Route planning and supercharger issue

[DougIngraham]

I've had my 2013 Model S for 9 months now and have put 15000 miles on her. About 6000 of those miles were during two long road trips with another coming up in late November. What I have learned is that the charge estimator works really really well except in places where the speed limit on the interstates is 80 mph or more. For the most part the interstate speed limits are posted at 75 mph or less except in the following states. Montana, Nebraska, Nevada, South Dakota, Texas, Utah, and Wyoming. There are large stretches of interstate that are posted at 80 mph or more in all of those states. And if you trust the on board nav you cannot go 80 mph because it will tell you that you have enough energy and you won't. Only 15 or 20 miles after leaving the supercharger you will get the notification that you need to slow down in order to make your destination.

I had two ideas as to why this might be happening and I think I know which one it is. Idea one was that the speed limit portion of the mapping database was incorrect and had those sections of interstate set to a lower speed. Idea two was that the calculator code has an over zealous sanity check such that speeds over 70 or 75 mph are treated as if they are incorrect and a more reasonable value is substituted. It occurred to me that it is easy to find out what the nav system thinks the speed limit is. Just get on the interstate and run up to 80 mph and then navigate to the next supercharger. The nav will tell you both the mileage and the time and with that you can calculate what it thinks the speed limit is. I live in Rapid City, SD and the speed limit between Rapid City and the Murdo SD supercharger is 80 mph. This is about 140 miles and about 130 miles of it is posted at 80. I waited until I reached the 80 mph section and punched up the Murdo supercharger and it gave me a driving distance of 131 miles and a travel time of 122 minutes. If you work this out you get a speed of 64.43 mph (60 * (131 miles /122 minutes). And if the energy predictor bases its calculation on you traveling at 65 mph and you go 80 you will fall short. I suspect the sanity check sets the speed to 65 miles per hour when it sees a speed in the map database greater than 70 or 75. This was either done in the Tesla code when the energy calculation takes place or at Google when the speed limit database was generated. After all the max speed limit in California is 65 and if you don't take road trips it would be easy to make the assumption that 80 or more mph would be an impossible number.

I am hoping that this is corrected in the about to come out 8.0 software release but I don't expect it. For the right person this would take only a few minutes to correct.

Doug Ingraham
2013 Model S 85 Multi-coat Red "The Woman in the Red Dress"

 

[TexasEV]

If this is true, it's another example of Tesla's California-centric thinking. Just like the range per charge calculator on the Tesla website tops out at 70 mph.

 

[DougIngraham]

I assumed the range per charge calculator having a 70 mph limit was for marketing reasons. Your range is going to go down dramatically at speeds above that. But I suppose in California 70 would seem plenty fast and for most of the country as well. I think I see a little over 400 wH/mile when going 80. And at 65 I see around 300 wH/mi.

Just keep in mind that if traveling at 80+ mph you will need between 20% and 25% additional capacity beyond what the calculator tells you. And even more if there if the road is wet or snow packed or a headwind or the temps are excessively low.

Still waiting for my update to 8.0 so I can report on if it is fixed.

Doug Ingraham
2013 Model S 85 Multi-coat Red "The Woman in the Red Dress"

 

[DougIngraham]

My car got it's update earlier this week and I was able to retest what it believes is the speed limit on I-90 between Rapid City and the Murdo SD super charger. And 8.0 still believes the speed limit is 65 mph.

Heading east out of Rapid City I set the Nav for Murdo SC as soon as the speed limit went to 80 mph (129 kph) and recorded the following:

Distance 121 miles.
Time 1 hour 55 minutes.
Current state of charge 73%
Arrival estimate of SOC 27%

I then drove to the next exit at New Underwood and just before exiting recorded this:

Distance 114 miles.
Time 1 hour 47 minutes.
Current state of charge 70%
Arrival estimate of SOC 23%
Consumption of 328 wH/mi.

Clearly from the 121 miles and the 1.92 hours of travel time the car believes the average speed is going to be about 63 miles per hour. In just the 7 miles traveled the arrival SOC estimate dropped by 4%. If it continued dropping at that rate the car would be telling me to slow down in just another 30 miles. It would reach an extimated 6% SOC remaining at that point and in the past that is where the car goes into Nag Mode. I know that I would make if I start with an additional 15% beyond what the software tells you. You arrive with about 10% remaining which is what the planner software seems to try to achieve.

In summary: The 8.0 release does not correct the Nav planner issue where the speed limit is posted at 80 or higher MPH. The car pretends the speed limit is 65 and calculates too little energy because of it.

Doug Ingraham

 

[hiroshiy]

Just a wild guess, depending on the highway route and time of the day, average speed of the car relative to the speed limit is different.

For example in Tokyo:
Shuto Expressway limit 60km/h traffic 80km/h 33% faster
Tomei Expressway limit 100km/h traffic 80km/h 20% slower
Ise Wangan Expressway limit 100km/h 130km/h 30% faster

So I wildly guess that Trip Planner just assume 65mph to calculate like Rated Range display. You will need to adjust based on your speed and wind speed/direction, elevation etc.

 

[ecarfan]

I assumed the range per charge calculator having a 70 mph limit was for marketing reasons. Your range is going to go down dramatically at speeds above that. But I suppose in California 70 would seem plenty fast and for most of the country as well. I think I see a little over 400 wH/mile when going 80. And at 65 I see around 300 wH/mi.

Yes, at speeds over 65 to 70 energy usage goes way up. While ideally the Nav should use the posted limit to calculate range to a Supercharger stop and the display should tell you that, if you are in a state where you can legally go 80 or more and you decide to travel that fast you may have trouble making it to the next Supercharger.

With current battery energy density it just doesn't make much sense to go much over 75 in a Tesla unless your next charging stop is your destination where you have lots of time to charge. If your next charging stop is a Supercharger, and that is not your final destination, it makes more sense to go slower so you don't have to spend as much time charging.

 

[DougIngraham]

I have driven enough to know that at posted speeds of 70 and below the planner gets it right within just a couple of percent. It is only on higher speeds that it just ignores the speed limit and assumes 65. For it to get it right at 70 but fail so badly at 80 shows that they are computing the additional drag losses that occur both with rolling resistance and air drag and then use the wrong speed in the calculation. If you are going to blow it you should do so on the side of caution. Underestimating charge is the wrong way to go. I would much rather spend 15 additional minutes to get the charge I need at a super charger than white nuckle it for an additional 25 worried I am not going to make it being forced to slow down to below 65. This is pretty much the only time I have had range anxiety. It especially annoys me that it is so good the rest of the time.

The equations to perform these computations are fairly simple. Tesla appears take into account everything but temperature, wind, road surface conditions, and variations in mass of the vehicle due to loading. Vehicle mass does make a small difference in the rolling resistance and affects energy usage when climbing hills and the amount of energy recovered by regen when descending. But climbing and descending tend to cancel out mostly. And the number they use appears to be for two people with quite a bit of luggage. I have traveled alone and I pack lightly so this tends to help me out. (I see the % of remaining charge estimate go up when climbing hills and go down when descending.)

It would be interesting to know if the Auto Pilot feature that shows you the speed limit on the dash display knows the correct speed on these 80 mph and greater posted stretches of interstate highway. If it does then the left hand does not know what the right hand is doing and the autopilot database needs to be consolidated with the navigation database.

These cars are so good that when you do find a problem it amplifies the severity. This issue is hugely annoying to me especially when I know I could fix it in less than a week and it would only take that long because I am not at all familiar with their code. And there is nowhere in California where you can see this problem. They could take a road trip to points east of Reno/Sparks NV because most of I-80 in Nevada is posted 80 mph. Some manager needs to schedule a road trip to show the developers there is an issue.

The only other really annoying issue with my Classic Model S is that the dash display is polarized at the wrong angle. The big screen is correctly polarized at 90 degrees. It is annoying to use polarized sun glasses because you have to tip your head 45 degrees to the left to get full brightness on the display. If you tip your head 45 degrees to the right the dash display goes completely black. How this didn't get caught early on is a complete mystery to me. The first time I got into my model S with polarized sun glasses I noticed it. I have not had a chance to check this on a recent vehicle. It could have been corrected at some point.

I am getting off topic on my own thread. Sorry about that.

Doug Ingraham
2013 Multi Coat Red 85 kWh model S "The Woman in the Red Dress"

 

[kort677]

FWIW: just like when you first start out your Whmiles are reported as quite high the trip planner gets confused, after the first 30 miles force a recalculation by ending the trip and restarting it, you'll get a clearer picture of your usage and trip estimations.

 

[jbcarioca]

Driving at a steady 80 mph in calm winds, OAT of 80F, my P85D averages about 450 WhMi. Thus I plan for an effective range of 145 miles. The Tesla calculator simply cannot make those calculations, so I just monitor my actual usage and change speeds as required.

Although ever so slightly OT, that is exactly why I think tesla should either make major electrical efficiency improvements and/or increase ASAP maximum battery capacity to at least 150 kWh. I expect both of those over the next couple of years.

in the meantime the OP, most Europeans, Australians and many others will have severe range anxiety even though teh slow drivers of the world will be delighted. Tesla has given us huge rises in expectations!

 

[aesculus]

increase ASAP maximum battery capacity to at least 150 kWh.

I think this is the perfect (minimal) size battery based on my use for longer trips.

EM's recent comments at capping battery size to 100 kWh seems to go against this happening.

 

[hiroshiy]

I have driven enough to know that at posted speeds of 70 and below the planner gets it right within just a couple of percent. It is only on higher speeds that it just ignores the speed limit and assumes 65. For it to get it right at 70 but fail so badly at 80 shows that they are computing the additional drag losses that occur both with rolling resistance and air drag and then use the wrong speed in the calculation. If you are going to blow it you should do so on the side of caution. Underestimating charge is the wrong way to go. I would much rather spend 15 additional minutes to get the charge I need at a super charger than white nuckle it for an additional 25 worried I am not going to make it being forced to slow down to below 65. This is pretty much the only time I have had range anxiety. It especially annoys me that it is so good the rest of the time.

The equations to perform these computations are fairly simple. Tesla appears take into account everything but temperature, wind, road surface conditions, and variations in mass of the vehicle due to loading. Vehicle mass does make a small difference in the rolling resistance and affects energy usage when climbing hills and the amount of energy recovered by regen when descending. But climbing and descending tend to cancel out mostly. And the number they use appears to be for two people with quite a bit of luggage. I have traveled alone and I pack lightly so this tends to help me out. (I see the % of remaining charge estimate go up when climbing hills and go down when descending.)

It would be interesting to know if the Auto Pilot feature that shows you the speed limit on the dash display knows the correct speed on these 80 mph and greater posted stretches of interstate highway. If it does then the left hand does not know what the right hand is doing and the autopilot database needs to be consolidated with the navigation database.

These cars are so good that when you do find a problem it amplifies the severity. This issue is hugely annoying to me especially when I know I could fix it in less than a week and it would only take that long because I am not at all familiar with their code. And there is nowhere in California where you can see this problem. They could take a road trip to points east of Reno/Sparks NV because most of I-80 in Nevada is posted 80 mph. Some manager needs to schedule a road trip to show the developers there is an issue.

The only other really annoying issue with my Classic Model S is that the dash display is polarized at the wrong angle. The big screen is correctly polarized at 90 degrees. It is annoying to use polarized sun glasses because you have to tip your head 45 degrees to the left to get full brightness on the display. If you tip your head 45 degrees to the right the dash display goes completely black. How this didn't get caught early on is a complete mystery to me. The first time I got into my model S with polarized sun glasses I noticed it. I have not had a chance to check this on a recent vehicle. It could have been corrected at some point.

I am getting off topic on my own thread. Sorry about that.

Doug Ingraham
2013 Multi Coat Red 85 kWh model S "The Woman in the Red Dress"

Hmm, if it calculates air drag and stuff correctly up to 70mph but not 80mph, I guess you can ask EVTripPlanner to fix that? Not sure whether he has some idea behind picking the wrong speed, or it could be a bug.