Tesla Model 3 Mid Range Highway Range Rating Is Actually 251 Miles


Model 3 Mid Range has 251 miles highway range according to EPA

Here is a screenshot from the EPA spreadsheet that has the new data. The numbers on the right side show 270.1 miles for city range and 247.6 miles for highway range. Combined range is 55% of city + 45% of highway range. In this instance, the combined range is exactly 260.0 miles because of 270.1*0.55 + 247.6*0.45= 260.0.

However, these are not the actual scores. The actual numbers are 274.1*0.55 + 251.2*0.45= 263.8. You can see these numbers when you select the cells and look at the formula bar. City range is 274.1 miles and highway range is 251.2 miles.

Here is a table that summarizes the current situation. Advertised range, EPA rated range and EPA combined range, are different names for the same thing. In an ideal world, the advertised range of the LR should be 318 miles, LRD and P should be 296 miles and MR should be 251 miles.

A detailed version of this table that shows the range numbers at different speeds can be found here: Tesla Range Table. I updated this table based on the latest data in this EPA document.

Other new data points in this document are the MPGe (miles per gallon equivalent) numbers. For Model 3 Mid Range, highway MPGe is 116.97. EPA defines MPGe as range per 33.7 kWh electric meter consumption. 1 gallon of gasoline is equal to 33.7 kWh energy according to EPA. That’s the conversion factor they use for all EVs. You can compare the MPGe numbers of different Teslas here on the EPA website.

Looking at the MPGe numbers, we can do some calculations. If highway range is 116.97 miles per 33.7 kWh electric meter consumption and total highway range is 251.2 miles, then we can calculate the total electric meter consumption for a full charge. It would be 33.7*251.2/116.97= 72.37 kWh for the Mid Range. This data alone wouldn’t mean anything but luckily we also know the following about the Long Range Model 3 battery:

  • 89.41 kWh electric meter consumption
  • 80.5 kWh total capacity
  • 78.27 kWh usable capacity
  • 4416 cells

If 89.41 kWh wall consumption corresponds to 78.27 kWh usable capacity, then 72.37 kWh wall consumption would mean 72.37*78.27/89.41= 63.35 kWh usable capacity for Mid Range.

The LR pack has 80.5-78.27= 2.23 kWh buffer (brick protection). Model S data shows that the buffer remains the same for 75, 70, and 60 kWh versions. Therefore it should be the same for LR and MR (2.23 kWh). That means the total capacity of the MR pack should be 63.35 kWh usable capacity + 2.23 kWh buffer= 65.58 kWh.

If 80.50 kWh corresponds to 4416 cells, then 65.58 kWh would be 65.58*4416/80.50= 3598 cells. This looks like a rounding error. I will assume 3,600 cells. Here is a summary:

I want to be clear that 3,600 cells for the Mid-Range battery is an estimate I’m calculating based on electric meter consumption numbers. 3,600 cells is kind of an odd number because all Model 3 packs have 96 bricks and if each brick had 37 cells, the cell count would be 3,552 cells. If it were 38 cells per brick, cell count would be 3,648. For 3,600 cells to be true, the pack needs to have 48 bricks with 37 cells and 48 bricks with 38 cells. That’s what the theory shows. I’m sure we will eventually find out whether 3,600 cell is the actual number. My previous estimate based on vehicle weight was 3,648 cells but based on this latest data, I have to go with 3,600.

I wish EPA would stop using the 55% city plus 45% highway method to calculate EPA rated range. They could continue using the combined range for MPGe and switch EPA rated range from combined to just highway range. Expecting Tesla to advertise realistic numbers while all other EV manufacturers continue to advertise whatever the EPA score shows would put Tesla at a disadvantage. Therefore the solution should be to making the range numbers more realistic for all EVs.

Even though EPA rated range is somewhat optimistic, it doesn’t look half as bad if we look at the messy situation in Europe. They used to have a range test called NEDC which was too unrealistic. They came up with a new range test called WLTP which was supposed to be better but it’s still too optimistic.

For example, the numbers are as follows for Model 3 LRD:

  • 296 mi (476 km) EPA highway range
  • 310 mi (499 km) EPA rated range in North America (5% over EPA highway range)
  • 338 mi (544 km) WLTP rated range in Europe (14% over EPA highway range)

Even the 296 mi EPA highway score is not exactly a good measure because it corresponds to the range at 68 mph. The range drops to 287 mi at 70 mph and 264 mi at 75 mph. For more details, see the Tesla Range Table.

Source: Teslike.com

Categories: Tesla

Tags: , , ,

Leave a Reply

34 Comments on "Tesla Model 3 Mid Range Highway Range Rating Is Actually 251 Miles"

newest oldest most voted

Can you imagine anyone splitting hairs like this about the range of a gas powered car? The more chargers, the less interesting this gets…

Right I feel it will be too technical for many people better make it simple.

This is pretty deep “Inside Tesla” all right. I know from discussion on the Tesla Motors Club forum that there are some who have this level of interest, but I think even on that forum it’s a pretty small percentage of owners.

The EPA highway score is not based on 68 mph. It is a test cycle with an average speed of 48 mph and a top speed of 60 mph: https://www.fueleconomy.gov/feg/fe_test_schedules.shtml

Correct. The EPA highway dyno test score for MR is 358.7 miles at 48 mph average speed. The document that will show this number has not been released yet. After the dyno test, EPA uses the 0.7 multiplier (it’s called the ‘derating factor’) to calculate the highway range. When they multiplied the 358.7 mi highway dyno score by 0.7, the result was 251.2 mi EPA highway range. This document shows the 251.2 mi highway range number which is what you get after multiplying the dyno score by 0.7. Nowhere it mentions that 251.2 mi is 68 mph. I came up with 68 mph because it corresponds to 68 mph based on my range table. For Model 3 LR, the EPA Highway dyno score is 454.7 mi (at 48 mph average) and the highway range is 454.7 mi*0.7= 318.3 mi. There is one document that shows the 454.7 mi dyno score and another document that shows the 318.3 highway range. However, for the MR, the document that shows the dyno score has not been released yet. We only have the one that shows the highway range which is the number you get after multiplying the dyno score by 0.7. To be… Read more »

I wouldn’t put too much faith in EPA numbers and numbers derived from them. It’s unknown which value is using new method, old method, old old, or old old old method. Ballpark for one car may be ok, but comparing to other model is dubious (ie, MR to LR).

Hi. All the details are known. For example, there is a spreadsheet that shows what multipliers they have used for each Tesla model to convert the dyno score to EPA highway range. They used 0.7 for MR and LR and 0.7032 for LRD and P. The dyno test is always performed the same way. It never changed. Therefore the dyno scores are the best way to compare one tesla model to another. All the marketing manipulation happens after the dyno scores. The dyno scores are not manipulated in any way.

EPA rate Tesla 3 LR to have higher MPGe rating that lighter MR. This makes no sense. As well, they rate BoltEV being bit more efficient that SparkEV when the real world usage shows SparkEV to be about 20% more efficient. Until these things are worked out (unlikely), I have no faith in EPA comparing different vehicles within 20% margin of error.

I appreciate you going to the effort to put this together, but I don’t think it’s appropriate to be presenting the conclusions in this article as facts. For example, you yourself said in the article that you are only estimating the number of cells in a Mid-Range pack.

Your numbers are partly based on some assumptions, rather than on solidly verified facts, and thus your conclusions can only be assumptions, or rather estimates… not facts.

* * * * *
I find it strange that you try to make the total number of cells fit the assumption of the same number of cells per module. We know this isn’t true for the LR version:

“Another particularly compelling observation highlighted by Rickard was that the four modules of the [Long Range] Model 3 battery pack were not identical, with two modules featuring 25 cell groups in series and the other two featuring 23 cell groups in series.” (source below)

There is no logical reason to assume the MR pack has the same number of cells in each module, any more than the LR pack does.


Did you skip the part where I said, “I want to be clear that 3,600 cells for the Mid-Range battery is an estimate I’m calculating.” 251 mi highway range is the number in this EPA document. There is a link to the document and you can open the document and see 251 mi highway range yourself. That is the fact. 3600 cells is an estimate. In the article it says estimate.

You are confusing the number of bricks per module with the number of cells per brick. The LR pack has 4 modules. 2 modules have 25 bricks and 2 have 23 bricks. In total there are 25+25+23+23= 96 bricks. Each brick has 46 cells. The pack has 96*46= 4416 cells. The SR pack also has 25+25+23+23= 96 bricks but it has 31 cells per brick. Therefore it has 96*31= 2976 cells. Both the SR and the LR packs have identical number of cells per brick. In other words, all bricks in the LR pack have 46 cells and all bricks in the SR pack have 31 cells. However, MR appears to be an exception because the calculation shows 96*37.5= 3600 cells. The reason for 96 bricks is cell voltage. Each cell and each brick is 3.7 Volts. With 96 bricks you get 3.9*96= 355.2 Volt. However, officially its rated at 35 Volt.

I’m not confusing anything, Troy. The article I linked to, which is based on an actual teardown of an actual TM3 battery pack, and not merely on a spreadsheet calculation, shows the LR battery pack is composed of 4 modules, and of the four, two have a different cell count than the other two:

2 x 25 x 46 = 2300
2 x 23 x 46 = 2116

Total 4416 cells.

In your article, you say:

“If 80.50 kWh corresponds to 4416 cells, then 65.58 kWh would be 65.58*4416/80.50= 3598 cells. This looks like a rounding error. I will assume 3,600 cells.”

No, it doesn’t look like a rounding error, and your assumption ignores the facts. It looks like the MR pack, just like the LR pack, is composed of modules with different cell counts.

It looks like 3598 might have been spot on after all. I have now updated the original article on Teslike dot com. My new calculation is 23*2*38+25*2*37= 3598. I have also posted a tweet about 3598 cells being my new estimate. Thanks


Your new calc for 3,598 active cells appears to assume that the smaller 23S modules have 38 cells in-parallel (38P) and the larger 25S modules have 37 cells in-parallel (37P).

That is not possible. All modules are connected in series. Every 2170 cell installed in the pack has the same amp-hour rating and nominal voltage and must carry the same amp load. For each cell to carry the same amp-load and discharge at the same amp-rate, the pack must have same # of cells in-parallel in all modules that are connected in series. If some modules have fewer cells in-parallel, there will be an amp imbalance at the cell level both during charge and discharge cycles. That is a cardinal sin in battery pack design as the more-heavily-loaded cells will discharge/charge faster plus get hotter than the other less-loaded cells. All bad for cell life and pack efficiency.

Whatever your nominal total kWh or A-h calcs suggest, the MR pack’s active cell count has to be in increments of 96: 96S36P = 3,456 cells, 96S37P = 3,552 cells or 96S38P = 3,648 cells. Pick one. There are no half-cells or “mid-way” counts.

I have looked into this and it turns out you are right. When connecting batteries in series, each battery needs to have the same voltage and same Ah capacity. Therefore it wouldn’t work if 46 bricks have 38 cells and 50 bricks have 37 cells. That means the correct answer should be 3,552 or 3,648 cells for the MR battery.

Pushy — This is the frustrating part of Troy’s numbers. They are constantly being revised, yet always repeated across the internet as if they were biblical fact, even it they are labeled estimates.

In reality, the very tests themselves simply aren’t accurate enough to support much of the math. Dyno’s simply aren’t that consistent. Even the same dyno on different days can vary. The math he is attempting simply can’t lead to the accuracy of battery count he would like to do. It is like looking at a speedometer and then calculating down the the second how long a 100 mile trip will take. The speedometer error just isn’t accurate enough to get that accurate of results, even when all the math is done right down to 3 decimal places.

Just between you and me, I wouldn’t go down the rabbit hole. It just isn’t worth it, and nobody outside Tesla knows the real answer until there is a teardown. And then a month later Tesla can make minor revisions to the packs that only show up as part number changes and we are back to zero.

Let’s put your theory to test: I’m sure you know that the Model 3 LR scored 310 miles in Consumer Reports’ range test with regen set to low and 350 with regen set to standard. I estimate that the MR would score 245 mi with low regen and 276 mi with standard regen in Consumer Reports’ test. Hopefully, they actually test it so we can find out whether my estimates are pretty close or wildly inaccurate.


Yes, thanks for that summary of the situation, and (as usual) I entirely agree with your comments.

Oh the other hand, HVACman (I think that’s George Bower?) says the number of cells have to be divisible by 96, and I certainly wouldn’t argue with him! I did play around a bit with the math and found that 4416 can be divided evenly by a surprising number of integers (2, 3, 4, 6, 8, 12, 16, 24, 32, 48, 64), but I stopped before getting to 96.

That doesn’t necessarily mean that every cell count in a Tesla pack has to be evenly divisible by as many integers as 4416, but it’s not surprising to me that the count does have to be evenly divisible by some integer larger than 10.

BTW — George Bower cited your analyses, but his estimate is 3446 cells for the Mid Range. It may well be that your estimate is more accurate than his, but both are just estimates.


Who said they are not estimates? You are talking as if I presented the cell count as a fact. I didn’t. I specifically said that it is an estimate.

Yes, you do specify at one point that you’re just making an estimate; yet your concluding paragraph reads:

“However, these are not the actual scores. The actual numbers are…”

I’d say the article sends mixed signals about whether the figures are just an estimate or are the actual figures.

Interesting to see that this article gives, on one of the charts, “advertised capacity” for the TM3’s three battery sizes. So far as I have ever seen, Tesla has officially refused to specify the kWh of any version of the Model 3.

I wonder what the author is using as his source for the so-called “advertised” capacity. Looks to me like it should be “anti-advertised”!

After seeing an increasing number of “my Model 3 LR is getting less than 200 miles of range” (some of which use some not so nice language about Tesla “lying” about the range) posts on the forums it is clear to me we need a better way of expressing range for BEVs at various temperatures and speeds. Having lived with a Volt for 5 years before getting my model 3 I learned one could count on the “my car lost range it is my battery failing?” type posts from new owners coinciding with the arrival of winter weather. But, folks new to EVs are shocked (pun intended) by the impact of cold weather and high speeds on EV range. For BEVs to become mainstream we need some way of expressing this to prospective buyers. Perhaps a table with temperatures on the Y axis and speeds on the X axis with estimated range in each cell? Obviously the climate control setting would need to be included so maybe just use 70 to keep it simple.

I’ve often wished that the EPA would give us a series of numbers for EV range, with speed being a variable: Different ranges for 35, 45, 55, 65, 75, and perhaps even 85 MPH. Since range varies so much by speed, I question how useful it is to have just one number for range.

Re temperatures: Yeah, we are seeing a lot of comments in the Clarity PHEV section of the IEVs Forum from shocked Clarity drivers experiencing their first winter, and their first exposure to range loss. Some are even reporting a range loss of 40% or more! And I find that shocking, because I thought the norm was 30% or less, with anything above 30% being an outlier figure. But then, the Clarity is a PHEV and not a BEV, so perhaps it’s wrong to assume the percentage loss should be the same as it is for Tesla’s BEVs.

The thing is, though, I don’t see how one could give a meaningful average for range loss due to temperature. It greatly depends on how much energy is used to run the cabin heater. If you drive slow for a long time and have the cabin heater cranked up, you’re going to lose a lot of range. If you’re driving at highway speed, depending mostly on seat/ steering wheel heating and minimizing use of the cabin heater, then that impact on range (for the same distance driven) will be much less.

There are too many variables there to easily sum up in a single chart.

75 kWh was written inside the first Model 3 LR owners’ manual on page 137. In later versions, they removed 75 kWh. 50 and 62 kWh is leaked information.

Thank you! 🙂

Attempting to simplify it down to one number is no use. In the UK we’re used to ICE cars being quoted with 3 mpg figures for city, combined and motorway. There’s no reason why EVs should be different – then we can supply our own weightings (especially since we don’t all live in cities).

Interesting analysis, but the EPA testing is based on historically the mix of what the majority of people actually drive. Rating cars on just highway rating would be meaningless to most people, as people don’t just do highway driving.

I’m surprised that Tesla has such a small buffer in the battery. They should just put in 10% or more, so that users don’t have to think about what percentage tp charge to, for best battery life.

Jean-Baptiste Labelle

No thanks. I much prefer having the best range when i need it. 98% if the other time, i charge to 80% which is enough for what i do.

What we do need is TeslaBjørn to do this test. Especially for us Nordic countries. Can’t be more than 3-4 months away now.

So, the point is, that the true EPA combined rating for the Model 3 MR is 263 mi? That’s all I want to know for direct comparisons. Where can I find this info for all BEVs on the US market?