[KarenRei]

@KarenRei, dyno tests are performed in a controlled environment. They don't contradict real-world data if you have a sample size that is large enough.

Except you're not simply reporting dyno data. You're trying to make some sort of "dyno-consumer reports-range conversion". Except that consumer reports isn't a drivecycle, isn't even self-consistent, and your "conversion" makes a number of unsupported assumptions.

That's way overcomplicating things. The goal is "what's the real-world consumption?" And the way to do that isn't the above; it's to simply look at real-world consumption.

 

[Troy]

You're trying to make some sort of "dyno-consumer reports-range conversion".

Yes, exactly. I'm fixing everything that is wrong with the EPA rated range. Here are the main problems:

1. EPA rated range combines two dyno scores: city and highway. I don't use the city dyno score. I only use the highway.
2. The default conversion multiplier is 70% when converting dyno scores to EPA rated range. That's what all car manufacturers use by default. See the first bullet point here. Instead, I'm using 65.55% because this is more accurate.

As for using Wh/mi numbers to measure or compare range, how well did that method work so far for Model S versions? For example, the Model S 75D has been around since mid-2016 and we still don't have precise Wh/mi numbers for 65 mph or any specific speed. It's not like people are reporting similar numbers so we can take those and then compare them to other Wh/mi numbers at 65 mph for other Tesla models.

In my previous message, I pointed out to the difference the elevation and wind makes. I found an example in a forum message here. The person is reporting these numbers:

I drove 11 miles at 65mph, 5.5 miles each way to minimize the impact of elevation changes.
Round trip: 286 wh/mi.
First leg: 394 wh/mi
Second leg: 177 wh/mi

As you can see, his consumption was 177 Wh/mi at 65 mph going one way and 394 Wh/mi at 65 mph going the other way. This is a huge difference when you are trying to compare the range of many different Tesla cars precisely. That means when people are reporting you Wh/mi numbers at 65 mph, you are going to get numbers that vary too much. What you need is a consistent test environment. You need to perform the same range test at the same location. That's why dyno scores are valuable because the cars are tested indoors on the same dyno by Tesla.

Right now we have EPA highway dyno scores but soon we will also have the European WLTP range test scores for all current Tesla cars. Tesla has until Sep 2018 to switch to these numbers on their European design studios. WLTP rated range is expected to be lower and more accurate than EPA rated range.

 

[Reef Club]

Hi, everybody. I have created these two tables mostly based on EPA highway dyno test scores.

Range in miles:

Range in km:

Why use EPA highway dyno test scores?
The reason I'm using EPA highway dyno test scores instead of EPA rated range is that EPA rated range is manipulated in two ways: 1. Different multipliers, 2. Voluntary reductions. The manipulation happens after the dyno test is done and before they release the official EPA rated range. More information about these manipulation methods can be found here. Therefore EPA rated range numbers are not comparable but EPA highway dyno test scores are.

Range at 65 mph
To convert EPA highway dyno test scores to range at 65 mph, I have used the range test numbers by Consumer Reports. Here is an example:

The Model S 75D scored 358.49 miles in EPA highway dyno tests (source: page 19 here)
It also scored 235 miles at 65 mph in Consumer Reports' range test (source: video and article)

The Model 3 LR (aka Model 3 75) scored 454.64 miles in EPA highway dyno tests (source: page 7 here)
Consumer Reports has not tested the Model 3 LR yet. We only have the EPA dyno test scores for the Model 3. Therefore I calculated what the range would be if/when Consumer Reports tests the Model 3 LR. The calculation is this: 454.64 * 235/358.49= 298 miles. That's the range at 65 mph.

Other than the Model S 75D, Consumer Reports also tested the range of the Model X 90D. You can read more about it here.

Range at 70, 75, 80 mph
After I had the numbers at 65, to calculate the numbers at 70, 75 and 80 mph, I used the first graph Tesla published here.

Supercharge percentages and 30-min Supercharge rates
I've used this video for the S/X supercharge percentages. For the Model 3 LR, I've used this calculation 170/310= 54.8% in 30 minutes based on the 170 miles in 30 minutes number Tesla published here.

Different wheel options:
There are 3 wheel configurations for the Model 3:

1. 18" wheels without aero covers
2. 18" wheels with aero covers
3. 19" wheels

I think the test was done with #1 because of two reasons:
1. On page 4 here you can see a photo of the car during the test.
2. On page 16 here, it shows these two numbers: 9.95 HP for Model 3 with 18" and 11.13 HP for the Model 3 with 19". The ratio is 9.95/11.13= 89.4%. Now let's look at the Model S RWD numbers. On page 5 here, it shows these numbers: 11.45 HP for the Model S with 19" wheels and 12.78 HP for the Model S with 21" wheels. The ratio is 11.45/12.78= 89.6%. The difference is almost identical and the Model S in this test doesn't have aero covers. Therefore the almost identical difference suggests that neither the Model 3 nor the Model S had aero covers.

To calculate #3, I have used the A, B, C coefficients on the same page (on page 16) here. It shows these numbers:

Model 3 18"
A= 38.51
B= ‐0.0811
C= 0.01610
Road load @50 mph= 9.95 HP

Model 3 19"
A= 42.30
B= ‐0.0212
C= 0.01691
Road load @50 mph= 11.13 HP

The HP numbers at 50 mph are useful but I needed the numbers at 65, 70, 75, 80 mph. On Reddit, some people said the gains with the smaller wheels would diminish at higher speeds. This is correct but the change is very small. The efficiency difference between 18" - aero vs 19" at different speeds is as follows:

• 65 mph 10.97%
• 70 mph 10.70%
• 75 mph 10.44%
• 80 mph 10.19%

To calculate these, I used the A, B, C numbers you see above which allow calculating the road load at any speed. We happen to have the road load at 50 mph. Therefore it is possible to double check whether the calculation is correct. Here is the calculation (50^2 means 50 * 50):

(42.30 * 50 + ‐0.0212 * 50^2 + 0.01691 * 50^3 )/375= 11.135 HP

In this formula, you can change 50 mph to 65 mph, then do the same calculation with the other A, B, C numbers for 18" wheels and then compare the results and you get 10.44% difference at 75 mph.

For #2, I used an estimated 6% improvement because there is no definitive data yet. In 2012 Tesla released aero covers for the Model S and people were reporting 5 to 10% improvement. I figured 6% would be a safe bet.

Degradation:
I added the purple columns to display the range after 5% degradation because 5% looks reasonable based on survey data. If you look at the degradation chart here for miles and here for km, you can see that the range drops to 95% at 45,000 miles or 72,500 km. For the Model 3, the drop to 95% will actually happen at a 25% higher mileage than the Model S because the Model 3 is more efficient and requires fewer charge cycles to achieve the same mileage.

Shortcomings:
The biggest problem I think is the Model S 85 data. The EPA test was done in February 2012 and it was never repeated again. However, Tesla improved the Model S 85 in 2013, 2014 and 2015. Therefore the numbers you see are for the 2012 Model S 85. It shows 219.7 miles range at 65 mph but I think the range was ~15 miles more in later years.

Troy, thank you for this post. I shared with my son. He went over the findings, we agreed that I should get the long range battery with the 18 inch aero wheels. We decided the Midnight Silver looks best with the aero wheels. So I am just waiting for my email so I can get delivery before July 1, 2018 and grab that $7,500 tax credit!

 

[Troy]

I found a message I wrote on Reddit a while ago. I want to copy it here.
---
Here is an example to explain the different range units:

North America:

• EPA rated range: 259 miles for the Model S 75D
• Ideal range: Unrealistically high

Rest of the world:

• NEDC rated range: Unrealistically high
• Typical range: 241 miles for the Model S 75D

Here is what you need to know:

1. Tesla advertises EPA rated range only in North America. If you open the Model S design studio, scroll down and change the country selection to somewhere in Europe, let's say the UK, you will see the NEDC rated range. NEDC rated range is completely unrealistic but it is the official range they use in Europe. NEDC will be discontinued by Sep 2018.
2. Tesla cars display EPA rated range only in North America. Outside of North America, EPA rated range is not a selectable option in the settings menu on the touchscreen. In Europe, the touchscreen has two options: Rated and Typical. The rated range in Europe should not be confused with the rated range in North America even though the touchscreen says "Rated" for both.
3. Typical range was created by Tesla and does not exist outside of Tesla cars. Unfortunately, Tesla does not publish the Typical range numbers anywhere but we do have survey-based data. Currently, Typical range is the most realistic range unit. For example, a new Model S 75D in Europe will display 241 mi Typical range at 100% charge. When Consumer Reports tested the range of the Model S 75D, it scored 235 miles. Therefore Typical range is very close to real-world range. Typical range is always less than EPA rated range. Unfortunately, Typical range is not available in North America.
4. In Europe, the NEDC rated range is being replaced with something new called WLTP rated range. Newly released cars already use WLTP since Sep 2017. Existing cars will switch to WLTP rated range by Sep 2018. WLTP rated range is expected to be lower than EPA rated range and very similar to Typical range.

 

[EV-lution]

@EV-lution,

298 miles is the range at 65 mph. It is not the official rating. I calculated the 298 mi number from EPA highway dyno test scores and Consumer Reports test data at 65 mph. The official EPA ratings are also calculated from EPA dyno tests but there are two dyno tests: city dyno test and highway dyno test. The official number combines these two. The city test has 55% weight and the highway test has 45% weight. Ideally, EPA range should not include any city test because it is the highway range that matters. However, the EPA is concerned about fuel economy. Their test is designed to measure MPG for gasoline cars. They are doing the exact same thing for electric cars. They are measuring MPGe (miles per gallon equivalent).

Another problem with the official number is that the calculation to convert dyno scores to EPA rated range has changed over time. This makes the advertised numbers incomparable. One more problem with the official numbers is voluntary reductions. This also makes the advertised numbers incomparable.

310 miles is the number Tesla picked when they voluntarily lowered the EPA rated range from 334 to 310 miles. It is not based on any tests. 334 miles is the actual EPA score before the voluntary reduction. It is based on city and highway EPA dyno tests. You can see both numbers on the last page of this EPA document.

The important thing you need to know about EPA rated range is that 1 EPA rated mile in one Tesla model is completely different than 1 EPA rated mile in another Tesla model because of voluntary reductions and changes to the calculations over the years. It would be completely wrong to say something like, the Model S 100D has 335 miles EPA rated range, therefore it has more range than the Model 3 LR RWD with 310 miles EPA. That's not the case. They both have the same range.

Similarly, it would be wrong to assume that a 2012 Model S 85 with 265 mi EPA has more range than a 2016 Model S 75 with 249 mi EPA. The Model S 75 has more range. Tesla improved the S85 in 2013 and 2014. Also, it would be completely wrong to assume the Model S 75 with 249 mi EPA has more range than the Model 3 SR RWD with 220 mi EPA. They both have the same range assuming aero covers are on.

In the food industry, there is something called shrinkflation which makes similar looking products incomparable. You have to look at the weight to have comparable data. EPA rated range is like that. The advertised numbers are not comparable but the dyno scores are.

In case people are curious, Model 3's 334 mi EPA rated range was calculated using this formula:
EPA rated range =round(0.7 * (0.55 * city dyno score + 0.45 * highway dyno range))

The dyno scores can be found here. The 0.7 multiplier is the default multiplier. The Model 3 is the first car where Tesla used the default multiplier. They used many multipliers in the past and they were all higher than 0.7. See the list here. I found these multipliers inside an EPA spreadsheet.

Here is a summary for you:

• Don't trust the advertised range numbers of Tesla cars unless it is a Model 3.
• Don't pick the larger wheels because you will lose too much range.
• Don't assume a Model S has more range than a Model 3 just because the advertised range is more.
• Don't assume a Model S 85 has a larger battery than a Model 3 LR because it doesn't. Details here.
• Don't trust the advertised 0-60 numbers. Some of them might be under-advertised. See the example here.

Thanks Troy, I really appreciate your detailed reply and all the calculations you have done to determine range. There are so many factors involved with range, but as long as we have comparable numbers we can make informed decisions. My main concern is getting a M3 with significantly more range than my current Model X 90D with 20" rims.

 

[EV-lution]

Except you're not simply reporting dyno data. You're trying to make some sort of "dyno-consumer reports-range conversion". Except that consumer reports isn't a drivecycle, isn't even self-consistent, and your "conversion" makes a number of unsupported assumptions.

That's way overcomplicating things. The goal is "what's the real-world consumption?" And the way to do that isn't the above; it's to simply look at real-world consumption.

Great idea, would you be willing to post some real-world consumption numbers in a table similar to what Troy presented that takes into account all factors involving range, such as: model type, wheel type, wind, elevation, temperature, speed etc.? The only way to get to the truth is to have a amicable exchange of empirical data. Science!

 

[3V Pilot]

Man, you're goood! @Troy , @garsh nailed it.

Personally value the highly data driven nature of your posts and not sure why @KarenRei takes exception. Short of talking to eachother which would be ideal, I always recommend direct contact via pm to resolve these kinds of situations. I am convinced you both genuinely mean well.

Take care, both of ya. And mostly Happy new T≡SLA year!! :rainbow::rainbow::rainbow:

I also wanted to say that I really do appreciate all the work that you and Karen have put into crunching the numbers of this car. It's all over my head for the most part and won't really matter when your driving into a 60mph headwind, uphill......or a 60mph tailwind downhill for that matter. My point being that real world mileage will always vary, but that is true of any car. I'm actually more curious to know how much the efficiency varies based on driving style once I get the car. I'm hoping that I can drive like madman and not suffer as much loss as I would in an typical ICE car. Any thoughts on that from either of you would be very interesting.

 

[MelindaV]

EPA rated range combines two dyno scores: city and highway

based on the sticker, highway is 95.2% of the combined 126MPGe. So based on this, 295.1 miles for highway only (or 322 miles for city only). these based on Tesla's reduced range #s.

 

[Troy]

Hi, everybody. I have now updated the tables to change the names from Model 3 75 to Model 3 LR.

By the way, there is a column called '30 min supercharge range' and it shows how much range you would add in 30 minutes starting from 0%. These range numbers are not rated range. They are equivalent to 65 mph range. I didn't use rated range because it is not a comparable base unit between models.

At 1:39 in this video, you can see the charge percentages of S/X packs in 30 minutes starting from 0%. The 75, 85, 90 packs Supercharge to 54, 56, 55% in 30 min. For the Model 3, I'm using 54.8% based on 170/310= 54.8%. All 4 percentages are very close. This looks fine to me but I will update it if real-world data shows a different percentage than 54.8% for the Model 3 LR. I multiplied the first orange column by this percentage to get the Supercharge range in 30 minutes.

@MelindaV, you have the correct idea with the 310 * 120/126= 295 mi highway range calculation. This number happens to be very close to the 298 mi number I calculated at 65 mph because the 310 mi number is actually a pretty good representation of combined range for the Model 3 which is unfortunately not the case for other Tesla cars.

What this all tells me is that if you are concerned mostly with range and supercharger charge speed, then the Model 3 LR is actually superior to both the Model S and X.

Yes, exactly. In fact, a Model 3 has already broken the cross-country road trip record (details here). Of course, the Model 3 LRD will improve that record.

 

[KarenRei]

In my previous message, I pointed out to the difference the elevation and wind makes. I found an example in a forum message here. The person is reporting these numbers: ,, As you can see, his consumption was 177 Wh/mi at 65 mph going one way and 394 Wh/mi at 65 mph going the other way. This is a huge difference when you are trying to compare the range of many different Tesla cars precisely.

Said user also drove only 11 miles. Super-short trips will obviously dramatically amplify variability. That's not that's being reported with figures like Maevra's - these are long-term averages.

Furthermore, such "variable" trips will artificially inflate the average. Take, for example, the case of driving in heavy winds. Let's go with a wind-caused 177 Wh/mi / 394 Wh/mi. Does that mean that a wind-free trip would get 285 Wh/mi? Not at all.

1) Crosswinds raise your drag coefficient. The tailwind bonus is for a perfectly aligned tailwind.

2) Aero drag is proportional to relative speed squared. Take for example a perfectly aligned 30mph wind and a car going 60mph. So the relative effective velocities are 30mph and 90mph. Aero drag at 30mph is cut to 1/4 while it's increased at 90 to 2 1/4. The average of 1/4 and 2 1/4 is not 1, it's 1,25; your net effect is a 25% increase in aero drag. So if, say, 60% of your energy loss is aero and 40% is rolling/parasitic, your range is cut by 15%, even though it was a round trip in the same wind conditions.

You have the same problem on significant elevation changes.

3) There's a slight increase in energy consumption regardless due to hill climbing due to I²R heating; climbing means more current while descending means less, which hits the same problem as #2 - if you linearly adjust X by a fixed amount positive and negative for the same trip length, and use X² to compute some figure, you're overall increasing the value of said figure (in the former case, drag; in this case, resistive heating)

4) Obviously, if you have to regen, that wastes part of the energy.

5) Climbing is typically associated with other loss factors, such as cornering losses and more uneven traffic speeds.

On long trips and accumulated data, all of these factors generally blur into an average consumption that takes into account both good and bad conditions. But on little 11-mile trips, you're obviously going to see a lot of distortion depending on which direction you drive.

Note that even aggressive driving is getting good reported figures. E.g., RiggerJon is a self-confessed "hypomiler". His running average - a roughly 50-50 mix of city and highway - 244 Wh/mi. This is from a guy posting pictures of rubber splatters from his tires on his wheel wells

By the way, said user that you quoted getting 177 Wh/mi and 394 Wh/mi on an 11 mile round trip? They were driving a Model X, not a Model 3. Model 3 would have used vastly less energy for said trip.

 

[Troy]

@KarenRei, I'm glad that you are enthusiastic about your Wh/mi based range estimate method. I hope you collect lots of data and create your own tables and then we can compare the tables. Good luck.

 

[Dr. J]

On long trips and accumulated data, all of these factors generally blur into an average consumption that takes into account both good and bad conditions.

And good, bad and indifferent drivers. The averages (mean and median) can be useful, but you're also going to get wide deviations. I don't know if I'm toward the top end or the bottom end of each of those factors.

@Troy is calculating one number (for each configuration at a given speed under one set of conditions) that makes comparing different Tesla models possible. It also provides me a benchmark for mileage and range. If MMVs, I can figure out what's causing that (weather, topography, tires, nut loose on the steering wheel, etc.).

Both ways of going about it are useful.

 

[KarenRei]

And good, bad and indifferent drivers. The averages (mean and median) can be useful, but you're also going to get wide deviations. I don't know if I'm toward the top end or the bottom end of each of those factors.

@Troy is calculating one number (for each configuration at a given speed under one set of conditions) that makes comparing different Tesla models possible. It also provides me a benchmark for mileage and range. If MMVs, I can figure out what's causing that (weather, topography, tires, nut loose on the steering wheel, etc.).

Both ways of going about it are useful.

Except that his numbers are very different from the sort of real-world numbers people are reporting. Which means that he's doing it wrong. And it's wrong because it relies on multiple faulty assumptions. The notion that CR is a steady-state 65 (it's not). The notion that CR is internally consistent between vehicles (it's not, you get very different ranges depending on which "comparison vehicle" you pick, and he happened to pick a particularly pessimistic one). The notion that CR is self consistent (it's not). And a bunch of others.

When your theory is contradicted by reality, your theory is wrong.

 

[Michael Russo]

@KarenRei , look, can we just agree that anyone getting the Model 3 LRB, as observed simply by @Dan Detweiler and myself earlier in this thread, will most likely get one of the longest range T≡SLA vehicles currently made in Fremont, with maybe just the exception of the S 100D...?

And leave it at that... :tmi:

In the end, I find this the most encouraging news of all (am even past the HUD now!! ) and one that could make me stretch my budget by 10-20% to make sure I get Midnight S≡R≡NITY to be right for me for the next decade!!

Thank you, T≡SLA!

 

[Reef Club]

Troy no good deed goes unpunished! We can all agree the 18" aero wheels will increase the range of the Model 3. So instead of paying $1,500 up front for 19" wheels, I say pocket the money and if the 18" aero wheels do not produce a nice ride and/or sufficient increase in range, we can use the $1,500 to pay toward aftermarket 19" wheels and tires. Save the 18" aeros or sell on Ebay. It is a win win.

 

[KarenRei]

Just to add to this, for real-world data, you can always consult the spreadsheet (and definitely add more as you find more!). Ranges are based on the EPA's charge-depleting capacity of 78,2 kWh (note: this is the point at which the car stops moving, not the point at which it reads 0). In the brackets I'll put the conversions to SR range assuming a battery capacity ratio of 31/46 and a 2% range bonus for the reduced weight.

LR 18" w/caps:
* 80mph: 240 Wh/mi (326 mi) [224 mi]
* 60-80mph, mainly 70-75: 241 Wh/mi (325 mi) [223 mi]

LR 19"
* Mixed city/highway "hypomiling": 244 Wh/mi (321 mi) [220 mi]
* Mixed city/highway: 246 Wh/mi (318 mi) [219 mi]
* 70mph: 290 Wh/mi, 58°F (289 mi) [185 mi]
* 80mph: 280 Wh/mi (270 mi) [192 mi]

Expect the weather, terrain, and traffic on these to be for the most part "Southern California in the winter" - cool, but not freezing, a mix of flat and non-flat terrain, and intermittent traffic, sometimes heavy on city driving.

The net takeaway: If highway range matters a lot to you, leave off the 19" wheels That said, even on the 19"-ers, you'll likely get more than the rated range in mixed driving - just not in pure highway driving. With the 18" aeros, you can expect an improvement over the rated range even in highway driving.

 

[3V Pilot]

All of this number crunching is great and I love the details but it will never be exact in real world conditions because there are too many variables. The best use of this data is to come up with estimated rules of thumb that are easy to remember and can be applied very simply. What I've noticed from all this is one that stands out pretty clear, at least for the LR with areo caps, but that's all I'm concerned with. This is just a quick estimate and YMMV but......

Simple range rule of thumb #1:

At highway speeds every 5mph=20 miles of range.

 

[Kbm3]

@KarenRei , look, can we just agree that anyone getting the Model 3 LRB, as observed simply by @Dan Detweiler and myself earlier in this thread, will most likely get one of the longest range T≡SLA vehicles currently made in Fremont, with maybe just the exception of the S 100D...?

And leave it at that... :tmi:

In the end, I find this the most encouraging news of all (am even past the HUD now!! ) and one that could make me stretch my budget by 10-20% to make sure I get Midnight S≡R≡NITY to be right for me for the next decade!!

Thank you, T≡SLA!

I love the back and forth between two of the posters I respect the most.

As long as it stays civil, and no tempers are lost, I see no problem.

 

[MichelT3]

Anyone who pretends to know 'the truth' concerning energy use, should go back to school. There are too many differentiating factors in roads, conditions, cars and drivers. We can only use estimates, know the deviation and outliers and get enough grips on our own driving style to factor that into our planning.
So can we please refrain from posing to know whats 'wrong' or 'the truth'? All gathered info will help us further. Questions are more important than answers.

 

[Reef Club]

Let's not forget that apparently a VP of Engineering at Tesla mentioned a 10% increased range with 18" aero wheels.
https://electrek.co/2017/08/22/tesla-model-3-aero-wheels-can-increase-efficiency-vp-engineering/

Recently Kim from Like Tesla drove by the SpaceX facility in Washington State. She saw two Model 3s in the parking lot (starting at 11:50) and both had aero wheels. Maybe there is something to these 18" aero wheels that the rocket scientists must know!

(starting at 11:50)