What's up with Tesla's range numbers?

No, this is not another "dieselgate".

If you’re shopping for an electric vehicle, it’s likely that one of your top concerns is how far you’ll be able to drive it before needing to stop to recharge. This is commonly referred to as the vehicle’s range. Unfortunately, answering that question isn’t quite as simple as we all wish it to be. Why is that?

First, let’s talk about range ratings. Every vehicle sold in the US provides a range figure based on a protocol established by the Environmental Protection Agency (EPA). Outside the US, a variety of testing protocols are used - so the rating you see for the same vehicle may differ somewhat (or a whole lot) depending on which country the vehicle is being sold in. Common rating systems include the EU’s WLTP standard, the largely obsolete NEDC standard, and China’s newer CLTC standard.

Of these, the EPA rating is the most conservative and the most accurate for EVs. The WLTP numbers tend to be an average of 11% higher than EPA range figures, and the CLTC numbers tend to be 15-25% higher than that. When you read about Toyota talking about having a 600 mile range EV later this decade, keep in mind that this aspiration is stated in CLTC terms - meaning the same vehicle may achieve something much closer to 400-500 miles in the EPA test.

And yet, the EPA range figures are generally not conservative enough. In many real-world tests, most EVs don’t achieve their EPA rated range, even when brand new. What gives?

Let’s say you’re shopping for an EV, and you find that the Tesla Model Y Long Range has an EPA range rating of 330 miles - can you expect to be able to charge to 100%, hop in, and drive 330 miles without stopping?

The answer is: Maybe, but probably not.

Let’s look at why that is, and then what can be done about it.

Is Tesla Cheating?

Recently, a number of claims have made the rounds in social media about Tesla’s range numbers. Some have said that Tesla is facing their own “Dieselgate” scandal. One person making this claim recently was Cory Doctorow, who wrote a long article attacking Tesla - mostly based on falsehoods and misunderstandings. When it came to this topic, Cory wrote…

Now, Tesla is having its own Dieselgate scandal. A stunning investigation by Steve Stecklow and Norihiko Shirouzu for Reuters reveals how Tesla was able to create its own demon-haunted car, which systematically deceived drivers about its driving range, and the increasingly desperate measures the company turned to as customers discovered the ruse:

https://www.reuters.com/investigates/special-report/tesla-batteries-range/

The root of the deception is very simple: Tesla mis-sells its cars by falsely claiming ranges that those cars can't attain. Every person who ever bought a Tesla was defrauded.

Fortunately (or unfortunately), none of this is true.

Cory goes on to tell a tale about how Tesla’s “range predictor” lies to owners, suggesting that this accusation is supported by the Reuters article he linked to. It isn’t. But the article is written in a way that seems intended for people like Cory to get the wrong idea.

So, what are the claims in the Reuters piece? There are two:

1. That at some point early in the company’s history, Tesla programmed its range-estimating software to give a more optimistic prediction when above 50% charge, and to switch to a more pessimistic one when below 50% charge.

2. That Tesla created a dedicated team to “suppress thousands of driving range complaints”.

The first of these claims is attributed to one unnamed individual. Specifically, the authors say this person was “familiar with an early design of the software for its in-dash readouts”. We can say with absolute certainty that this is not true today (as the range predictor has been thoroughly tested and even reverse-engineered), and that it has not been true since the introduction of the Model 3 in 2017.

Was it true for older Model S or Roadster vehicles? Thus far, I haven’t seen any evidence of this. It is possible, but even if it did happen, it hardly seems scandalous. What’s described is a simple heuristic, which for a v1 product might be entirely reasonable. Indeed, many combustion engine trip computers are overly optimistic when you have a full tank of gas, and get more accurate as your fuel level gets lower. However, the statement about someone having seen an “early design” could even mean they saw something that was never released to the public. Sadly, the authors didn’t seek any hard evidence to support or refute this claim.

As for the second claim… I’m going to wait to address that until a bit later.

Nowhere in the article does it suggest that Tesla’s EPA numbers are “faked” or manipulated in any dishonest way. In fact, it says just the opposite, revealing that the EPA validated Tesla’s numbers, and in some cases required small adjustments before Tesla advertised their figures. Here’s what the article says:

Whatever an automaker decides, the EPA must approve the window-sticker numbers. The agency told Reuters it conducts its own tests on 15% to 20% of new electric vehicles each year as part of an audit program and has tested six Tesla models since the 2020 model year.

EPA data obtained by Reuters through the Freedom of Information Act showed that the audits resulted in Tesla being required to lower all the cars’ estimated ranges by an average of 3%. The projected range for one vehicle, the 2021 Model Y Long Range AWD (all-wheel drive), dropped by 5.15%. The EPA said all the changes to Tesla’s range estimates were made before the company used the figures on window stickers.

And later…

“I’m not suggesting they’re cheating,” Pannone said of Tesla. “What they’re doing, at least minimally, is leveraging the current procedures more than the other manufacturers.”

Despite what some would have you believe, Tesla’s EPA range numbers are in fact legitimate, and have been validated by the EPA as well as independent testers.

Then why don’t drivers get the advertised range?

Sometimes, they do! There’s a reason the phrase “Your Mileage May Vary” has been the standard disclaimer for EPA range and efficiency metrics for decades. These kinds of range estimates are, well, estimates - and they can only tell you how the vehicle performs in the tested conditions.

There are five main reasons why most EVs don’t achieve their EPA rated range in common tests or many real-world scenarios. Let’s go through them one by one.

Reason #1: The “combined cycle” test

You may recall a time when new vehicles would advertise two fuel economy ratings - one for city driving, and one for highway driving. For combustion engine vehicles, the city rating was almost always significantly lower than the highway rating. One reason for that is the energy lost to braking, combined with the energy needed to accelerate the car back up to cruising speed in between stop signs and traffic lights.

Hybrid vehicles were the first to start challenging this conventional wisdom, with the Toyota Prius generally achieving very similar numbers for city and highway efficiency - in most cases, even higher for city driving. This is primarily due to the use of regenerative braking, which uses the electric motors to slow the car, and in doing so captures energy and puts it back into the battery. With braking losses minimized, the lower air resistance associated with driving at city speeds makes up for the rest and can result in getting greater efficiency than on a highway.

This is true for EVs as well. In fact, the effect is exaggerated with fully electric cars, largely because of two factors: the aforementioned regenerative braking, and just how incredibly efficient they are to begin with. Combustion engine cars rarely achieve above 30% efficiency (the measure of how much energy from the fuel is used to propel the vehicle, rather than being lost as heat or noise). EVs are typically 90% or more efficient at converting the electricity stored in their batteries into propulsion. This means that factors other than the drivetrain’s efficiency have a more pronounced effect on overall efficiency and thus, range.

The Monroney sticker in the window of a new car will still list both highway and city efficiency, but the larger and more obvious rating is for what the EPA calls a “combined cycle”. This is what the EPA’s single range estimate is based on. The particular EPA test is called the Multi-Cycle City/Highway Test Procedure (which, for EVs, is based on the SAE J1634 standard). This is designed to approximate a mix of city and highway driving.

With a gasoline or diesel-powered car, you would be right to expect that in favorable conditions you would get more than the stated combined cycle range if you drove the entire tank from full to empty on the highway. However, with an EV (and to some extent, hybrids), highway range should be expected to be lower than city range, and thus lower than the combined cycle range.

If you look up the current Model Y Long Range on the EPA’s FuelEconomy.gov website, you can see that the efficiency numbers listed include 127 MPGe (“miles per gallon equivalent”) for city driving, and a lower value of 117 MPGe for highway driving. The combined value is right in the middle at 122 MPGe.

This also means that a manufacturer can increase their combined cycle range by improving only the result of the city range test. Some of Tesla’s engineering investments, such as their highly optimized regenerative braking scheme, largely affect city range - with minimal benefit to highway driving. Those benefits are real, but they mean little or nothing if your goal is to assess how far you can drive on an interstate road trip.

The last relevant detail I’ll point out here is that the combined cycle value is not an even split between city and highway results, but rather weighted 55% toward the city result.

Reason #2: Speed matters more than you think

While the mix of city and highway driving covered in the EPA’s combined cycle test begins to explain the discrepancy between some buyer expectations and reality, another problem is how the EPA defines “highway” driving.

The EPA’s testing procedure involves testing the vehicle on a dynamometer (a “treadmill for cars”) where it performs simulated city and highway loops with average speeds of 20 MPH and 48 MPH respectively. If that sounds low, it’s because it is. In between these loops the car is run at a steady 55 MPH or in some cases 65 MPH. Since this isn’t representative of the real-world (both because of speed and a lack of air resistance, which we’ll talk more about in a moment), the EPA then applies an “adjustment factor” to the results. By default, it reduces the resulting range figure by 30 percent. Since this 30% value isn’t very scientific, the EPA allows manufacturers to perform additional tests which can be used to reduce that adjustment factor. Tesla and Audi have both been reported to take advantage of this.

Air resistance is generally the dominant factor in determining relative efficiency and range at highway speeds for EVs. You may find it intuitive that air resistance becomes a bigger problem the faster you go, but what you might not have realized is that the relationship between speed and air resistance is not linear. In fact, air resistance increases quadratically (that is, as a square) with respect to speed. This means the difference between 70 and 75 MPH is substantially more noticeable than the difference between 65 and 70 MPH (let alone a 5 MPH increase at even lower speeds).

Depending on where you live and what I’ll call “driving style”, your typical highway speeds probably start at 60 MPH at the low end and go upwards of 80 MPH at the high end. If you stuck to the former, in favorable conditions you could often exceed the EPA rated range for any EV, including any Tesla. However, at 80 MPH, you’re not going to come close.

Just how big of a difference does speed make? A few years back, @TroyTeslike maintained some incredibly handy tables showing a number of range-related figures for every model and configuration of Tesla vehicle. While he unfortunately no longer produces these, I find even these 2019 tables incredibly helpful for understanding the effect of speed as well as certain configuration options, like wheel + tire choice.

What does this tell us? Well, let’s say we’re back in 2019 and you’re looking at a then-new Tesla Model 3 Long Range with an EPA rated range of 310 miles. If you opt for the most efficient configuration, with 18” wheels (and you leave the provided aerodynamic covers in place), you should find that at a steady 65 MPH you could exceed that rated range. However, at 70 MPH, you’ll fall 12 miles short. At 75 MPH, you’ll only go 275 miles - a full 35 miles less than the EPA range estimate. Push it to 80 MPH, and you’re only going 251 miles - 59 miles below the rated value.

However, if you drove this same car at a steady 50 MPH, you could achieve over 400 miles of range! This just isn’t something any of us are likely to do very often. That said, I did get a taste of this higher-than-normal efficiency while driving down the Pacific coast last summer. Scenic highways like US 101 can be the ideal kind of place to get the most out of your EV battery.

This also means that, counterintuitive though it may be, traffic is good for highway EV range. One of the only reasons my car’s prediction about our battery level on arrival ever fluctuates is due to changing traffic conditions. If a traffic jam appears, that means it will take more time but less energy to reach our destination. If traffic clears up while we’re en route, we’ll get to our destination faster, but we’ll use more energy to do so.

Tesla drivers will sometimes notice a message from the car telling them to keep their speed under a certain value (e.g., 80 or 75 MPH) to reach their destination without running out of juice.

Image source: CleanTechnica

Reason #3: Configuration matters

Back in 2017-2019, Tesla rated all Long Range and Performance versions of the Model 3 with the same 310 mile EPA range figure. They actually voluntarily downrated the rear-wheel drive model (which achieved a 325 mile range in the test) to match the dual-motor AWD version. My guess is they did in order to avoid encouraging customers to buy a cheaper, lower-margin model.

At the same time, they didn’t give a unique range rating to the Performance model, nor to alternate configurations such as the “sport” wheel option. They were not required to do so under the EPA’s rules - many manufacturers do the same and have for decades. Published EPA ratings tend to be for the base version of a given model and in the cheapest (and often range-optimal) configuration. However, I think this was a mistake, as it led to buyers of the Performance model or the 19” Sport Wheel option more likely to be disappointed by their real-world range.

As the chart in the previous section shows, wheel options can have a substantial effect on range. For our hypothetical 2019 Model 3 LR AWD at 65 MPH, the standard 18” aero wheels were good for 6% greater range - enough to make the difference between being 15 miles over the rated range versus 4 miles under it. The effect even becomes slightly more pronounced at higher speeds. Most of that difference is due to the aerodynamic wheel cover - with the rest likely due to differences between rolling resistance of the tires and the weight of the wheels.

Fortunately, a few years ago Tesla began giving separate range ratings not just for different variations like the Performance models, but also for different wheel options. For example, Tesla’s longest-range vehicle today is the standard Model S. It has an EPA rated range for 405 miles on the default 19” aerodynamic wheels. However, Tesla drops that to 375 (technically an unofficial rating, not validated by the EPA) if you select the fancier looking 21” Arachnid wheel option.

I think this is a great thing Tesla does, even though they are not required to do so. What’s more, as of 2+ years ago, the car’s own “rated range” display is based on the chosen wheel configuration (which the owner can change in the car’s Settings screen - useful if you change your wheels or have a different summer versus winter set).

Reason #4: Conditions matter

The EPA range estimate does little if anything to account for variable conditions which can affect range. When looking at the published EPA range figures for any EV, it’s best to assume that the figure represents:

• A brand-new battery

• No headwind

• Relatively flat terrain

• Moderate temperature and humidity

• The vehicle’s default settings for acceleration, regenerative braking, etc.

Battery age is a factor because, just like your phone battery, an EV battery will slowly lose capacity over time (and charging cycles). The rate and degree to which the battery degrades depends on a wide variety of factors including how much capacity the manufacturer reserves as a “buffer” against degradation, the battery chemistry, and in some cases how much time the battery spends fully or nearly-fully charged.

Many EVs, including many Tesla models, experience faster battery degradation if the battery is left fully charged at or near 100%. To help improve longevity, Tesla and many others recommend that you set a charge limit on the car in daily use, capping it to 80% or 90% by default. Charging to 100% is then reserved for road trips or situations where you expect you won’t be able to charge for a while. Many enthusiasts who want to preserve their batteries as much as possible will go so far as to time it so that the car reaches 100% charge just before their scheduled departure. I do this on my Model S, though I don’t really know how much of a difference it really makes.

It’s worth noting that some Tesla “standard range” models now use Lithium Iron Phosphate (LFP) batteries, which don’t experience this effect. In these vehicles, Tesla recommends leaving the charge limiter at the default of 100%. The car itself and the manual will tell you which is recommended for your particular model.

Temperature is a factor as well. You may have heard that EVs get reduced range in the winter. This is generally true. In the hypothetical 2019 Model 3 we discussed above, winter weather had a very sizable effect on range, largely because the cabin was heated using a resistive heater. Beginning with the Model Y, and soon after extending to all current models, Tesla switched to a much more advanced thermal management system which relies on a heat pump + waste heat rather than a resistive heater. The result is that today’s Tesla vehicles experience far less range impact in the cold. It’s still impacted, but the effect is closer to that of running your car’s air conditioning on a hot day.

Tip: If you do elect to have a separate winter and summer wheel + tire set, I think a smart strategy is to get smaller, aerodynamic wheels for winter, to offset the impact of cold temperatures on range.

Reason 5: Phantom Drain

A combustion vehicle’s gas tank normally doesn’t lose any of its contents while it’s sitting around (unless you have one of those silly hydrogen combustion vehicles). However, you may be aware that if you leave such a vehicle parked for an extended period, you can find that the 12V lead-acid battery has been depleted, and the vehicle will no longer start. This is because the battery’s charge is slowly drained while the car is off, as it is used to power the door locks, car alarm, clock, and sometimes other components.

A similar thing happens with EVs. In fact, all currently available EVs rely on a regular 12V battery (usually lead-acid, but sometimes lithium-ion as in Tesla’s newer vehicles) in addition to the high-voltage “traction” battery (the giant one usually under the floor which lets you drive around using only electricity). While parked, the vehicle will drain one or both of these batteries.

Generally speaking, most EVs disconnect the high-voltage battery when parked. You can usually hear the contactors engaging or disengaging after you park or when you approach or unlock the car. They then drain the 12V battery over time, often faster than a typical combustion engine car. The reason they drain it faster is because they tend to have more electronics running even while idle, including for things like automatic software updates. When the 12V battery gets low, the car will re-engage the contactors and connect the high-voltage battery, and then use that to charge up the 12V battery before disconnecting again.

Normally, this “phantom drain” is relatively minor. Though even in normal conditions it can be around 1% per day. Some vehicles will go into a “deep sleep” when left parked for long periods to reduce this level somewhat. If you’re on a long road trip and you lose 0.5-1% while parked overnight, it probably isn’t a huge deal. However, if you don’t have the ability to charge your car at home and rely on weekly visits to a DC fast charger, this effect may be more of an inconvenience.

If you use certain features, idle drain can be much higher. Tesla offers a few features which can significantly increase your idle battery drain. They are:

• Sentry Mode - an advanced security mode where the car keeps the Autopilot cameras active while parked, monitoring your surroundings, and recording video if someone gets uncomfortably close to your car

• Summon Standby - an optional feature which keeps the Autopilot system running while parked in order to make the Summon feature immediately available (normally, it can take 20-30 seconds for the car to wake up and be ready for you to use this feature)

• Cabin Overheat Protection - This is enabled by default and turns on the air conditioning if the interior temperature exceeds a specified value (the default is 105F). This can be set to Fan Only or disabled, and only operates for 12 hours after you leave the car and if the battery level is over 20%.

• Keep Climate On, Dog Mode, Camp Mode, and remote climate control - These features keep the climate control system (and in some cases more) running while parked. These all require the user to enable them on-demand, and the car will warn you if you try to use them when the battery is below 20%.

Sentry Mode is probably the top culprit for high battery drain while parked, though cabin overheat protection can also surprise owners, particularly because it’s enabled by default. Some features, like Sentry Mode, can be configured to be location-aware, e.g., not turning on when at home, at work, and/or at any locations you set as “favorites” in the navigation system. This can help balance the utility of the feature with unnecessary power usage in locations you consider safe. Owners can also control Sentry Mode (turning it on or off on-demand) remotely from the mobile app. If I plan to have the car parked somewhere for an extended period while on a trip, I will often manually disable Sentry Mode from the app to preserve power.

In my experience, both Sentry Mode and Summon Standy can increase idle drain to around 5% per day, but the exact amount can vary depending on a variety of factors (including vehicle model, software version, number of detected events, and even outside temperature).

Recapping real-world range versus EPA range

I hope you now have a solid understanding of why EVs often don’t achieve their EPA rated range values in common use cases. The next time you see an article like this one from Car & Driver showing this phenomenon, you can notice the part where they say their test involves driving non-stop at 75MPH and say, “ah ha! We should not expect these cars to achieve their EPA combined cycle rating in this test!” Isn’t being a huge nerd well-informed great?

Rated Range vs Predicted Range

With all of that out of the way, I wanted to take a moment to explain something Tesla does that I think causes confusion among some buyers and owners, and which I would like to see them change.

Tesla provides two ways to display the battery’s state of charge (AKA “battery level”, or what Tesla labels as “Energy Display”). They call these two options percentage and distance. The latter is, last I knew, the default. I would prefer that it did not exist at all.

The percentage option displays your battery level as a value between 0% and 100%, much as your phone and laptop probably do. The distance option displays your battery level in miles (or kilometers). What does it mean to display battery level in miles? Simple. You take the car’s rated range value (e.g., 405 miles for a new Model S) and display that when the battery is charged to 100%, and you scale the figure down linearly as the battery is depleted.

It is very, very important to understand that this is not a prediction for how far the car will go based on your driving habits or typical routes. It’s a very simple (“dumb”) calculation. It’s often referred to as the rated range value. Unless you are driving in favorable conditions and following the same mix of driving scenarios covered in the EPA’s Combined Cycle test, you should not expect to be able to drive the displayed number of miles.

One interesting thing about this rated range value is that the software adjusts it as the battery ages. When my Model S was new, it had a rated range of 375 miles with its wheel configuration. When fully charged, that’s what would be displayed. Now, after 2 years and 12,500 miles or so, the rated range at 100% is 364 miles. This is the one very minor way in which I think you could argue this option is useful. I’d rather this “capacity at full charge” value be displayed in a settings page (as it is on an iPhone) instead.

While I have heard that some people like this method of displaying the battery capacity, I believe Tesla should simply remove this option. Or at least make the percentage option the default. This rated range figure causes much more confusion than it is worth, and too many people (whether owners or just armchair critics) point to this as a predicted range figure, then point out that it’s not very good at being that thing that it isn’t.

Tesla’s range predictions are next-level

Tesla does have a range prediction solution, and it’s kind of bonkers how sophisticated and accurate it is. In my experience, it’s always been pretty accurate, but an update in 2021 made it very accurate, and last year they updated it again to factor in a crazy number of details including energy loss to phone charging + 12V accessories, air density, predicted battery heating/cooling needs, loaded vehicle weight (measured by the motors based on inertia), and more.

For some reason, the Reuters article not only failed to distinguish rated versus predicted range, but it also gave a terribly misleading description of how the prediction algorithm works. It’s almost like they wrote the piece without investigating the subject matter or talking to anyone knowledgeable about these things 🤷‍♂️

Range predictions manifest a couple different ways in the car. The primary way this shows up is as part of the navigation UI. Once you’ve selected your destination, the car will tell you what it expects the battery capacity to be when you arrive. If you scroll down, you can also see a round trip prediction.

In this case, I’ve selected a destination 19 miles away. My car’s current state of charge is 79%. This is predicting that I’ll consume 6% of the battery’s capacity to go 19 miles. That works out to 3.16 miles per percentage point of battery. That means at this predicted level of consumption, my range is 316 miles, not the EPA rated range of 375 (or the car’s adjusted rated range of 364 miles for my 2+ year old battery).

The other place where this prediction algorithm shows up is in the car’s Energy app. This app does a really great job showing predictions (and real-time feedback) for your current trip, as well as letting you view data from recent trips along with tips for how to maximize range in the future.

In the screenshot above, the Energy app explains why the car consumed more than the EPA rated range worth of miles in the time since my last charge. The largest factors were driving/speed, air conditioning, and elevation change.

If you tap the little circled “i”, the app displays some helpful explanations about all of this.

You can also find out some more about range and efficiency via the Trips screen in the main vehicle UI. If you have a Tesla or end up buying one in the future, I recommend renaming the second “trip” memory entry to “Lifetime” as I have, and never resetting it. Tesla provides helpful “current trip” and “since last charge” figures, so having just one manually controlled slot is sufficient for me. That said, I’m surprised they don’t just provide a “lifetime” entry here by default.

Over the lifetime of my Model S thus far, it has achieved an average energy consumption of 306 Wh/mi, which can be restated as 3.27 miles per kilowatt hour. The available capacity of the battery pack in this model is a little shy of 100 kWh. If I use a value of 95kWh, that means on average I get a range of about 310 miles - considerably short of the EPA rating of 375 (as well as the adjusted figure of 364 given the battery degradation).

On the surface, that might seem poor. However, knowing that most of my driving on this car has been on interstates, with long periods at 70-80 MPH, as well as factoring in winter trips and our relatively frequent drives across Snoqualmie Pass (with sizable elevation changes), this actually seems pretty reasonable to me. It also fits with my own sense from experience that I can generally plan on comfortably driving at least 300 miles on a full charge.

If I look back at Troy’s tables for the old 2019 Model S Long Range (where at the time, Tesla didn’t distinguish the 19” and 21” wheel configurations), his calculation for the expected range at 75 MPH (with 21” wheels) was 266 miles when new. The 2021+ Model S is an entirely new design, weighing nearly 500lbs less than the old one, so I think if he did his calculations against the new version the number would probably be somewhere around 300-310 miles.

How To Think About EV Range

If you’re shopping for an EV, there are two different ways you may want to think about range, and they depend on the answer to one simple question:

Can you charge at home?

If your answer is yes, then you probably have two things to consider:

1. The farthest you drive in a typical day where you’ll be charging up at night while parked at home

2. How far you’ll be able to go between charging stops on a road trip

For most people, #1 is not a particularly large number. Even if it’s 100 or 150 miles, most any EV with an EPA range above 200 miles will suit you just fine. Though, if you’re driving more than 30-50 miles per day, you will probably want to install a “level 2” charging station - which can be as simple as installing or using a NEMA 14-30 “dryer outlet” or a 14-50 outlet (typically used for RV shore power or electric stoves) in your garage or on an exterior wall near your driveway. This will ensure you can recover the charge you depleted each night.

So long as you have home charging which can recharge your battery from daily use each night, you will almost certainly find your EV to be far more convenient than your old combustion engine car. You will never have to worry about “stopping for gas” in regular use.

For you, the only case where you’ll ever have to give any thought at all to charging is for road trips. If you don’t plan to ever do any of those, then you probably don’t even need a long range EV. However, if you’re like us, and you go on trips ranging from 300 to thousands of miles, this is where range matters to you.

Do not expect to achieve the EPA range rating while on a road trip! While you might achieve it on certain scenic drives, if you’re expecting to drive 70+ MPH for most of the trip, you are almost certainly going to achieve lower range than that. Further, you probably aren’t going to actually run your battery down to 0% between charges, and you most likely won’t charge all the way to 100% at charging stops, either. That’s because charging speeds get slower as you get closer to 100% state of charge, and so it’s generally better to charge only enough to comfortably get to your next stop. Good EVs will help you plan an optimal route and will tell you how long to charge at each stop. There are also tools like A Better Route Planner which can be handy if your car doesn’t do a good job at routing, or if you want to explore your options and hyper-optimize your route yourself.

If you can’t charge at home, then you have additional considerations:

1. Where and when will you charge your vehicle

2. How often will you be inconvenienced by the need to get to or wait for a charger

I know some people who do not have a way to charge at home, many of whom live in apartment complexes or condos. If you’re in this situation, you unfortunately don’t get to enjoy the bliss of effectively-unlimited range in daily use. Whether you find the experience better or worse than fueling up a gas or diesel car depends a lot on the answer to #1.

If you have easy access to reliable charging at your workplace, this may serve as a substitute for home charging, and is probably the next-best thing from a convenience perspective. Otherwise, you may have to rely on a mix of “level 2” chargers at grocery stores, malls, and other locations, as well as “level 3” DC fast chargers, such as Tesla’s Supercharger stations, to fuel up. This is doable! But you do have to think about it and it will add some cognitive overhead to driving an EV.

In that situation, the EPA’s combined cycle rating may be more relevant, if your normal use includes a mix of city and highway driving. That said, your mileage may vary, and indeed will.

If you want an EV for road trips in the US, Tesla is unmatched

If you are looking for an EV to take on long road trips, you really can’t do better than a long range Tesla. The Model S Long Range really shines here, but it’s an expensive, high-end vehicle. The Model 3 and Y Long Range variants are the next best option. This comes down to two things:

1. Range

2. Charging network

Tesla’s efficiency is the best in the business (leaving aside the super expensive niche vehicles from Lucid), and this is reflected in their range numbers.

Car & Driver did a great breakdown of Tesla’s efficiency advantages back in 2020, and this all holds true today.

Source: Car & Driver

Since 2020, Tesla has made further improvements, such as the addition of the heat pump and weight reductions I mentioned earlier. This is why they lead in both EPA figures and real-world range.

The second reason is the charging network. They are beginning to open this up to non-Tesla vehicles, but for now it’s only a few stations, and not all vehicles are compatible. Over time, it looks like this situation will improve for other makes of EV. However, even with hypothetical access to an increasing number of Supercharger stations, it remains to be seen which vehicles will actually work well with the Supercharger network.

You can see the effect of Tesla’s advantages in a variety of real-world tests, including Car & Driver’s “EV 1000” test, in which the three tested Tesla vehicles had the three best results. The Model S Long Range (prior generation) spent 1 hour and 36 minutes charging out of a 16 hour and 14 minute trip. The Model Y Performance and Model 3 Performance each spent about 3 hours charging. The next best showing came from a Ford Mustang Mach-E with over 5 hours spent charging and more time spent driving (due to having to divert farther to find a working charging station).

And that that was the best non-Tesla result! The Polestar 2 spent 8 hours charging and over 4 hours more driving than the Model S. And that was in the old Model S which couldn’t charge as fast as the newer generation.

Some other real-world EV road trip comparisons you can watch on YouTube:

• CarWow’s 2023 EV road trip comparison

• Out Of Spec Motoring’s EV Road Trip Race! Lucid Air vs Tesla Model S vs Porsche Taycan - Race To VegasReturn

• MKBHD - Driving 1000 Miles in 3 Cars: Gas vs Electric!

Spoiler alert: Versus other EVs, Tesla always wins.

Oh right, the other Reuters claim

I mentioned earlier that the recent Reuters article claimed that Tesla created a dedicated team whose mission was to “suppress thousands of driving range complaints”. I find this framing to be misleading at best.

For a bit of context… one of the reasons Tesla is currently the only profitable EV manufacturer is their relentless pursuit of operational efficiency. This emphasis is apparent in how they’ve approached scalability of their service department. Service appointments are handled via the Tesla mobile app, which provides benefits for both customers and Tesla themselves. For example, they don’t have to pay someone to answer the phone, and you don’t have to wait on hold and juggle your phone plus your calendar to figure out when is a good time for an appointment. You can also attach photos and videos to your service request to aid in diagnosis. Then you don’t have to worry as much about the problem being reproducible on-demand at the service center or when a mobile service technician arrives.

Another way Tesla has leveraged this to optimize utilization of their service resources has been to preemptively diagnose issues remotely. This can help avoid requiring follow-up appointments, ensure the right technician and parts are available, and can make service appointments shorter. In addition, some appointments can be avoided altogether. For example, this can save a lot of time for everyone if the customer reports a problem that is going to be addressed via a forthcoming software update.

In mid 2022, Tesla took some steps to apply this approach to a specific kind of service request - where the customer reports an issue or concern regarding range. There were two key steps they took as part of this effort:

1. Enabling a thorough remote battery diagnostic check which could be performed and reviewed remotely by Tesla’s service department

2. Investing in owner education solutions to help customers understand the kinds of things I discussed in this article (such as the effects of Sentry Mode and Cabin Overhead Protection on idle power usage, the effects of speed and temperature, etc)

With these in place, Tesla can look into service requests related to range, battery drain, and battery degradation without the owner having to bring the car to a service center. If the diagnostic reveals a problem, they can set up a service appointment for the customer. If it does not, they can point the customer at educational resources to help them understand how and why their experienced range may differ from the EPA or “rated range” figure, as well as steps they can take to improve efficiency based on their usage.

Everything here is good for both customers and Tesla:

• The customer reporting the issue avoids an unnecessary trip to a service center

• Tesla avoids costs related to those avoided appointments

• Tesla reduces the load on busy service centers

• Other customers with real service needs don’t have to wait as long for an appointment

Somehow the Reuters article tries to spin this as a negative, and I just don’t see how that could be the case.

There’s one more reason I find the Reuters article to be problematic. They seem to imply that Tesla has received an unusually high number of such service requests. However, this is misleading for two reasons. First, no data is provided for the number of such requests, the trend over time, or a comparison to similar service requests made by customers of other EV manufacturers.

The second reason is perhaps even more important, because if the authors of the article had investigated this situation, they would certainly have learned a very important fact…

Customers use this even when they don’t have a problem

Last summer, Tesla enthusiasts discovered that they could perform a diagnostic battery test and retrieve a bunch of interesting information about the health of their battery by simply filing a service request citing “loss of range” as the reason - even if they hadn’t experienced any loss of range!

This “trick” was shared throughout a variety of Tesla and EV communities. For example, sites like DriveTeslaCanada posted entire articles about this:

To see if your vehicle’s range is within normal parameters, open your mobile app and tap on Service, and then ‘Request Service.’

Then tap on ‘Battery’, select ‘Range’, and type in ‘loss of range’ when asked to describe the problem. Tap ‘Next’ and the app will run a self-diagnosis to check on your battery health.

If everything is normal with your battery and range, the app will show ‘No issue detected’. The app then also provides a short description of how your vehicle’s range can be impacted by your personal driving habits and that fluctuation in your displayed range is normal and expected. (h/t: @tesla_adri)

Article: You can now check your Tesla's battery health through the mobile app (driveteslacanada.ca)

More such articles appear here, here, and here.

This trick was also shared widely on Reddit and Twitter (1, 2, 3, 4).

Hopefully you can see the problem with the claim in the Reuters piece (and related takes based on it). Whatever the number of these reports is, and whatever you think that number means, it’s clearly been inflated by owners who have no range problem just following these widely shared instructions to check their battery health.

What Tesla could do better

While I don’t think Tesla is doing anything dishonest here, I do think their choice to show battery level as a % of the EPA rated range can be confusing, and I have some suggestions for how they could improve that. These are, in order of ROI (in my estimation):

• Change the default energy indicator to “percentage” instead of “distance”

• Remove the “distance” option entirely and only show battery level as a percentage

• Rename “distance” to “rated range” in the Settings page to better clarify that it’s not a prediction

• Add an alternative “predicted range” option (though they may be hesitant to do this because predicting range is highly dependent on route and conditions, and building a good UX for this which isn’t more confusing could be tricky)

If you own a Tesla or buy one in the future, I suggest you change the Energy Display option to “percentage” and don’t look back.

What the industry could do better

Today’s published EPA range ratings are not nearly as helpful as they could be. In most cases, they don’t convey the information the buyer really cares about (e.g., how far can I go between stops on a road trip), and they aren’t even particularly useful as points of comparison across makes and models. They are better than nothing, but that’s a low bar.

Some manufacturers, like Porsche, go so far as to downrate their published EPA range (something the EPA permits manufactures to do - as Tesla did with the original RWD Model 3). Ostensibly this is to better communicate actual highway range to their customers. On one hand, that seems very considerate of them. On the other, it just muddies the waters (and may hurt their own sales!). Instead, I’d rather see a solution which better informs customers.

If it were up to me, I’d develop a replacement or supplement to the current “combined cycle” range rating system. Ideally this would be driven by the EPA, but manufacturers could also decide to agree on a solution on their own. In fact, Tesla (or any other manufacturer) could choose to lead here by developing their own range rating indicators and testing protocols. At least, I don’t think the EPA bars them from doing so, so long as they continue to also publish the standard EPA data.

Here’s a quick and dirty example of something I think would really help customers understand what to expect from an EV like the Model S w/ 21” wheels:

Sure, it’s more complicated than a single number to consider. However, the one-size-fits-all approach of the EPA’s combined cycle estimate just isn’t effective because of how much variability exists based on speed alone. While you could potentially simplify this to a single City vs Highway number (and ideally indicate the highway speed), I think having the 60/70/80 split really helps convey to drivers just how much of an effect speed has.

You could go further and include some kind of winter range loss indication here, ideally taking into account how the specific vehicle performs (e.g., due to having a heat pump versus a resistive heater). Perhaps a simple “range loss at 35F” percentage figure would suffice.

Some might worry that getting into these details and acknowledging these realities up front will have a negative effect on EV adoption. That may be true to an extent, but I think that’s shortsighted and that ultimately EV adoption benefits from buyers and potential buyers having a better understanding of these things.

Key takeaways

• EV range is complicated

• EVs usually get better mileage in city driving than highway driving

• Speed has a large impact on highway range

• You should not expect to achieve the EPA “combined cycle” rated range on a highway road trip, especially if driving 70+

• Tesla does not cheat in their EPA figures

• Tesla really does have the best efficiency and range available in mass production EVs, though EPA ratings may exaggerate the advantage in some cases

• Tesla’s rated range display is not a range prediction. It’s just an alternative to battery %, translated to a distance based on the EPA “combined cycle” rating

• Tesla’s actual prediction algorithm is really good, but it is exposed via the navigation UI and the Energy app, not the battery meter

• Addressing service requests via remote diagnostics is a good thing for everyone involved

• Service requests reporting “loss of range” is not a useful measure of customer satisfaction or vehicle problems, because many customers are following instructions on various web pages to submit such a service request even if they have no range problems, just to see their battery stats and proactively check for issues

• The EPA and/or auto industry should provide potential buyers and owners with more transparency and detail into what kind of real-world range can be expected in key real-world driving scenarios

If you’re in the market for an EV (or a new vehicle in general), I hope that the information I’ve provided here helps you make an informed decision, and better calibrates your expectations. I also hope that it does not deter you from considering an EV, as I really think for most people they are a great choice. This is particularly true if you can charge it at home - and a simple 110V AC outlet is sufficient in many cases.

If you think I missed anything, or have further questions, please reply in the comments below.