Weird Road Trip Problem: bad arrival percentages!

[Chris Barnes]

Took my first road trip in my 2024 Tesla Model Y LR DM and for the most part it was FANTASTIC. The drive from Richmond, VA to Houston, TX was LONG and required several stops to charge. The good news is that EVERY ONE of the SuperChargers I encountered worked perfectly - even the ones along I-10 in Mississippi and Louisiana! The car road comfortably and didn't have any physical problems during the trip.

However...

I did start having this weird problem on day 2 of the trip (Montgomery, AL to Houston, TX). I'd charge up and it would show my arrival at the next SC would be with ~20% charge remaining. Cool.

Then I hit the interstate and watched that number drift down to 19...18...17...16... and by the time it got down to 8% I decided to change my plans and stop at the next SC I came across instead of pushing it to Mississippi. Charged up again. Again: 20+ SoC predicted at the Mississippi SC. Hit the road again...

19... 18...17... 16...

WTF?

So I stop at another SC, the 3rd SC on I-65 between Montgomery and Mobile.

Same thing again.

By the time I got to Mobile I was resigned to the idea I'd be stopping at every single SC I came across between Alabama and Texas. But I didn't know why.

Somewhere in Louisiana I tried rebooting the system. Same issue.

At one point I decided to start looking at miles of charge remaining vs distance to the next SC. This is where I noticed what you see in the pic: 193 miles charge remaining, 194 miles to the SC. That means I should get there with 0% remaining. But you can see it says I'll have 11% remaining!

WTF again?

I realize weather (rain, headwinds, etc) and terrain play a role in miles you get from a charge but this was just ODD.

For the rest of the trip I completely ignored percentages and only looked at miles SoC and miles to the SC I was shooting for and everything worked out.

Anyone have any suggestions on what's happening here and/or why I didn't see this happening on Day 1 from Richmond, VA to Montgomery, AL?

Thanks!

 

[vanjwilson]

I had the opposite experience when I first got a 2021 MY LR three years ago: the miles remaining dropped faster than the miles I was driving. I soon switched to percentage instead and I've been happier ever since.

I think Tesla does occasionally tweak their estimation algorithm during OTA updates (like when they started taking more factors into account). And sometimes, a new software update introduces bugs in the estimation, so the version you have may just be off, and the next update will make it better.

I haven't done many long drives in my car since the beginning of the year, so I'm not sure how good it has been lately.

Of course, it is possible that the arrival percentage is dropping because some factor is not what it anticipated. The wind could have picked up, or you could be driving more spiritedly than you realize--speed kills range.

I found myself in a bad cross-wind heading from Death Valley to Las Vegas a few months ago. I watched my 20% arrival percentage start dropping, and dropping at a rate than would have me arriving at a negative SOC! So, I started slowing down until the rate of decline stopped at 7%, and then started climbing back up. I ended up arriving at the first supercharger on the northwest side of Vegas with 11%.

 

[Chris Barnes]

I had the opposite experience when I first got a 2021 MY LR three years ago: the miles remaining dropped faster than the miles I was driving. I soon switched to percentage instead and I've been happier ever since.

I think Tesla does occasionally tweak their estimation algorithm during OTA updates (like when they started taking more factors into account). And sometimes, a new software update introduces bugs in the estimation, so the version you have may just be off, and the next update will make it better.

I haven't done many long drives in my car since the beginning of the year, so I'm not sure how good it has been lately.

Of course, it is possible that the arrival percentage is dropping because some factor is not what it anticipated. The wind could have picked up, or you could be driving more spiritedly than you realize--speed kills range.

I found myself in a bad cross-wind heading from Death Valley to Las Vegas a few months ago. I watched my 20% arrival percentage start dropping, and dropping at a rate than would have me arriving at a negative SOC! So, I started slowing down until the rate of decline stopped at 7%, and then started climbing back up. I ended up arriving at the first supercharger on the northwest side of Vegas with 11%.

Ohhhh... it's all AI... as opposed to actual intelligence. LOL. Thanks for the reply - bottom line is that I made it in one piece... but not exactly user friendly.

 

[scottf200]

You did not mention one of the *main* ingredients and that is how fast you were driving.

In all my trips I look on the energy graph and the 'projected range' for the last 30 miles if I'm questioning how it is going. I probably have 100K of road tripping experience at this point.

Tips here:

How to accurately predict range in a Tesla

When I was researching Tesla, I found countless videos and blog posts on how range is impacted by cold, wind, speed, terrain, and various settings on the car itself, which led me to believe that road trips were going to require a lot of planning and energy management. Now that

www.gregsramblings.com

Some terminology that I had saved off.

More 'realistic' range: 'rated' in USA & 'typical' in non-USA
'Constant value' range: 'ideal' in USA & 'rated' in non-USA
Typical, rated and ideal range explained

 

[IronCoffee_Max]

My suggestion - if it's over 2%, stick with it. I prefer to get to a SC at 1-5%. 8% is good. 20% is overkill. Too conservative.

 

[karkrazy]

Headwinds are also a killer on range; can’t really predict this.

 

[Chris Barnes]

20% is overkill. Too conservative.

Well... to be fair - it was 2 AM on I-65 in southern Alabama. Not sure if you've ever driven that stretch but it's no place to run out of "gas".

 

[vanjwilson]

Well... to be fair - it was 2 AM on I-65 in southern Alabama. Not sure if you've ever driven that stretch but it's no place to run out of "gas".

On a college road trip back in the 90s, my friends and I were driving through the night down I-20, when we stopped to take a picture of the exit sign for Chunky, MS, because we thought it was a funny name.

About half an hour later, a little before dawn, we stopped at another exit to switch drivers. While we were in the parking lot of a closed gas station, a big 4x4 lifted pickup truck with a gun rack pull up beside us. We were all very worried.

The driver rolled down his window, leaned out of the truck, and said, "We're supposed to meet some guys to go hunting this morning. They said to take the exit for a place called Chunky. Do y'all know where that is?"

 

[Mr. Nuke]

Anyone have any suggestions on what's happening here and/or why I didn't see this happening on Day 1 from Richmond, VA to Montgomery, AL?

Trying to figure out what happened in the absence of more information post hoc is going to be difficult if not impossible. Assuming you were driving in a similar manner both days, given that it only happened in the 2nd day of the trip and that it seemingly repeated itself; my best guess is you were encountering stronger headwinds or (potentially lesser tailwinds or a crosswind having more impact) than the car was anticipating on the 2nd day.

I did a 400 mile round trip from Omaha to Kansas City and back last week. The car was incredibly efficient on the way down exceeding the car's initial estimate despite it not knowing I'd be driving 8-9 mph over the speed limit, as well as ABRP's estimate with it knowing my intended speed. On the way back after supercharging in St. Joseph Missouri, I experienced similar behavior as you described where the estimated arrival percentage dropped as I drove back towards Omaha.

In my situation I had the benefit of having done this drive the opposite direction 3 hours prior to know that this was likely going to happen and knew to alter my charging in St. Joseph appropriately. I had made a mental note of what my percentage was at passing this charger on the way down and knew that I had benefited from a tailwind and would be heading into the wind on the way back.

At one point I decided to start looking at miles of charge remaining vs distance to the next SC. This is where I noticed what you see in the pic: 193 miles charge remaining, 194 miles to the SC. That means I should get there with 0% remaining. But you can see it says I'll have 11% remaining!

@scottf200 got into this with their post, but one key thing to recognize is the 193 miles on the upper left corner is rated range. Some still swear by it, but the numerical value is simply a function of your current battery percentage related to the "rated" EPA wh/mi/mileage test. If you are driving on a flat road in good weather conditions at 55 mph it is going to be highly accurate. In most other conditions, I personally think it has significantly less value. This is why a lot of people like @vanjwilson and myself have switched that to battey percentage.

I think it does have some value in this instance, because if you do realize the 193 miles is the current EPA rated state of charge on the battery and you know you are about to drive 194 miles on the interstate where you are unlikely to acheive EPA rated mileage, yeah it should be a red flag that you need to charge more. In a theoretical sense here though if the car was expecting something like a strong tailwind, It would be entirely possible for your 193 miles in the upper left to get you 194 actual miles with 10% remaining. Something like that would potentially explain why the estimates kept dropping when the car started to realize it wasn't doing as well as it anticipated. Going back to my KC drive, had I been driving the speed limit on the way down I would've likely exceeded "rated" miles while doing 70 mph, because I got very close doing 78-79.

For the rest of the trip I completely ignored percentages and only looked at miles SoC and miles to the SC I was shooting for and everything worked out.

This is probably the last thing I personally would've done, but I'm glad it worked out for you. If you were driving in a manner though to make sure the rated miles in the upper left alway exceeded the remaining miles to your destination it will work. Arguably though there are a couple of ways of doing this better/easier and I think more importantly one of them may have helped figure out what was wrong with the estimates.

The first thing if the arrival percentage is dropping faster than you anticipated and below where you want it to be is to slow down as @vanjwilson mentioned. I would also get familar with the energy screen.

This is going to basically be the same screen that @scottf200 posted above from the energy app. A couple of things to note. Again the 251 miles in the upper left is the rated miles. However, based on the last 30 miles of driving here, the car only thinks it will make it 198 miles. Driving in a straight line on a relatively flat interstate with constant speed cruise control that line is going to be a lot smoother as your consumption is going to be relatively constant. That number though has a lot more value to me in the upper left. If you are good within your margin of safety there you are going to be good (unless you anticipate conditions to change at some point i.e. weather, etc.).

The other screen that is highly valuable in a road trip and the one that I would've liked to see for your drive is the trip screen from the energy app.

This screen will show you exactly how well you are doing versus what the car predicted and if it is deviating from the prediction why. The vehicle above is doing 1.8 miles worse than expected over a 10.2 mile drive. In your case, it is pretty obvious on the 2nd day the car was having a hard time estimating something. This screen likely would've indicated the issue. From there you kind of know what is going on and what you might need to do. If it were stronger or weaker winds it would let you know there. Regardless though if the car was continuously underestimating the 20% arrival time at the very least it was a sign to charge more at each supercharger than what the car was telling you.

 

[Chris Barnes]

Trying to figure out what happened in the absence of more information post hoc is going to be difficult if not impossible. Assuming you were driving in a similar manner both days, given that it only happened in the 2nd day of the trip and that it seemingly repeated itself; my best guess is you were encountering stronger headwinds or (potentially lesser tailwinds or a crosswind having more impact) than the car was anticipating on the 2nd day.

I did a 400 mile round trip from Omaha to Kansas City and back last week. The car was incredibly efficient on the way down exceeding the car's initial estimate despite it not knowing I'd be driving 8-9 mph over the speed limit, as well as ABRP's estimate with it knowing my intended speed. On the way back after supercharging in St. Joseph Missouri, I experienced similar behavior as you described where the estimated arrival percentage dropped as I drove back towards Omaha.

In my situation I had the benefit of having done this drive the opposite direction 3 hours prior to know that this was likely going to happen and knew to alter my charging in St. Joseph appropriately. I had made a mental note of what my percentage was at passing this charger on the way down and knew that I had benefited from a tailwind and would be heading into the wind on the way back.

@scottf200 got into this with their post, but one key thing to recognize is the 193 miles on the upper left corner is rated range. Some still swear by it, but the numerical value is simply a function of your current battery percentage related to the "rated" EPA wh/mi/mileage test. If you are driving on a flat road in good weather conditions at 55 mph it is going to be highly accurate. In most other conditions, I personally think it has significantly less value. This is why a lot of people like @vanjwilson and myself have switched that to battey percentage.

I think it does have some value in this instance, because if you do realize the 193 miles is the current EPA rated state of charge on the battery and you know you are about to drive 194 miles on the interstate where you are unlikely to acheive EPA rated mileage, yeah it should be a red flag that you need to charge more. In a theoretical sense here though if the car was expecting something like a strong tailwind, It would be entirely possible for your 193 miles in the upper left to get you 194 actual miles with 10% remaining. Something like that would potentially explain why the estimates kept dropping when the car started to realize it wasn't doing as well as it anticipated. Going back to my KC drive, had I been driving the speed limit on the way down I would've likely exceeded "rated" miles while doing 70 mph, because I got very close doing 78-79.

This is probably the last thing I personally would've done, but I'm glad it worked out for you. If you were driving in a manner though to make sure the rated miles in the upper left alway exceeded the remaining miles to your destination it will work. Arguably though there are a couple of ways of doing this better/easier and I think more importantly one of them may have helped figure out what was wrong with the estimates.

The first thing if the arrival percentage is dropping faster than you anticipated and below where you want it to be is to slow down as @vanjwilson mentioned. I would also get familar with the energy screen.


View attachment 1047842
This is going to basically be the same screen that @scottf200 posted above from the energy app. A couple of things to note. Again the 251 miles in the upper left is the rated miles. However, based on the last 30 miles of driving here, the car only thinks it will make it 198 miles. Driving in a straight line on a relatively flat interstate with constant speed cruise control that line is going to be a lot smoother as your consumption is going to be relatively constant. That number though has a lot more value to me in the upper left. If you are good within your margin of safety there you are going to be good (unless you anticipate conditions to change at some point i.e. weather, etc.).

The other screen that is highly valuable in a road trip and the one that I would've liked to see for your drive is the trip screen from the energy app.

This screen will show you exactly how well you are doing versus what the car predicted and if it is deviating from the prediction why. The vehicle above is doing 1.8 miles worse than expected over a 10.2 mile drive. In your case, it is pretty obvious on the 2nd day the car was having a hard time estimating something. This screen likely would've indicated the issue. From there you kind of know what is going on and what you might need to do. If it were stronger or weaker winds it would let you know there. Regardless though if the car was continuously underestimating the 20% arrival time at the very least it was a sign to charge more at each supercharger than what the car was telling you.

Thanks for the detailed response! I made it work in the end but this was the first unexpected thing I've experienced after owning 2 EVs.


I do worry about people coming over from gas vehicles that are used to watching the needle go from F to E and otherwise not thinking about anything. To me this was the equivalent of watching the gas needle go from 3/4 full to 1/4 full in a matter of seconds. It would cause anyone to pull over and figure out what the heck had happened!

 

[scottf200]

Thanks for the detailed response! I made it work in the end but this was the first unexpected thing I've experienced after owning 2 EVs.

Unless I missed it you haven't explained your what speeds that you were traveling as some of us alluded to being a major factor.

 

[LostVector]

Unless I missed it you haven't explained your what speeds that you were traveling as some of us alluded to being a major factor.

Yes, speed is probably the only actual number that matters re understanding the discrepancy.

 

[Chris Barnes]

Apologies if I missed it... both days I drove the same: speed limit or speed limit +5 (otherwise people will run you off the road). So max speed both days was 75 mph. I drive like grandpa and shake my fist at the rest of the world for driving like FREAKING IDIOTS out there.

 

[Ciaopec]

Apologies if I missed it... both days I drove the same: speed limit or speed limit +5 (otherwise people will run you off the road). So max speed both days was 75 mph. I drive like grandpa and shake my fist at the rest of the world for driving like FREAKING IDIOTS out there.

As someone mentioned: perhaps high headwinds. My 2016 MS and 2021 MY both seem to like 75mph as a sweet spot. Slowing down is the best thing to reduce consumption. Slowing by 5mph on a 2 hour leg is only about 10 minutes. I still continue to charge to about 55% max at my charge stops and then head off to my next stop, usually a charger or two away shooting to reach it at about 10%. If things go south I can always stop at the intermediate charger if slowing down doesn't rectify the problem.