If the official EPA range is....

[SeaGeo]

Te problem is that, apparently it is not really consistent between manufacturers. There is a minimum two test series and then there are three additional tests. Tesla does all five as does, I believe, Audi. The full set tends to produce larger values than the minimum testing. The other problem is that people don't seem to understand that the EPA range is a synthetic number intended to be representative of a mix of typical driving conditions and that actual range will deffier dramatically depending on how much one deviates nor do they seem to appreciate that the deviations will not be the same. Sometjhing like the R1T or CT are going to deviate more than the MX with respect to speed because the latter has less frontal area and lower drag coefficient.

The Etron numbers are believable but the Taycan numbers are not. How can one experience less drag at highway speeds than in a mix of low speed highway and in town driving?

As I have driven only Tesla's I can say that based on a sample size of 2, the Tesla EPA range number serves as a good basis for prediction of realized performance under actuall driving conditions. I cannot say the same for any other car(s).

Porsche applies a deduction to their range, and has generally done the test with their worst performing tires. Literally every test people have run at highway speeds massively exceed the rate range. The Mach e tends to as well, but not nearly as much.

Sponsored

 

[ajdelange]

While the EPA test might not be meant to indicate highway range accurately, it is highway range when driving on long trips in real-world conditions that is of most concern to EV owners. Range is far less critical for local driving for most EV owners, especially when home charging is used.

Correct. Learn how to interpret EPA range properly and you will be able to use it to predict highway range quite accurately.

In repeated testing, Edmunds has found Teslas chronically to fall shorter of their range estimates in highway driving than other brands:

https://www.edmunds.com/car-news/testing-teslas-range-anxiety.html
[/QUOTE]I skimmed it fast - it's not worth reading carefully. Tesla tests according to the protocol and publishes a range based on that. Edmunds argues yes, but people don't drive according to the protocol. Duh!

I'm clearly the vox clamantis here. I don't think I want to be that anymore.

 

[Hmp10]

Learn how to interpret EPA range properly and you will be able to use it to predict highway range quite accurately.

No need to bother. After six years of driving my Tesla, I have figured it out. On the most frequent road trip I make (from Naples to Miami, FL) I get about 83% of the EPA-rated range, with about 80% of the trip on open interstate and 20% in Miami traffic congestion.

 

[irish26]

As it is unlikely that the hardware or software in your car(s) is different from mine this is just further evidence that you don't understand what it does and what it is telling you. It is worth making the effort to get to the point where you do as the information it provides is incredibly useful both in terms of planning, making en-route decisons (we've go enough to make it to xxx - the restaurants are better there....) and in reassuring driver and passengers that you will not run out of juice.

Maybe I've just been lucky.

I expect the Rivian system to be as good and will be most disappointed if it isn't.

I understand exactly what it’s telling me and how to interpret that data. That doesn’t change the fact that it is consistently wrong. If I start with 200 miles tomorrow morning on the estimator I will never reach 200 miles in normal driving conditions.

I just went and checked my trip odometer. I’ve gone 70 miles since my last charge Saturday morning and I have 78 miles remaining. I always charge to 202 miles (80%). So we’re on pace for 148 miles for this current charge out of an original estimate of 202.

 

[ajdelange]

No need to bother. After six years of driving my Tesla, I have figured it out. On the most frequent road trip I make (from Naples to Miami, FL) I get about 83% of the EPA-rated range, with about 80% of the trip on open interstate and 20% in Miami traffic congestion.

Well there it is. You have learned that the range of your vehicle in this particular mix of city and freeway is 83% of the EPA range. Note that is is very close to the average experienced by S drivers probaby because most of their driving patterns are similar to yours.

More important for planning purposes is to recognize that what this means is that highway consumption is 20% higher that rated consumption. This is, of course, to be expected. If your EPA range is 300 miles and you intend to go 100 miles you will use nominally 33% of your battery but you have learned that this trip does not represent nominal conditions and that you will use 20% more i.e. 40%. Based on this you can calculate your remaining battery at destination quite accurately. You would also have learned by now that if it rains you are going to have even higher consumption say another 20% so that your trip will consume 47% of your battery. But of course you don't have to do any of this because the energy monitor in the car does it for you to the extent it can. It can't predict the rainstorm and it can't predict a headwind but barring events like that it is amazingly accurate. It works on the rated consumption. If the rated consumption were wrong then it wouldn't work. I conclude, therefore, that Tesla's rated consumption numbers are darn good. Why do you think they aren't? Have you found the energy monitor to be inaccurate?

Interestingly enough I just read a piece last night on how FUD is spread and survives. It used the recent Texas accident/fire in which the cop told the press that there was no one in the drivers seat at the time. It was later proven and NTSB agreed that this was impossible but the world thinks there was no one in the drivers seat and can't be convinced otherwise. Same here. Everyone thinks that Teslas EPA numbers are bad (usually because Tesla cheats) and no amount of reason or data or analysis is going to convince them otherwise.

 

[ajdelange]

I understand exactly what it’s telling me and how to interpret that data.

It's going to seem that I am picking on you and I'm sorry for that but the rest of your post makes it clear that you don't.

That doesn’t change the fact that it is consistently wrong. If I start with 200 miles tomorrow morning on the estimator I will never reach 200 miles in normal driving conditions.

My goal is to help people understand so let's see if we can explain why this isn't wrong. For starters the number you see next to the battery symbol on the display is NOT an estimate of range. It is the SoC of your battery converted to miles by multiplying SoC by the batteries usable capacity and then by the rated consumption for your car. For example if my battery meter reads 50% and I converted the display to miles it would read 0.5*98.5/.282 = 175 miles because the battery's usable capacity is 98.5 kW and the rated consumption for my car is 282 Wh/mi. This number is often referred to as the "rated miles" because it represents how far the car would go if it were always driven at the rated consumption rate which it seldom is. I tend to deprecate the use of the rated miles display because it is misleading to many as it has clearly misled you. It is much more valuable to know what percent of the usable capacity is available to you as you can accurately predict range from that by multiplying by the consumption as modified for your driving conditions. I have learned that around town consumption is about 275 (less that the rated consumption for my car) so half a battery means that I can travel 0.5*98.5/.272 = 181 mi whereas on the interstate i can expect 300 Wh/mi for 0.5*98.5/.3 = 164. This is great for planning but, of course, in the car I don't have to do these calculations as the car does them for me and, in fact, I don't even have to do them for planning as ABRP does them for me

I just went and checked my trip odometer. I’ve gone 70 miles since my last charge Saturday morning and I have 78 miles remaining. I always charge to 202 miles (80%). So we’re on pace for 148 miles for this current charge out of an original estimate of 202.

You start with 202 rated miles and this morning you have 78 remaining so you used
202 - 78 = 124 rated miles over the weekend. You drove 70 miles and assuming you drove them as does the average S driver that would have taken 84 rated miles out of your battery. Taking those off the 124 you used that leaves 124 - 70 = 54 rated miles that went some where else. That's an awful lot for phantom drain which should have taken maybe 15 leaving 39 unaccounted for. Perhaps you left the air conditioning on or the car cooled the battery. In any case your confusion seems to stem from conflating the "rated mile" which is a proxy for energy and 5280 feet which is a distance. The number of rated miles it takes to go 5280 feet depends on how you drive (1 rated mile per mile around town and 1.2 on the freeway are typical) and rated miles are consumed even when the car is standing still.

It's clear you are losing energy when the car isn't being driven. You need to figure out where that is going. Some for phantom drain is normal and the rest probably is too but if I had that much phantom drain and didn't know why I'd be worried.

So perhaps the most important point for you is that the miles number next to the battery display is not an estimate of range. It is telling you how much energy is available in the battery in the proxy unit of rated miles. This energy can be used to move the car or play the radio or heat it in winter of cool it in summer. As noted above I deprecate the use of rated miles in this way because it confuses folks such as yourself. I suggest changing to the % display after which you can read that meter just as you would the gas gauge in your ICE car. Just as you had a fairly good idea about how far you could go on 1/4 tank of gas you will develop a sense of how far you can go on 1/4 of the battery. Of course you can judge how full the battery is by the length of the illuminated bar if the associated number is the rated miles number too but that, I think, will pull you back into thinking that the number is a range estimate.

Range estimates are available but they are on the energy display screens. As I have noted in earlier posts they are quite accurate based as they are on Tesla's superior EPA testing.

 

[guernsej]

Here's what I get when I recognize that F(0) has to be 0.
a =0 ± 0
b =1.5354 ± 0.117
c =0.01011 ± 0.00159

Clearly these are much better quality and, as the residuals only change a tiny bit it's clear that the improvement comes from the DOP. We can argue as to whether I really have the right to do this but if I don't the available data is not really good enough for me to draw any conclusions from.

You're estimating the coefficients of an existing quadratic equation based on visual analysis of its graph using an imprecise plot of 6 points. You know it's first coefficient should be at or near zero - of course setting a = 0 helped!

That doesn't make it a necessary assumption to fit the original data, nor does is imply the engineers estimating road load have had it wrong this whole time doing otherwise.

 

[ajdelange]

Porsche applies a deduction to their range, and has generally done the test with their worst performing tires. Literally every test people have run at highway speeds massively exceed the rate range. The Mach e tends to as well, but not nearly as much.

You list your profession as engineer. As an engineer you should understand why this is the case and be perfectly comfortable with it.

 

[ajdelange]

You're estimating the coefficients of an existing quadratic equation based on visual analysis of its graph using an imprecise plot of 6 points. You know it's first coefficient should be at or near zero - of course setting a = 0 helped!

Actually no, I don't know that. I know that I have a curve that came from a pop car mag. I assume they got the curve from some source but I don't know what that source is. I assume that that source took 3 coefficients from the manufacturer and plotted a polynomial evaluated using those coefficients. Given that a source I trust, Engineering Explained, cites Tesla's polynomial as having a ≠ 0 I probably should not make the assumption that a = 0. But if I don't the consequences are seen in the tolerances in the estimates.

a =23.345 ± 21
b =0.79463 ± 0.673
c =0.015647 ± 0.00515

a =0 ± 0
b =1.5354 ± 0.117
c =0.01011 ± 0.00159

That doesn't make it a necessary assumption to fit the original data, nor does is imply the engineers estimating road load have had it wrong this whole time doing otherwise.

I have to ask: are you familiar with geometric dilution of precision? Perhaps you know it by another name?

What I am doing here is trying to estimate drag from the road load curve as represented in this magazine article. I don't have the data that the engineers used to derive the coefficients from which this curve was probably derived. I don't even have the curve. I must eyeball data from it. I must assume the the curve is quadratic but in fact I don't really know that. I do know that a quadratic fits the points I can eyeball pretty well and I know if I allow 3 degrees of freedom (estimate a) that the CV in the estimate of c is about 33% but that if I use what the physics tells me (set a = 0) it drops to less than half that. I'm inclined to do the latter but I recognize that it may not be justified. The load may be quadratic only in some defined subregion and in that subregion a ≠ 0 may be valid.

All of this has nothing to do with the data taken by the Tesla engineers. I would hope that their analytic skills would be such that they would take enough data data taken over a wide enough range of speeds (this is what gets the GDOP matrix down - lots of measurements over a wide range of the independent variable) to get coefficient estimates with low enough error bands but I have worked with lots and lots of engineers and was always amazed at how few thoroughly understood this aspect of data collection and analysis.

 

[Hmp10]

Well there it is. You have learned that the range of your vehicle in this particular mix of city and freeway is 83% of the EPA range.

For all your equations and calculations, I think you're missing the big picture here. As you wrote earlier, there are different methods and vehicle setups for arriving at an EPA range, some within the control of the automaker.

Porsche chooses a method that yields a very conservative range. A Porsche buyer can look at the Monroney sticker and be assured that in real-world driving in all but extreme conditions he will get at least that much range out of the car.

Tesla chooses a method that yields a very optimistic range. A Tesla buyer has to resort to calculations and/or accumulated experience to figure out his actual minimum range in real-world driving, as the posted range will only be attained while jumping on one leg, rubbing his stomach, and patting his head.

The same scenario plays out in other marketing. Porsche publishes the acceleration numbers for the Taycan based on true 0-60 times on normal asphalt. Tesla publishes the acceleration number for the Model S Plaid based on a track prepped with VHT TrackBite and using a one-foot rollout.

It's not about mathematics and who plays the numbers game better. It's about integrity in marketing.

 

[Jarico75]

txtravwill said:
Since that would compare to a Tesla like the Y that says 316 miles rated range but ends up with closer to 240 miles, i.e. way over estimated for regular driving styles.


I have a Y and based on your assumptions you don't. We get very near the state numbers, but it definitely depends on who is driving. My wife gets more, I get less. We have driven from Salt Lake City to Palm springs. Drive was about 40 minutes longer with the Y than an ice vehicle, but we arrived less stressed than when we drove in an ICE vehicle because of the 2 extra stops. This trip was a good test because there are mountain passes with colder weather an flat out desert. We did this at the end of March so it was freezing in the mountains and hot in the desert. We averaged 300 miles on full charge.

 

[Hmp10]

I have a Y . . .

Edmunds testing recently found that the Model Y comes closest to delivering its EPA-rated range in real-world driving of any Tesla to date. So, at least for the non-performance models with which Tesla plays all kinds of numbers games, they're getting more frank.

 

[skyote]

For all your equations and calculations, I think you're missing the big picture here. As you wrote earlier, there are different methods and vehicle setups for arriving at an EPA range, some within the control of the automaker.

Bingo.

 

[ajdelange]

For all your equations and calculations, I think you're missing the big picture here. As you wrote earlier, there are different methods and vehicle setups for arriving at an EPA range, some within the control of the automaker.

Well of course I think you are missing the big picture. If I thought otherwise there wouldn't be much reason to continue this rather interesting discussion. The EPA and SAE have promulgated rules for determining a range that is intended to benchmark vehicle performance. Manufacturers whined that the complete procedure was too expensive so EPA gave them a copout. They don't have to do the whole procedure but their findings will be fudged to reflect this. You can argue that this should not be allowed and that all manufacturers should be forced to follow the full procedure. I'd support that argument.

Porsche chooses a method that yields a very conservative range. A Porsche buyer can look at the Monroney sticker and be assured that in real-world driving in all but extreme conditions he will get at least that much range out of the car.

What you are missing here is that this is not the intent of the SAE rating. It is intendend as a benchmark. The obvious fallacy in your argument is that "all but extreme conditions" are quite different in the the northeast and midatlantic states than they are in the west. If the Monroney sticker was adjusted to reflect the "all but extreme conditions" of one of those regions people in the other would complain that the number was valueless or deceptive or whatever it is you folks like to think.

Tesla chooses a method that yields a very optimistic range.

No. Tesla chooses a method that follows the EPA/SAE recommendation precisely. The answer is not optimistic or pessimistic. It is accurate.

A Tesla buyer has to resort to calculations and/or accumulated experience to figure out his actual minimum range in real-world driving, as the posted range will only be attained while jumping on one leg, rubbing his stomach, and patting his head.

Just as a Porsche driver has to resort to calculations to figure out what his in-town consumption will be (if the Porsche EPA number resembles highway driving at 75 it is clearly inaccurate for in town driving).

The same scenario plays out in other marketing. Porsche publishes the acceleration numbers for the Taycan based on true 0-60 times on normal asphalt. Tesla publishes the acceleration number for the Model S Plaid based on a track prepped with VHT TrackBite and using a one-foot rollout.

It's not about mathematics and who plays the numbers game better. It's about integrity in marketing.

I'd say the guy that follows the mandated protocol has more integrity than the guy that sand bags but that's just me.

 

[mkhuffman]

It may be in some places but on the east coast speed limits are 65 and 55 mph. Some choose to exceed those speed limits by various amounts. Some don't.

FYI - if you drive down the east coast on I-95, you will find many places where the speed limit exceeds 65. In Virginia, we have 70 mph limit highways, so driving 75 mph is pretty average on those roads.

I agree the EPA test for city range is a reasonable benchmark, but we need a real highway range test. As others have posted, highway range is the most important range estimate for BEV owners, because that is when you need to know if you can make it to the next charger. And that is something we should know prior to purchasing a new BEV. Well, at least that is something I want to know before I buy one. Unfortunately, since manufacturers don't tell us that range estimate, we have to wait for independent testers to do it.

Sponsored