Question... if you knew you were going to be tight, why didn't you drive at 65 mph (EPA tests) to achieve EPA rated range, or at 55 mph (Tesla "ideal") in order to get more? Driving 70 mph and then saying it was a "nail biter" seems a little contradictory when you did not take all of the proper steps to maximize your range.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Badly needed changes to Navigation and Energy apps
- Thread starter jeffnorman
- Start date
-
- Tags
- Model S User Interface
Well, I added 90 minutes of charging at 40 amps. The actual trip was 268 so I thought that by adding a 90 minute charge I would be pretty safe to drive "normally" and not have to hypermile. Now that I have seen the energy vs speed vs temp/Ac usage graph, I know my mistake. But the car should have given me that info - that is my complaint. I shouldn't have to dig up the graph on this forum.
Also, I wasn't aware that EPA tests were at 65. My understanding is (and was) that the EPA test resulting in 265 miles was in a variety of conditions including city driving, stop and go, etc. By watching the energy app, I was sure I could beat this because 90% of my route was open road. I didn't think the 65->70 mph would make much difference really. I normally drive 80 in my old ICE, so going 70 (the speed limit most places) seemed slow but I really thoiught that by keeping the energy below rated for most of the trip, I would be okay and I even thought I would end up not needing the 90 minute top of because if I drove carefully (simply avoiding acceleration and speed changes) I thought I would beat the rated. ... not true as it turned out. Keep in mind that I KNOW I can get a lot more; I got energy consumption down to 215 Wh/mile for the last 50 miles when I saw it was going to be a nailbiter. But I didn't fully realize this until I was well past my top off charge location.
- - - Updated - - -
I really don't know. For some reason, my car did not update firmware until last night?! I finally got the .5 update last night so maybe it was a glitch? I do know that running a 30 mile average I ALWAYS stayed below rated at least when I looked. I drove using the 5 mile average most of the time, checking the 30 mile average on occasion. On the 5 mile average, it would go slightly above rated while in the city, stop and go, at toll booths (after accelerating back up to speed), and in traffic. I pretty much ignore instant because as you said it was useless noise.
What was especially strange was that the "projected miles" on the energy app were always much greater than what I achieved in real life. In other words, at one point (coming in to Chicago when I was driving slow and really watching energy usage), my projected miles were at 95, and my rated miles were 84. I had 85 miles to go and was thinking I was doing pretty good. My 5 mile average was 250 at that point and I kept it low, and even hit 215 for the last 15-20 miles. But guess what? When I pulled into the garage I was at zero. So the projected miles seems like fiction. Something was not being accounting for but I have no idea what. Would love to have the source code so I can understand what the app programmers did wrong.
I will test out the new firmware and see if I notice the same issue next weekend.
Have you tried the following site: http://evtripper.com/planner/carview_beta/
I used for my recent weekend drive and it predicted my energy usage very closely. I'm definitely going to use it from now on, and it's what the Nav system in the car should be. I'd be interested in how closely it comes to the energy usage you observed. I do find I need to use advanced and put in a 1.1 or 1.2 speed multiplier, but that's because I don't stick to the posted speed limit.
I used for my recent weekend drive and it predicted my energy usage very closely. I'm definitely going to use it from now on, and it's what the Nav system in the car should be. I'd be interested in how closely it comes to the energy usage you observed. I do find I need to use advanced and put in a 1.1 or 1.2 speed multiplier, but that's because I don't stick to the posted speed limit.
Agreed there are some improvements that can be made in the nav and energy calculators on the S, that being said this post should be required reading for all new MS owners. How to plan a road trip - how long will it take?
I have successfully completed several multiday road trips without any range anxiety, but planning and understanding the cars limits are very important.
Good luck on your next adventure.
I have successfully completed several multiday road trips without any range anxiety, but planning and understanding the cars limits are very important.
Good luck on your next adventure.
Another technique I've applied recently is to reset one of the counters at the start of the trip (I keep the other as my lifetime counter), and then show the counter on the dash. I can see the average "burn rate" for my trip to the current point, and compare that to the 5, 15 and 30 mile averages as well as the miles remaining on the Nav.
I usually start conservatively with the cruise control set at 65 or so, then as I get more comfortable with my average burn I usually move it up above 70 to 72 or 75. On a recent trip from Boston to Milford via a stop in Westchester for a meeting, I drove 237 miles at 299 wh/mi, arriving with 18 rated miles remaining.
I get the impression this was roughly his trip:
http://binged.it/14fbS6S
Bing says 282.mi so it's probably not that far off from his actual route.
He reported 360-375 wh/mile at highway speed with cruise control.
He reported "Just finished" so the weather can be swagged as "last 4 days" from any weather service that exposes historical climate conditions.
Putting all the together you could probably come up with a decent model for expected efficiency and see if you get similar numbers.
I'm too lazy to do that, but just throwing out the idea for consideration.
- - - Updated - - -
1. Chicago
2. Detroit
3. Tesla P85S (21inch tires)
4. Click update
The bottom two-thirds of the narrative (i.e. everything past the first 36.3 miles) on the right shows 364-366 Wh/mi.
Sounds like the OP was right on track with this beta.
Olle
Active Member
I am driving a p85 with 21s and got sligtly above rated range last 184 mile round trip from Manhattan to Minnewaska state park with 4 adults and 2 kids + heavy luggage in the car. 1500 feet elevation gain. Followed posted speed limits. (35-65, mostly bet 50-65).
Had 80 miles of charge left when returned. Also noticed that I had 7 miles over rated range displayed after max charge, so the battery is probably slightly bigger than the TM claimed 85kWh.
Jeff, when I hear your experience from driving the same car I realize there must be something wrong with your car.
Had 80 miles of charge left when returned. Also noticed that I had 7 miles over rated range displayed after max charge, so the battery is probably slightly bigger than the TM claimed 85kWh.
Jeff, when I hear your experience from driving the same car I realize there must be something wrong with your car.
The evtripper is awesome. It is exactly what I needed but did not have; i have it bookmarked on the Tesla now. My sincere appreciate and thanks to whomever wrote and provided that.
Actual numbers from my trip (including the slight detour for charging) [FOR CLARITY, by "actual" I mean "as predicted by evplanner based on my actual route, actual weather, actual load, 1x the speed limit, and my 21" tires; I did not collect enough log data to figure out my actual for comparison but it had to be pretty close to the predicted results]
Outside temp: 75 degrees
Headwind: 1 mile SSE (disregarded because the east vector of this would be minimal)
Load: 500 pounds (3 people, plus lots of equipment for kayaking and camping)
Speed multiplier 1x (I kept strictly to speed limit)
Car: P85S with 21" tires
Cabin Temp: 70
Actual Route Livonia, MI to Chicago, IL via Dowagiac, MI: 273.1
Results:
Energy: 95.6 Kwh
Avg Wh/mile: 350
Avg speed: 61
Up feet (elevation - hills): 6107 (!!! I NEVER noticed hills before driving this route in my Tesla!! Now I know to check this out; it is a huge factor I think)
Down feet: 5950
My "fill up" of 90 minutes with a 40A HPWC (single charger) added at most 15Kw (1.5 hours x 10Kw/hour max charge rate)... so in theory I had 100KwH and should have ended with 5KwH instead of zero.. but I suspect I did actually end up with something like 5KwH even though the speedo said zero .. because the car was still working, I did not get the "charge now" warning, and based on other posts it seems likely that there is close to a 5KwH buffer when the speedo says zero miles remaining. There are a couple of factors that introduce error that could account for some of the 5KwH difference too: (a) I had to charge at a public station and drove my car several miles in stop and go traffic to get back to my parent's house the night before, so I probably started with less than max charge (I wish I had looked that morning but I didn't); and (b) we made a 20 minute pit stop along the way to get coffee. I might have charged for slightly less than 90 minutes, and the HPWC probably put out slightly less than max of 10Kw (it read 28 miles per hour during the charge time whereas I think 31 is max). Thoughts?
The 350 is probably almost exactly what I experienced, though it is hard to calculate because I did not keep good logs. I know I was at at least 360-370 for the first half of the trip (arriving in Dowagiac, I was down to 70 rated miles whereas I thought I should have had 100+). I kept to rated pretty well after that, and got my consumption down to 215-250 Kwh/mile once I got into Chicago... but that probably didn't matter much because it was the last few miles.
Some interesting observations after running this simulation and trying some variations:
1) My "detour" through Dowagiac gave actually saved me 3.2Kw even though it added 8 miles to my trip (as above post notes, running "straight through" Livonia to Chicago the evtripper calculates 99Kw expenditure or 372 Wh/mile! This is exactly why I think it is critical to integrate the Energy App, something like the evtripper beta, and the Nav app... Had I followed the Nav app "planned" route, I would have been 3.2Kw shorter and probably would not have made it home even if I stopped to charge at a J1772 for several hours! Even without my "top off" in Dowagiac taking the county road detour save me 3.2Kw! But there is NO WAY in the UI of the car to determine a "best route" that will result in the least expenditure of Kw's (versus shortest route or so-called fastest route). If you look at the map, though, it really is not obvious that the detour saves any Kw's.
2) If you look at the segments along M14 (which runs through Ann Arbor and is especially hilly) my average Kwh/mile was predicted at 369 - 384 .. and the segment of I-94 up to the exit for Dowagiac average predicted Kwh/mile was 384. That accounts for the assymmetry that really concerned me (the fact that I was so short when I reached Dowagiac - almost all of my deficit versus the "rated" came from this part of the trip I think). I think that this "power suckage" was mostly caused by the hills - lots of up and down. It is an interesting failure of the nav app that elevations are not even displayed but they have such a big impact on Kw usage.
3) Slowing to 85% of the posted speed limits would have allowed to me make the entire trip without stopping in Dowagiac to top off. This would have increased the trip time from a little over 4 hours to 5.25 hours, but eliminating the charging stop would have given me back that time... though I would not have had the chance to meet S.C. and share a few cups of coffee at his place. All in all, I am glad I made that stop but it is interesting to note that such great savings can be had by slowing down. But back to my original point, which was that slowing down to 85% of the posted speed limit leaves you open to lots of road rage and I really think it is bad for Tesla's image to be consistently passed by Prius's. Most of us (I think) bought the Tesla at least in part because we don't want to have to drive slow and hypermile everywhere we go.
4) Changing to 19" tires would have given me an extra 5 Kwh (90 Kwh for the trip, avg Kwh/mile 331)!
5) Turning the AC to match outside temps (or presumably turning off Ac) would have given me back 1Kw.
6) Traveling on a non-obvious route (Livonia -> Romulous MI -> Cambridge MI -> Coldwater MI -> South Bend IN -> Chicago, IL would have allowed me to make the trip with a total expenditure of only 92.8 Kwh! but still achieve average speeds of 54 mph for a 284 mile distance driving posted speed limits all the way, and not changing any other factors (versus 273 miles at average speed of 61 and energy expended of 99Kwh for the straight through route recommended by the Tesla Nav app). By taking this slightly longer but not much slower alternate route, changing to 19" tires, and turning my AC to match outside temp of 75, I probably could have made the trip without stopping to charge (estimated 86.1 Kwh avg 303 Kwh/mile)... A very brief stop to charge at a public J1772 would have done the trick.
All of this reinforces my belief that the Nav/Energy apps need major upgrading to be more useful and not misleading. At the average expected expenditure for the hilly part of the trip of around 380 Kwh/mile, I would have had only 214 miles of range. But my "projected range" on the energy app never said so except when we were actually going uphill. Without my detour through Dowagiac, I would have only had 228 miles of range; and even with the detour I only had about 240 miles of range. That is still pretty darn good so don't take my criticism as a knock on the Tesla itself, just the UI.
Yes I am new to EVs but I really don't think the issue is my driving but rather lack of information (and now that I know, need for much more detailed planning). My point is, there is just no way for a human being to know the actual range to expect on an unfamiliar route, or the shortest energy path between two points on that route, without running a tool like evtripper. It might be obvious to a hypermiler that it is better to go around a mountain at a cost of adding extra miles, versus the shortest mile path which might be over the mountain .. but when you are traveling over hills that you barely notice when driving, a computer is needed. Same for alternate routes that add miles but at lower speeds. Or travelling in the morning or late night when the weather is cooler versus mid-day. Or turning off your AC entirely and opening the windows. It is easy to know that all these things affect range, but by how much? I like to know exactly how many miles I have left. Rules of thumb and even intuition from experienced EV drivers won't provide that level of info if the route isn't familiar.
From now on, I will use evtripper every time I road trip. But what if I hit construction? Or decide to take a detour to sightsee? Etc. The nav system in our 100K car should be able to handle the task.
Personally, if Tesla would provide the APIs and the ability to load my app on the car, I would be happy to rewrite the code and donate it to the community. Rewriting the navigation app to calculate the least energy expenditure route would NOT be difficult. I was a programmer in a former life, and when writing navigation applications you break the geography into segments .. assigning "weights" to each segment based on miles divided by traffic speed (actual or posted depending on available data) and then do a search using one of many well known algorithms to determine shortest path based on the weights. "Fixing" the nav app would simply involve changing the assigned weights on each segment. Rather than using traffic speed, you would use Kilowatts for each segment (which would be determined by traffic speed, temp, elevation up/down, load, wind, etc.). Once the weights are calculated for the area, the shortest path algorithm would not need to be modified at all - its determination of shortest path would be the least Kw usage based on the calculated segment weights. This would take a programmer knowledgeable with the Nav and Energy apps source code a couple of days, max. It would really be great if Tesla opened up the source code for the Nav and Energy up so we could play with it if they don't have the resources to do it.
It is hard to see because i had to shrink the image to get a good screen shot, but I have attached the route planner detail...you will definitely need to zoom in to see numbers though
Actual numbers from my trip (including the slight detour for charging) [FOR CLARITY, by "actual" I mean "as predicted by evplanner based on my actual route, actual weather, actual load, 1x the speed limit, and my 21" tires; I did not collect enough log data to figure out my actual for comparison but it had to be pretty close to the predicted results]
Outside temp: 75 degrees
Headwind: 1 mile SSE (disregarded because the east vector of this would be minimal)
Load: 500 pounds (3 people, plus lots of equipment for kayaking and camping)
Speed multiplier 1x (I kept strictly to speed limit)
Car: P85S with 21" tires
Cabin Temp: 70
Actual Route Livonia, MI to Chicago, IL via Dowagiac, MI: 273.1
Results:
Energy: 95.6 Kwh
Avg Wh/mile: 350
Avg speed: 61
Up feet (elevation - hills): 6107 (!!! I NEVER noticed hills before driving this route in my Tesla!! Now I know to check this out; it is a huge factor I think)
Down feet: 5950
My "fill up" of 90 minutes with a 40A HPWC (single charger) added at most 15Kw (1.5 hours x 10Kw/hour max charge rate)... so in theory I had 100KwH and should have ended with 5KwH instead of zero.. but I suspect I did actually end up with something like 5KwH even though the speedo said zero .. because the car was still working, I did not get the "charge now" warning, and based on other posts it seems likely that there is close to a 5KwH buffer when the speedo says zero miles remaining. There are a couple of factors that introduce error that could account for some of the 5KwH difference too: (a) I had to charge at a public station and drove my car several miles in stop and go traffic to get back to my parent's house the night before, so I probably started with less than max charge (I wish I had looked that morning but I didn't); and (b) we made a 20 minute pit stop along the way to get coffee. I might have charged for slightly less than 90 minutes, and the HPWC probably put out slightly less than max of 10Kw (it read 28 miles per hour during the charge time whereas I think 31 is max). Thoughts?
The 350 is probably almost exactly what I experienced, though it is hard to calculate because I did not keep good logs. I know I was at at least 360-370 for the first half of the trip (arriving in Dowagiac, I was down to 70 rated miles whereas I thought I should have had 100+). I kept to rated pretty well after that, and got my consumption down to 215-250 Kwh/mile once I got into Chicago... but that probably didn't matter much because it was the last few miles.
Some interesting observations after running this simulation and trying some variations:
1) My "detour" through Dowagiac gave actually saved me 3.2Kw even though it added 8 miles to my trip (as above post notes, running "straight through" Livonia to Chicago the evtripper calculates 99Kw expenditure or 372 Wh/mile! This is exactly why I think it is critical to integrate the Energy App, something like the evtripper beta, and the Nav app... Had I followed the Nav app "planned" route, I would have been 3.2Kw shorter and probably would not have made it home even if I stopped to charge at a J1772 for several hours! Even without my "top off" in Dowagiac taking the county road detour save me 3.2Kw! But there is NO WAY in the UI of the car to determine a "best route" that will result in the least expenditure of Kw's (versus shortest route or so-called fastest route). If you look at the map, though, it really is not obvious that the detour saves any Kw's.
2) If you look at the segments along M14 (which runs through Ann Arbor and is especially hilly) my average Kwh/mile was predicted at 369 - 384 .. and the segment of I-94 up to the exit for Dowagiac average predicted Kwh/mile was 384. That accounts for the assymmetry that really concerned me (the fact that I was so short when I reached Dowagiac - almost all of my deficit versus the "rated" came from this part of the trip I think). I think that this "power suckage" was mostly caused by the hills - lots of up and down. It is an interesting failure of the nav app that elevations are not even displayed but they have such a big impact on Kw usage.
3) Slowing to 85% of the posted speed limits would have allowed to me make the entire trip without stopping in Dowagiac to top off. This would have increased the trip time from a little over 4 hours to 5.25 hours, but eliminating the charging stop would have given me back that time... though I would not have had the chance to meet S.C. and share a few cups of coffee at his place. All in all, I am glad I made that stop but it is interesting to note that such great savings can be had by slowing down. But back to my original point, which was that slowing down to 85% of the posted speed limit leaves you open to lots of road rage and I really think it is bad for Tesla's image to be consistently passed by Prius's. Most of us (I think) bought the Tesla at least in part because we don't want to have to drive slow and hypermile everywhere we go.
4) Changing to 19" tires would have given me an extra 5 Kwh (90 Kwh for the trip, avg Kwh/mile 331)!
5) Turning the AC to match outside temps (or presumably turning off Ac) would have given me back 1Kw.
6) Traveling on a non-obvious route (Livonia -> Romulous MI -> Cambridge MI -> Coldwater MI -> South Bend IN -> Chicago, IL would have allowed me to make the trip with a total expenditure of only 92.8 Kwh! but still achieve average speeds of 54 mph for a 284 mile distance driving posted speed limits all the way, and not changing any other factors (versus 273 miles at average speed of 61 and energy expended of 99Kwh for the straight through route recommended by the Tesla Nav app). By taking this slightly longer but not much slower alternate route, changing to 19" tires, and turning my AC to match outside temp of 75, I probably could have made the trip without stopping to charge (estimated 86.1 Kwh avg 303 Kwh/mile)... A very brief stop to charge at a public J1772 would have done the trick.
All of this reinforces my belief that the Nav/Energy apps need major upgrading to be more useful and not misleading. At the average expected expenditure for the hilly part of the trip of around 380 Kwh/mile, I would have had only 214 miles of range. But my "projected range" on the energy app never said so except when we were actually going uphill. Without my detour through Dowagiac, I would have only had 228 miles of range; and even with the detour I only had about 240 miles of range. That is still pretty darn good so don't take my criticism as a knock on the Tesla itself, just the UI.
Yes I am new to EVs but I really don't think the issue is my driving but rather lack of information (and now that I know, need for much more detailed planning). My point is, there is just no way for a human being to know the actual range to expect on an unfamiliar route, or the shortest energy path between two points on that route, without running a tool like evtripper. It might be obvious to a hypermiler that it is better to go around a mountain at a cost of adding extra miles, versus the shortest mile path which might be over the mountain .. but when you are traveling over hills that you barely notice when driving, a computer is needed. Same for alternate routes that add miles but at lower speeds. Or travelling in the morning or late night when the weather is cooler versus mid-day. Or turning off your AC entirely and opening the windows. It is easy to know that all these things affect range, but by how much? I like to know exactly how many miles I have left. Rules of thumb and even intuition from experienced EV drivers won't provide that level of info if the route isn't familiar.
From now on, I will use evtripper every time I road trip. But what if I hit construction? Or decide to take a detour to sightsee? Etc. The nav system in our 100K car should be able to handle the task.
Personally, if Tesla would provide the APIs and the ability to load my app on the car, I would be happy to rewrite the code and donate it to the community. Rewriting the navigation app to calculate the least energy expenditure route would NOT be difficult. I was a programmer in a former life, and when writing navigation applications you break the geography into segments .. assigning "weights" to each segment based on miles divided by traffic speed (actual or posted depending on available data) and then do a search using one of many well known algorithms to determine shortest path based on the weights. "Fixing" the nav app would simply involve changing the assigned weights on each segment. Rather than using traffic speed, you would use Kilowatts for each segment (which would be determined by traffic speed, temp, elevation up/down, load, wind, etc.). Once the weights are calculated for the area, the shortest path algorithm would not need to be modified at all - its determination of shortest path would be the least Kw usage based on the calculated segment weights. This would take a programmer knowledgeable with the Nav and Energy apps source code a couple of days, max. It would really be great if Tesla opened up the source code for the Nav and Energy up so we could play with it if they don't have the resources to do it.
It is hard to see because i had to shrink the image to get a good screen shot, but I have attached the route planner detail...you will definitely need to zoom in to see numbers though
Last edited:
Actually, EV trip planner predicts 321 wh/mile for your round trip, so you did slightly better than it expects, but it predicts 365 wh/mile for Chicago to Detroit. Why that drive would be so much worse I have no idea, but the fact that it is predictable to a computer that it would be but not to us really speaks to the original point that we need guidance to tell us such things.
(Actually it's interesting. EV trip planner shows an average speed of 69 mph on the freeway segment of your NY trip, but you say you followed the 65 mph speed limit. When I tweak the speed factor so it shows exactly 65 it shows that you should have gotten slightly better than rated, exactly what you did.)
While we're on the subject, I think it's probably worth mentioning "while driving" adjustments. This most useful -- and conceptually simple -- strategy for range management while underway is watching the Rated (instrument cluster) vs. Distance remaining (navigation).
Francis and Discoducky might be able to explain it more elegantly, but the basic gist is this...
All the below is in miles:
Prep
1. You're disconnecting from your charging point, about to embark on the next leg of your journey. Set the nav and make note of the Rated range and the Nav distance. From here forward I will call these Rated and Nav. If you have a poor memory (or health/stress makes that effectively the case), write it down on your hand or a piece of paper.
Define some terms
2. Calculate (Rated-Nav) and call it Buffer Absolute.
3. Calculate ((Rated-Nav)/Nav) and call it Buffer Percentage.
Some starting out guidance
4. If Buffer Absolute < 5, you probably want to reconsider your trip unless Nav < 20.
5. If Buffer Absolute is more than 30 then you can theoretically make it if you're (a) not in a desert, (b) not in a blizzard, (c) not on a race track, and (d) not climbing a mountain face. You may not enjoy the journey, and you may have cars hammering their horn behind you but you can make it. Even with 21" wheels.
6. If Buffer Percentage is > 50%, you will likely have a fun ride and make it easily.
Evaluating while driving
7. Now that you've wrangled these numbers a bit you might forget. So chant to yourself the Rated number. "My buffer was 57mi" (or whatever). Now start your journey.
8. Whenever you get mildly bored in traffic, glance down at the nav, then back to the road, then to your instrument cluster. From this you want updated real-time values for Rated and Nav.
9. Calculate your current Buffer Absolute.
10. Compare your Buffer Absolute to your original Buffer Absolute. Compare specifically not generally. "Less" isn't interesting. "How much less" is. "I started at 57 Buffer Absolute and now I'm at 51 Buffer Absolute. I've lost 6."
11. Compare your Nav to your original Nav. "I started with 120 Nav and now I'm at 100 Nav. I've completed 20."
12. With my examples from 10 and 11, I'm feeling comfortable because I've done this a lot. If you're at all uncomfortable or unsure, then follow the next steps. Otherwise, return to step 8 in a few minutes.
13. Calculate your new Buffer Percentage. "51/100 ... 51%" That's a pretty solid buffer you're fine.
14. (Optional) If you happen to remember your original Buffer Percentage you can compare. "Was 57/120 ... 47% and now it's 51%. My current driving pattern is better than required, given my starting buffer. I can relax a bit."
15. Return to step 8 in a few minutes.
Reading the above it sounds really complicated but the short version is:
Calculate how much buffer you have and watch as it decays. If the decay rate is sharper than your nav distance completion (i.e. linear prediction from 'now' suggests you'll go below zero buffer), then adjust your driving to be more conservative.
- - - Updated - - -
Now that I see this written out, I have yet another reason to want an SDK so I can just write a Range Management app for the car and save some brain use. (I have limited cycles remaining, I can already feel it.)
Without an SDK, I could perhaps write a phone app that updates periodically using REST data for the consumption and range part. I'm not sure how to tap into the Nav remaining distance though. I need to think about this some more.
Francis and Discoducky might be able to explain it more elegantly, but the basic gist is this...
All the below is in miles:
Prep
1. You're disconnecting from your charging point, about to embark on the next leg of your journey. Set the nav and make note of the Rated range and the Nav distance. From here forward I will call these Rated and Nav. If you have a poor memory (or health/stress makes that effectively the case), write it down on your hand or a piece of paper.
Define some terms
2. Calculate (Rated-Nav) and call it Buffer Absolute.
3. Calculate ((Rated-Nav)/Nav) and call it Buffer Percentage.
Some starting out guidance
4. If Buffer Absolute < 5, you probably want to reconsider your trip unless Nav < 20.
5. If Buffer Absolute is more than 30 then you can theoretically make it if you're (a) not in a desert, (b) not in a blizzard, (c) not on a race track, and (d) not climbing a mountain face. You may not enjoy the journey, and you may have cars hammering their horn behind you but you can make it. Even with 21" wheels.
6. If Buffer Percentage is > 50%, you will likely have a fun ride and make it easily.
Evaluating while driving
7. Now that you've wrangled these numbers a bit you might forget. So chant to yourself the Rated number. "My buffer was 57mi" (or whatever). Now start your journey.
8. Whenever you get mildly bored in traffic, glance down at the nav, then back to the road, then to your instrument cluster. From this you want updated real-time values for Rated and Nav.
9. Calculate your current Buffer Absolute.
10. Compare your Buffer Absolute to your original Buffer Absolute. Compare specifically not generally. "Less" isn't interesting. "How much less" is. "I started at 57 Buffer Absolute and now I'm at 51 Buffer Absolute. I've lost 6."
11. Compare your Nav to your original Nav. "I started with 120 Nav and now I'm at 100 Nav. I've completed 20."
12. With my examples from 10 and 11, I'm feeling comfortable because I've done this a lot. If you're at all uncomfortable or unsure, then follow the next steps. Otherwise, return to step 8 in a few minutes.
13. Calculate your new Buffer Percentage. "51/100 ... 51%" That's a pretty solid buffer you're fine.
14. (Optional) If you happen to remember your original Buffer Percentage you can compare. "Was 57/120 ... 47% and now it's 51%. My current driving pattern is better than required, given my starting buffer. I can relax a bit."
15. Return to step 8 in a few minutes.
Reading the above it sounds really complicated but the short version is:
Calculate how much buffer you have and watch as it decays. If the decay rate is sharper than your nav distance completion (i.e. linear prediction from 'now' suggests you'll go below zero buffer), then adjust your driving to be more conservative.
- - - Updated - - -
Now that I see this written out, I have yet another reason to want an SDK so I can just write a Range Management app for the car and save some brain use. (I have limited cycles remaining, I can already feel it.)
Without an SDK, I could perhaps write a phone app that updates periodically using REST data for the consumption and range part. I'm not sure how to tap into the Nav remaining distance though. I need to think about this some more.
With a combination of the evtripper algorithm and the "update while driving" algorithm programmed into the Nav system, it seems like the car would be able to give you really accurate guidance as to how many rated miles you could expect to have at your destination and whether you need to slow down to get there. I hope it's only a matter of time before Tesla comes out with something like that.
GlennAlanBerry
Member
I actually enjoy periodically calculating my rated range vs. my remaining distance in my head as I drive, pretty much as brianman describes, but I know that many people can't be bothered to do that. It would be nice to have an application to automate some simple mental math and common sense. Sooner or later, someone will run out of battery and be stranded, and of course it will be the car's fault...
People should be used to making this sort of calculation, since they have to do it for an ICE the exact same way. The major difference is that there are many more gas stations available when people miscalculate, so they don't really have to worry about it. As the EV charging infrastructure improves over the next few years, it will become harder to run out of energy unless you are trying to Broder the car!
It would be nice if the Tesla did have an app that integrated with PlugShare, Recargo, etc., so that you could very easily find the nearest charger with much less effort. I think something like this will be required by Gen III time, as Tesla goes more mass-market.
People should be used to making this sort of calculation, since they have to do it for an ICE the exact same way. The major difference is that there are many more gas stations available when people miscalculate, so they don't really have to worry about it. As the EV charging infrastructure improves over the next few years, it will become harder to run out of energy unless you are trying to Broder the car!
It would be nice if the Tesla did have an app that integrated with PlugShare, Recargo, etc., so that you could very easily find the nearest charger with much less effort. I think something like this will be required by Gen III time, as Tesla goes more mass-market.
That technique works fine as long as you are driving on a level freeway at a constant speed. Unfortunately, when I need to calculate range it's rarely that simple. Quick quiz: If the first 130 miles are at 70 mph on a level freeway, but the last 50 miles you are going 50 mph but gaining 1000 feet in elevation, will your energy usage increase or decrease for the last section? The fact is that I have had my wh/mile vary significantly on different drives due to elevation changes that I otherwise wouldn't have paid any attention to and faster or slower roads make an enormous difference. On my daily commute my average usage is 580 wh/mile for the first 10 miles, and then the average goes down to about 320 wh/mile when I get to work (30 miles). On my way home it goes back to 380 wh/mile and then to about 310 wh/mile when I get home. I still have never figured out why I use more energy going to work than coming home, but it's consistent.
In an ICE you just leave a huge margin for error, and if in doubt fill up before any questionable section. People don't even notice how much their mileage varies under different conditions, which is why they get caught out like the OP in an electric car.
Vger
Active Member
OK, I have posted about this before, but I clearly need to on this thread!
WIND, WIND, WIND!
The OP's return trip was westbound, into the prevailing wind. I have seen this over and over on our many trips across WA State west to east and east to west. Wh/mi can vary by 10-20% EASILY depending on wind direction and wind speed.
It is particularly deceptive in the Model S since the aero is so good, you do not hear and rarely feel it when you are in a head wind.
I would really like to see Tesla put a pitot tube on the car to measure apparent wind (air speed) but I don't think that is going to happen, so...
On long trips that will push range boundaries, especially east-west in North America, notice the wind direction and intensity--watch flags, trees, grass, smoke plumes. If you want to be more analytical and plan. Look at a wind forecasting web site like sailors use, for example: Error.
We have had instance driving westward where we left with a full range charge to do a 200 mile trip that we usually can do with 40-50 miles of rated range to spare, and found ourselves crawling the last 5 miles approaching the dread "0", despite using good "technique"-- smooth accel and decel, never over the speed limit, drafting when safe and easy, etc.
Hills, you can see, the speedo you can see and control, climate you can feel, but the wind can suck the life out of your battery, and you can be totally unaware it is happening.
After more than 25,000 miles of road trips in my Roadster and my Model S, my rule of thumb is: if your average consumption (since your last charge) is running above about 340 Wh/mi or 225 Wh/km, and you cannot account for it by speed, terrain, rain, snow or freezing cold, CHECK THE WIND!
WIND, WIND, WIND!
The OP's return trip was westbound, into the prevailing wind. I have seen this over and over on our many trips across WA State west to east and east to west. Wh/mi can vary by 10-20% EASILY depending on wind direction and wind speed.
It is particularly deceptive in the Model S since the aero is so good, you do not hear and rarely feel it when you are in a head wind.
I would really like to see Tesla put a pitot tube on the car to measure apparent wind (air speed) but I don't think that is going to happen, so...
On long trips that will push range boundaries, especially east-west in North America, notice the wind direction and intensity--watch flags, trees, grass, smoke plumes. If you want to be more analytical and plan. Look at a wind forecasting web site like sailors use, for example: Error.
We have had instance driving westward where we left with a full range charge to do a 200 mile trip that we usually can do with 40-50 miles of rated range to spare, and found ourselves crawling the last 5 miles approaching the dread "0", despite using good "technique"-- smooth accel and decel, never over the speed limit, drafting when safe and easy, etc.
Hills, you can see, the speedo you can see and control, climate you can feel, but the wind can suck the life out of your battery, and you can be totally unaware it is happening.
After more than 25,000 miles of road trips in my Roadster and my Model S, my rule of thumb is: if your average consumption (since your last charge) is running above about 340 Wh/mi or 225 Wh/km, and you cannot account for it by speed, terrain, rain, snow or freezing cold, CHECK THE WIND!
Agreed, EVTripPlanner is great! EVTripPlanner
It's interesting to run different route scenarios with EVTripPlanner and see the effect of headwind or tailwind, speed, outside temperature, hills, etc.
EVTripPlanner needs to be integrated into the Model S nav system.
That is actually one of the reasons I prefer to keep a close eye on the instant Wh meter. If you have a feel for how much energy it takes to keep you at speed on a level area, and your energy usage is higher than that or it fluctuates, you know you are in a headwind.
wcalvin
Member
Tesla ought to be providing us with a trip planner via the 17", the App, or a web page. It ought, like Google Maps, show us three routes. And the wh/mi estimate ought to show brackets for temp/wind variability.
Ideally, Google Maps ought to start using a Model S to do their street views, so there would be a watts/mile in the spreadsheet in the medium speed segments.
Ideally, Google Maps ought to start using a Model S to do their street views, so there would be a watts/mile in the spreadsheet in the medium speed segments.
Beavis
Model S Signature 991
You are absolutely right. Every road biker knows to watch the wind because it makes a huge difference. Having a wind forecast integrated with the nav, along with the other suggestions, would be a nice addition.
Spartha | AUGUST 9, 2013
I got my metallic green Tesla S today.
On the plus side, besides what is written often, is that it synchronizes effortlessly with iPhone (including music). I am also pleasantly surprised by the fairly dark windows (I may not have to tint but I live in Florida!)
I am a little puzzled that the navigation doesn't have route preferences, such as avoid Toll road, avoid freeway etc (In Florida, they are a must!). I am also puzzled that while Google map could identify my address in a new subdivision, Tesla-Google map doesn't. I had to give a fake nearby address in an older neighborhood to get going. The lack of saving "Home" "work" etc is also of my concern.
Could anyone verify that there is no alternate route preference etc? I vaguely remember something about a "short" and "long" route.
A Tesla navigation "simulation" would be great!
I got my metallic green Tesla S today.
On the plus side, besides what is written often, is that it synchronizes effortlessly with iPhone (including music). I am also pleasantly surprised by the fairly dark windows (I may not have to tint but I live in Florida!)
I am a little puzzled that the navigation doesn't have route preferences, such as avoid Toll road, avoid freeway etc (In Florida, they are a must!). I am also puzzled that while Google map could identify my address in a new subdivision, Tesla-Google map doesn't. I had to give a fake nearby address in an older neighborhood to get going. The lack of saving "Home" "work" etc is also of my concern.
Could anyone verify that there is no alternate route preference etc? I vaguely remember something about a "short" and "long" route.
A Tesla navigation "simulation" would be great!
Similar threads
- Question
