would be a different discussion, if one at all.

It is scary to think that the car would screw up the range estimate enough to drop more than 11% off the actual range. While Blanche Dubois may have always depended on the kindness of strangers, it's not how I want to plan trips in my Tesla.

What could have happened that the range estimate would continue to drop like this?

I haven't been able to test on any long trips since this post, but I do have a short commute that regularly adds 2 mins through a congested area, then makes up 2 minutes after. Assuming the algorithms work similarly, they are overreacting to something that is expected. This is exemplified by your new destinations having more pessimistic estimates. My guess is you drove slower than expected on the highway, had a slight tailwind, or some other situation that lowered your Wh/mi on the first part of your trip, causing the estimate to rise.

Maybe the initial estimate isn't any better or worse than any updated estimate, but it does not give you any variance. If the initial estimate can change by 7%, that could go either way. If it mistakenly starts trending up, the correction down will be more severe. One thing that usually helps my range anxiety is the one free tow for running out of battery. Have you used that? Don't you ever wonder how much longer you could've lasted?

 

[mxnym]

One thing that usually helps my range anxiety is the one free tow for running out of battery. Have you used that? Don't you ever wonder how much longer you could've lasted?

Not to cause undue stress, but are you sure this is a thing? IIRC, the fine print for the roadside assistance specifically excludes running out of battery without any implicit or explicit guarantee of a one time exception.

 

[MXn00b]

Not to cause undue stress, but are you sure this is a thing?

You did very poorly trying to avoid undue stress. There is no "one time" exception specified in my roadside assistance policy. I guess I won't risk leaving with 5% or less anymore. As I recall I heard this from everyone including my salesperson up to my delivery specialist. FWIW, I popped a tire on a rock covered in deep snow and was not charged for that tow.

 

[Krazaak]

You did very poorly trying to avoid undue stress. There is no "one time" exception specified in my roadside assistance policy. I guess I won't risk leaving with 5% or less anymore. As I recall I heard this from everyone including my salesperson up to my delivery specialist. FWIW, I popped a tire on a rock covered in deep snow and was not charged for that tow.

A flat is usually out of your control, running out of charge usually isn't.

 

[ditzpro]

Hopefully this is pertaining to this thread ..I am a little freaked ..I have a 1 year (26K miles) X...Very little issues with it ..I make a 250 mile trip about 3 times a year which is the only time I use a supercharger . In December I charged to 292 miles at home to make the trip , stopping at a supercharger about 1/2 way to top up .. In March (this year , 4 months later ) My max 100% charge is 256 miles . Repeated charges to 100% are the same 256 miles ..This seems to me to be an excessive loss is only 4 months .. The little l
tle battery indicator only shows about 90% charge ( a small space still at the end ,empty) ...If the car thinks it's at 100% shouldn't this show full ?...I'm really confused and trying to call service to discuss was pretty much fruitless ..Any comments would be very helpful .
Thanks

 

[GHammer]

Hopefully this is pertaining to this thread ..I am a little freaked ..I have a 1 year (26K miles) X...Very little issues with it ..I make a 250 mile trip about 3 times a year which is the only time I use a supercharger . In December I charged to 292 miles at home to make the trip , stopping at a supercharger about 1/2 way to top up .. In March (this year , 4 months later ) My max 100% charge is 256 miles . Repeated charges to 100% are the same 256 miles ..This seems to me to be an excessive loss is only 4 months .. The little l
tle battery indicator only shows about 90% charge ( a small space still at the end ,empty) ...If the car thinks it's at 100% shouldn't this show full ?...I'm really confused and trying to call service to discuss was pretty much fruitless ..Any comments would be very helpful .
Thanks

Sounds to me like you have the charge limit set at 90%. When you switch back and forth between miles and percentage display what does it say for each?

 

[Krazaak]

Hopefully this is pertaining to this thread ..I am a little freaked ..I have a 1 year (26K miles) X...Very little issues with it ..I make a 250 mile trip about 3 times a year which is the only time I use a supercharger . In December I charged to 292 miles at home to make the trip , stopping at a supercharger about 1/2 way to top up .. In March (this year , 4 months later ) My max 100% charge is 256 miles . Repeated charges to 100% are the same 256 miles ..This seems to me to be an excessive loss is only 4 months .. The little l
tle battery indicator only shows about 90% charge ( a small space still at the end ,empty) ...If the car thinks it's at 100% shouldn't this show full ?...I'm really confused and trying to call service to discuss was pretty much fruitless ..Any comments would be very helpful .
Thanks

My 2016 P100D w/ 30k miles charges to 256-258 at 90%. If you charged to 256 miles at 100%, there's likely nothing that could be done because 10% degradation is high, but not covered. If you can't charge past 90% though, that's an entirely different issue. After double checking that you've selected 100% charge, I'd make an SC appointment.

 

[Rocky_H]

The little l
tle battery indicator only shows about 90% charge ( a small space still at the end ,empty) ...If the car thinks it's at 100% shouldn't this show full ?...I'm really confused and trying to call service to discuss was pretty much fruitless

Sounds to me like you have the charge limit set at 90%. When you switch back and forth between miles and percentage display what does it say for each?

I also was a little unclear of what was actually being done and shown. In the charging screen, make sure the limit is set all the way up, and as @[email protected] suggested, see what the display says on % to see what the thinks it's doing. I have heard of this symptom before a few times, and it's just the car's systems having a little glitch and being confused. A reboot generally fixes it.

 

[mxnym]

You did very poorly trying to avoid undue stress. There is no "one time" exception specified in my roadside assistance policy. I guess I won't risk leaving with 5% or less anymore. As I recall I heard this from everyone including my salesperson up to my delivery specialist. FWIW, I popped a tire on a rock covered in deep snow and was not charged for that tow.

Undue stress would be if I was wrong. The stress is both due and beneficial if I was right. Better to find out now than on the side of the road. If it ever happens, I'd still call in case they do it once as a courtesy, but as you have seen, they certainly don't as a matter of contract. Personally, I signed up for AAA+ when I found this (and that they don't always have a spare tire near you) out, and while I hope to never need to use either, I probably won't be finding out whether or not Tesla would give me a courtesy tow for a dead battery.

 

[acoste]

Letting it get to a higher temperature is different from heating it.

It is literally heating it until target temperature is met. One user showed that the incoming (into the battery) coolant temperature became higher than the outgoing when he turned range mode on.

 

[mxnym]

It is literally heating it until target temperature is met. One user showed that the incoming (into the battery) coolant temperature became higher than the outgoing when he turned range mode on.

Just because the coolant is hotter than the battery doesn't mean it was heated. There is no logical scenario where energy wasted on heat* will increase range, that's like 7th grade physics.

*ETA: while not plugged in (heating the battery while plugged in could increase range, but we are discussing using range mode while driving)

 

[Krazaak]

Just because the coolant is hotter than the battery doesn't mean it was heated. There is no logical scenario where energy wasted on heat* will increase range, that's like 7th grade physics.

*ETA: while not plugged in (heating the battery while plugged in could increase range, but we are discussing using range mode while driving)

Definitely, I'd suggest that "allowing the battery to heat" is a more accurate description than "heating the battery" it sounds to me that it's simply using drive unit waste heat to warm everything, bypassing the radiator and keeping the chiller or battery heater offline.

 

[Alphonso14]

5% in particular is a dangerous number. If you were, say, slightly below 5% under normal expectations, sometimes the nav will “helpfully” assume that you’ll slow down and then give you that 5% number again. To show this, enter a destination that’s beyond your range right before you supercharge. Hit “continue trip”. Then bring up the trip graph. As you supercharge, once the projected arrival battery percentage gets to 5%, it stays there for awhile while the usage curve steepens. Then once it’s done steepening fully (as it assumes faster and faster driving speeds for the trip) the percent increases again beyond 5%.

TL;DR: 6% is actually way, way safer than 5% — much more than 1% more margin.

Only had the Model X 75D for 6 months, 6,000 miles. Two trips, one 550 miles, one 300 miles. The longer trip was from Newport Beach CA to Incline Village NV in cold weather. Highway driving usually at 75+ mph. Overall energy used was 25% (for the entire trip) more that what it should have been for a range of 237 miles. The last segment was 98 miles going from sea level to a max of 7700 feet in temperatures in the 20s. It took a bit more than 150 miles of energy which means I paid a 50% energy premium because of climbing and cold weather. Still suffer from occasional range anxiety so I appreciate the comments on how to monitor "distance remaining".

 

[Alphonso14]

I'm responding to comments like this...



I have been driving the car for 15,000 miles and been talking about my experiences here for pretty much all of that time. I get every response in the book with some telling me "this" is the way to figure your range and others saying that is a terrible idea and you should always do "that". I'm pretty fed up with so much of it being unsubstantiated and based on anecdote.

I read everything everyone writes. Maybe I misunderstand sometimes, like the graph I had forgotten about. But I really don't see where it gives me any more data than I have using the estimated remaining charge number. Seeing the graph diverge is harder than seeing the number diverge and tells me little more. I suppose it might shed some light on the remaining charge estimate rising early in the trip. But it wouldn't be useful for seeing what to expect.

With the limited range of the cars and the relative paucity of charging in some areas, it is important for the car to reliably predict the range. If EVs are ever going to pan out for the masses they have to deal with the issue so that the typical person won't have to think about it any more than they do now with gas cars. I just spoke to someone while I was charging who appeared to have a negative impression of the whole electric car thing and I think wanted to know about them so he could dissuade his son. I told him the facts rather than sugar coating it. I pointed out that if you aren't taking trips in the car charging at home is the best way and nightly charging is perfectly acceptable. However on trips you have to plan and drive very differently than in an ICE vehicle.

 

[Alphonso14]

IMHO the problem with electric cars and range anxiety and public perception has not as much to do with 100% range calculation accuracy as it does with availability of Supercharge stations along the way. ICEs don't give you an accurate estimate of distance remaining in the tank but you know that if you are running low you can just pull off the freeway at the next exit and find a gas station...at least in So Cal. Trips in an EV require more planning just as it did 70 years ago when my Dad would take us on road trips out west.

 

[Freerster]

Each Tesla is different. You need to know your car. What your car can and cant do in different senerios. The only way to do that is to test it out and experiance with electric cars. I have ran my X 2 times to 0% and the super charger was 6 and 16 miles away. I made it both times to the super charger. It depends on many different veriables. (Dont get me wrong both times I had puckered checks). Just get to know your car and it will take care of you. Dont be a fool and pass a super charger if your not sure. Pop in for 15 mins and be off again and make it with no problem. The super charging network is getting better all the time. Use the web site supercharge.info it helps alot too if your not sure.

 

[Krazaak]

The single greatest tool at your disposal for range planning is the trip graph in the energy app. You can see projected battery SoC over every mile of your trip, accounting for speed and elevation changes. I recently drove from Belmont, NC to the supercharger in Bristol, TN through the mountains via Morganton, NC. I paid extra attention to the trip view, since this thread was recent in my mind. Started at 94% SoC and my first observation was that regen is severely limited above 90% SoC. Even at 91%, I was getting at best half of the max regen that it's capable of, meaning any stop/go driving is going to be much less efficient unless you can adapt. I'm very used to full regen so I found myself having to brake for the first 5 or so miles to the highway.

From the start, predicted arrival was 31% SoC, 160 miles away, meaning 63% of my battery. That translates to 262 miles of capacity at 100% charge vs. the rated range of 289 miles. Pretty good in my opinion, since I'd be driving over the mountains. My speed typically averages 5-10 over the limit, so maxing out around 75 mph on the highway stretches.

About 75 miles into the trip, about the point where I was getting to the mountain roads past Morganton, predicted arrival had improved to 35% SoC. I was all set to see how efficiently I could get through the mountains until two important factors converged. First, nobody was in front of me and second, my wife wasn't in the car to complain about getting motion sick. So, any thought of efficiency went out the window and I proceeded to haul ass up the mountain and see how the X stacked up to my torque vectored BMW X6M. The next 15 miles were well over 900 Wh/mi and well worth it, and as the arrival SoC dropped, it only went back to 31% It was obvious from the slope of the graph that it had already anticipated at least double the consumption in the mountains, my spirited driving only consumed the power I'd saved on the previous 75 miles, despite making no conscious effort to be efficient.

I continued my driving entertainment on the down slope as well and although the trip graph had shown me recouping a percent or so of battery through regen, the energy wasted by driving like a lunatic through corners meant I was still burning power instead of gaining.

I picked up efficiency again once I hit boring straight roads and arrived with 32%, 1% over the original target.

 

[gnuarm]

IMHO the problem with electric cars and range anxiety and public perception has not as much to do with 100% range calculation accuracy as it does with availability of Supercharge stations along the way. ICEs don't give you an accurate estimate of distance remaining in the tank but you know that if you are running low you can just pull off the freeway at the next exit and find a gas station...at least in So Cal. Trips in an EV require more planning just as it did 70 years ago when my Dad would take us on road trips out west.

That is somewhat correct. But there is nothing we can do about the location of chargers, so it is no small importance to get the trip planning and range thing right.

I would disagree with the idea that you don't have any better estimate of range in an ICE auto. I had a truck that I could drive into the filling station with less than half a gallon of gas with no range failures... ever - in 20 years! The gas gauge had a warning light that came on with more than 2 gallons left. In addition the mileage was consistent enough that I could always be assured of getting 400 miles out of the max possible 450 miles. In the Tesla I have a hard time using the all but 50 miles on a 290 mile range.

I'm learning to live with the idea that I have to be very vigilant about charging. Winter trips are going to need to be planned much more carefully. But I have multiple bones to pick with Tesla about the tools provided. I'm sorry, but the present in car trip planner is far short of what is desired. The two big short comings is that it can only plan between your present location and a single destination and that it won't let you plan any route other than the one it comes up with. For route planning any real trips even Google is better than the car is. If only I could get the browser in the car to work properly that would be a big benefit. Weren't they supposed to upgrade the browser to Chrome or something?

 

[gnuarm]

The single greatest tool at your disposal for range planning is the trip graph in the energy app. You can see projected battery SoC over every mile of your trip, accounting for speed and elevation changes. I recently drove from Belmont, NC to the supercharger in Bristol, TN through the mountains via Morganton, NC. I paid extra attention to the trip view, since this thread was recent in my mind. Started at 94% SoC and my first observation was that regen is severely limited above 90% SoC. Even at 91%, I was getting at best half of the max regen that it's capable of, meaning any stop/go driving is going to be much less efficient unless you can adapt. I'm very used to full regen so I found myself having to brake for the first 5 or so miles to the highway.

I don't agree about the "severely" limited regen at 91% charge. I monitored it carefully one time and at 100% obviously the regen was zip. At 98% it was usable for gentle driving. By 95% it was about half of full capability which means no degradation unless you do hard engine braking. If you don't drive aggressively (and don't use the blinking autopilot) you don't need aggressive engine braking. Even at 98%, it's not like you have lost all benefit. You lose the peaks only. So I think the mileage impact of this is often overstated.

From the start, predicted arrival was 31% SoC, 160 miles away, meaning 63% of my battery. That translates to 262 miles of capacity at 100% charge vs. the rated range of 289 miles. Pretty good in my opinion, since I'd be driving over the mountains. My speed typically averages 5-10 over the limit, so maxing out around 75 mph on the highway stretches.

About 75 miles into the trip, about the point where I was getting to the mountain roads past Morganton, predicted arrival had improved to 35% SoC. I was all set to see how efficiently I could get through the mountains until two important factors converged. First, nobody was in front of me and second, my wife wasn't in the car to complain about getting motion sick. So, any thought of efficiency went out the window and I proceeded to haul ass up the mountain and see how the X stacked up to my torque vectored BMW X6M. The next 15 miles were well over 900 Wh/mi and well worth it, and as the arrival SoC dropped, it only went back to 31% It was obvious from the slope of the graph that it had already anticipated at least double the consumption in the mountains, my spirited driving only consumed the power I'd saved on the previous 75 miles, despite making no conscious effort to be efficient.

I continued my driving entertainment on the down slope as well and although the trip graph had shown me recouping a percent or so of battery through regen, the energy wasted by driving like a lunatic through corners meant I was still burning power instead of gaining.

I picked up efficiency again once I hit boring straight roads and arrived with 32%, 1% over the original target.

I'm not sure what we should learn from this. You originally mentioned that the trip graph was the "greatest tool" available, then you don't really mention it any further.

I have been using the trip graph and I don't see any information I don't get from watching the estimated remaining charge. In fact, the remaining charge estimate is much more sensitive at the beginning showing numbers. While those same numbers are shown in the trip graph, the curve itself doesn't diverge much early in the trip since the change in final charge requires a much smaller deviation in the slope of the line in the early part than subsequent changes later along the trip which require more significant changes in the line slope. I took some photos of my trip graphs and they nearly all show no apparent change in the graph in the earlier portion of the trip even though most of the predicted improvement in final SoC happened there.



So why is this view any better than just monitoring the estimated final SoC?