Tesla's EPA Driving Ranges Didn't Hold Up to Testing | Hacker News Reader

Tesla's EPA Driving Ranges Didn't Hold Up to Testing

thedrive.com

74 pointssamizdis85 comments

I think this is a bit misleading. Edmunds and the EPA ran two very different tests. For instance, Edmunds drives the car until it indicates it's empty. However, in the EPA test, "the vehicle is driven over successive city cycles until the battery becomes discharged" [1]. Tesla vehicles, like many others, have a not so insignificant buffer of "extra" miles that you can still use after the vehicle indicates that its "empty".

Furthermore, Edmunds tested the Teslas at temperatures that are roughly 14F lower than the 68F used by the EPA. This can make a big difference in the range you get.

My point is that these articles are acting like they are comparing apples to apples--the EPA test vs the Edmunds test--but they're not.

I admit that I'm a Tesla fan and shareholder. But my first loyalty is to the truth. Ding Tesla and Elon when they mess up. Nobody is above reproach or constructive criticism.

So let's not pretend that Edmunds (or The Drive for that matter) are either. Edmunds ran what appears to be a not very scientific test, and came to a highly debatable conclusion that Tesla vehicles don't meet the EPA mileage guidelines. I think the tests are not comparable. Maybe that's just me.

[1] https://www.fueleconomy.gov/feg/pdfs/EPA%20test%20procedure%...

There are only two temperatures it is valid to test range at: -50c, and +50c (120f). The extremes matter because the 1% of the time when you are in the extreme you need to know real numbers, not ideal ones. When the weather is potentially deadly if you run out of power you need to plan to ensure that you have enough range to safely make it to the recharge station.

Hopefully you never have to deal with the above extremes, but if you do the car better be there for you.

While this would certainly be good information I'm not sure they are the _only valid_ temps to test range at. Surely what would be most reasonable to report would be your day-to-day environment. That being said electric cars do seem to have their range affected by temp a lot more than IC cars, and as someone who lives in the Pacific Northwest I do feel like their ranges should be quoted regionally, though that does seem like a lot of work for something that was pretty easily researchable going in. That all being said, it would have been nice to know that my average in WA would be about 5% below the Southern California average.

Full disclosure - I own a model 3 and love it

That opinion is as extreme as the temperatures you post. Sure, let's ignore real-world scenarios for valid results. Only edge-cases need apply?

If the edge cases work everything else does too.

If edge cases fail that says nothing about whether the mainstream use case fare well or not.

I’m not a big Tesla fan because I dislike their lack of quality control. On the other hand I do like how they single handedly proved viability of electric cars and jump started the market. I am not a shareholder.

I agree the analysis was unfair to Tesla and bad science. Tesla gets hit by both extremes, people who love to bash them and people who love to defend them. Both extremes don’t bother questioning shoddy analysis if it proves their point. Yeah this is common with basically every topic, but with Tesla it’s extra visible.

This new test seems slightly better in that they appear to have run the two non Tesla’s at the same time, but it still doesn’t sound like measuring the same thing as the epa.

"Testing" implies the Edmunds report is in some way scientific while it is anything but. They drove the Tesla cars in temperatures as much as 20 degrees colder than the other cars. The EPA ranges are derived by testing in a controlled environment to ensure cars are treated in the same manner. There is variability in range when driving an EV, but this report is deeply skewed and unscientific.

Also, the 1st version of the Edmunds report didn't account for the buffer all batteries have while showing "zero range." Turns out the Teslas have more of that buffer. (Though Tesla itself is kinda coy about how big that buffer is, for obvious reasons.)

Yeah, running an EV battery completely dead can be catastrophic to the battery so they try to scare the driver into recharging early. I think the Model 3 LR has about 25 miles of buffer.

Hey man, you try getting headlines (clicks) with the truth all the time.

On the other hand, just seeing the headline I can't say I found it hard to believe - among the major car manufacturers, Tesla seems particularly prone to tweaking the performance numbers they put out. 0-60 figures measured from a rolling start instead of dead stop (that no one seems to be able to reproduce either way), hp figures that are a "peak system power" calculated by adding things up in weird ways and which can't be sustained for any length of time, range figures no one hits in the real world, ... they're very aggressive at this, and it'd be a mistake to think everyone does it this way - for example professional testers regularly exceed 0-60 figures Porsche releases officially.

I've seen a fair amount of professional testers/journalists reprimand them along the lines of "every manufacturer is guilty of this to some extent, but Tesla really ought to tone it down" (e.g. Alexander Bloch, who is generally very positive on the brand). As EV tech is commodifying swiftly and the competition closing in, I expect this to become a larger discussion.

> 0-60 figures measured from a rolling start instead of dead stop

That's pretty normal. One magazine (I don't remember if it is one of the car magazines or Consumer Reports) offers both numbers, from a true standstill and with a slightly rolling start. The latter is much more common, I believe.

Hagarty has a program with Jason Cammisa where this was recently debunked as well [0]. Many manufacturers do this.

[0] https://www.youtube.com/watch?v=JOWKJHvYK54

> range figures no one hits in the real world

I was curious so I looked...

Wikipedia[0] says the energy of a gallon of gas is 33.70 kWh and MPGe is:

    (total miles driven) * (energy of one gallon of gasoline) / (total energy of all fuels consumed)

My Tesla Model 3 LR RWD has a government rating[1] of 130 MPGe, the odometer[2] reads 26,328 mi and 6,727 kWh:

    26328 * 33.7 / 6727 = 131.89 MPGe

So I'm doing better than the rated MPGe. I don't hypermile either, my partner would kindly describe my driving as aggressive.

[0] https://en.wikipedia.org/wiki/Miles_per_gallon_gasoline_equi...

[1] https://www.fueleconomy.gov/feg/PowerSearch.do?action=noform...

[2] https://i.imgur.com/AACiz7d.jpg

Guessing you didn't read the article? This article is saying that when Tesla made those complaints, Edmunds reran the test the way that Tesla suggested and got the following results. Sure, maybe its not a big deal that the cars come up a few miles short of the advertised range under these conditions, but the Tesla ones were the _only_ ones to do so, which is at least interesting.

It would be a very interesting result if it were a controlled test in any way. Tell me, how do you test two cars in real-world driving conditions when one care goes 100 miles further than the other? The longer range car must go through more 'real world' than the other, right? Is Edmunds going on a 10 mile identical loop in identical weather again and again, or did they extend their existing circuit? I guess we'll never know.

Also, they "re-ran" the test? The went from real-world, to a test track, and then _only_ released the results of the 'buffer' range. Then they "re-ran" again but with just two of the cars. Then, they only shared the results for one of the cars they tested (noticeably missing is the data from the other car) calling it "validation" of their results.

You have to at least agree that not releasing the full data on their results is suspicious, no? It seems like they got caught with bad data and a hyperbolic headline ("Every Tesla we've tested has failed to hit its EPA range estimate") and then rushed to find any data that would help confirm their results. That's confirmation bias if I've ever seen it.

They drove one of the Teslas in a temperature 20 degrees below the warmest temperature they tested in, and that was the Tesla test that underperformed EPA range by the least. Most of the Teslas were tested in the same temperature range as all the other cars and still undershot EPA range.

Source: https://www.edmunds.com/car-news/electric-car-range-and-cons...

The range estimates my 2014 Tesla provides are, at best, very generous. It is frequently generous to the tune of 50 miles on a 200 mile range. Sometimes it's dead on.

Yes, I might not always be driving the most efficiently, but even when I set the cruise and drive on a flat, constant speed I can see swings of 40 miles. I'm generally a conservative, relaxed driver and rarely exceed the speed limit, happily just cruising along in the slow lane on the right.

I no longer trust Tesla to accurately measure mileage and assume the I'll get 70-80% of the car says I will.

My 2019 Model 3 is pretty good at estimating the range remaining, but it drops significantly overnight to the tune of 40-50 MPH in the dead of winter. That's not as big of a problem for me since I have a home charger, but it's definitely led to a few close calls getting home from work.

Have you considered asking your employer to install EV charging stations for employee use? There are metered solutions if they’re not comfortable providing the power for free.

A client installed ChargePoint chargers for myself (Model S) and their full time employees with EVs (Plug in Prius, Bolts, various Teslas) when I provided them with incentive information and helped them through the process gratis (EVs and electrical infra are my jam).

My employer refused to remove a wasp nest that had formed in the NOC room. I was told that $10/month for offsite backup was too expensive. There's no way they'd install a multi-thousand-dollar charger for one employee's car.

I wish you the best in your job search when the time comes.

> when the time comes

You underestimate my haste in the matter, my friend.

Who will feed the wasps?

Seems like the cars didn't fall that far short, but I think the implied grievance is that Tesla is systematically optimistic about the range of their cars where other automakers (Mini Cooper, Porsche, etc) make very conservative estimates. The "surprise" being that published specifications like driving range are really marketing tools, and some companies are more aggressive than others.

If I was a butcher selling you a pound of meat, and "surprisingly" was shorting you 10% on a consistent basis, would you say I was just doing more aggressive marketing than my competitors?

Different industries have different standards. Hard disks are several % under advertised capacity. Audio amplifiers ratings are completely disconnected from reality.

Packaged meat at places like Wal-Mart contain between 6 and 12% injected saline solution, so what you describe is already common practice.

EPA estimates are in no way equivalent to weight. lol. Bad analogy.

Edit: Yeah, I messed up my point. I was just responding to OP that getting less meat than what you paid for at a store is not equivalent to not getting the EPA rated mileage on your car. No one can really reproduce the EPA range without going to the EPA's test area.

In some ways they are. Run the test in one exact way and those are the results that you'd expect.

Run it differently and the results will vary in fairly predictable ways too.

Yeah, I messed up my point. I was just responding to OP that getting less meat than what you paid for at a store is equivalent to not getting the EPA rated mileage on your car. No one can really reproduce the EPA range without going to the EPA's test area.

To be fair, its easier than that. You can get a really close approximation just by using their speeds on a level closed track, ideally on a fair-weather day.

No. The analogy would be that every butcher sells meat by the lb, but some include an extra .1 lb, and others include an extra .05 lb.

And then a third party comes along and calls you deceptive because they measured yours at .9lb when measured on an airplane flying a parabolic trajectory, but the others still measured 1.0 lb.

Which one was doing good marketing? The ones that gave you what you paid for? The ones that gave you a little extra? Or the ones that got lots of clicks for running a different test? Maybe they all were...

Haha, or e.g. "Welcome to our gas station. We're optimistic that you're getting the number of gallons that you paid for."

But in reality I think the other responses get it right. I would say it's a marketing decision to publish the range in best case and worst case. In this case, the other automakers are publishing closer to the worst case or average case, while Tesla publishes best case. By comparison, the mass of meat doesn't really have a best or worst case (unless you market "our packaged meat weighs 1lb when travelling at 0.9c!!")

> I think the implied grievance is that Tesla is systematically optimistic about the range of their cars where other automakers (Mini Cooper, Porsche, etc) make very conservative estimates

Seems like it would be a rather better article had it included facts detailing that allegation rather than just implying it.

FWIW: I don't doubt that Tesla spins their numbers, but then automakers have been cheating on mileage statistics with testing conditions for decades. Is this really so different?

The big difference is that with electric car ranges it's even more important for the buyer. People care about that extra range so much, that it's in the name of the car type (long range).

The actual range numbers (like, the real value of how far the car can go between charges) are almost universally in Tesla's favor here, though. The article is about the delta between the advertisement and the result, not the literal value being delivered to the customer.

The question is how much is the difference. I really really like the Taycan 4S, but if it has a much worse range than a Model 3 LR, I go with the later one (Model S is not available in Europe until September :( ). Of course the final price matters as well. It's just that highway range is an important part of the comparision.

Yeah, I've been criticizing the article for that. Its missing the forest for the trees. The real range matters a great deal, as do other metrics like miles added per minute charging.

This report focuses on the least relevant bit.

Tesla Daily did an interview with one of the Edmunds people:

https://www.youtube.com/watch?v=rb1toreNTrU

Rob Mauer's early objections came from Edmunds knowingly introducing variations in conditions which would disadvantage the Teslas (like lower temperatures) but not being very clear about that.

Also, Edmunds originally did this thing, where they didn't account for the "zero remaining" buffer. Teslas put more of their capacity under the reading "zero remaining," to encourage low battery level drivers to seek charging sooner. Edmunds also had to be called out for that.

Temperatures in the 40s are not "low temperatures." It has been in the 40s at night in Los Angeles for the past month, and that is in one of the most favorable locations for EVs.

Tesla absolutely should be called out for giving range estimates that require unicorn temperatures. Edmunds testing more accurately reflects real world usage, aka, what a normal person would get if they weren't trying to artificially goose their measurements.

Yeah, but to do it properly the "real world" aspects need to be the same across all of the tested cars.

Edmunds tested all of the cars the same way: on roads, highways, and freeways in the LA area. It tested each car using the same settings for A/C, etc.

Tesla simply failed to live up to its marketing claims. Again. Even after Edmunds re-tested the vehicle to accomodate some of Tesla's requests. (And in fact, Edmunds notes that no Tesla has ever reached or exceeded its EPA rating in its tests, which is unique among all automakers, ICE or EV.)

In contrast, most of the other EVs outperformed their stated EPA ranges, and even with zero range buffers Tesla still underperformed its EPA ratings, while the other vehicles simply exceeded their EPA ratings even more once their zero-range buffers were included.

How does Edmunds conduct its testing? The roads

Edmunds drives on specific road routes that cover both highway and city driving around the greater Los Angeles area. We aim for a mix of 60% city driving and 40% highway, assuming that most electric vehicle owners will likely spend more time in stop-and-go traffic than they will on the open highway. Since no electric vehicle has exactly the same range, the route length is adapted to suit each vehicle. The methodology

In EPA tests, a vehicle is run in the default settings at startup. If there are more efficient drive modes available, or if you can increase the level of regenerative braking, but the vehicle doesn't default to these settings, they won't be utilized. Edmunds' standard practice is to use the most efficient drive mode as long as it doesn't affect safety or practical comfort levels, such as deactivating the climate control system or significantly reducing power for accelerating or maintaining appropriate highway speeds.

We run with windows up and the climate control set to auto at 72 degrees, and we maximize regenerative braking during stops. We follow the posted speed limits and keep within 5 mph of them, traffic and conditions permitting.

> Edmunds tested all of the cars the same way: on roads, highways, and freeways in the LA area. It tested each car using the same settings for A/C, etc.

That's... fairly generous to Edmunds. There was almost a 20 degree difference in temperatures between cars.

> How does Edmunds conduct its testing? The roads [...] driving around the greater Los Angeles area

Which can exaggerate the difference between cars. Suppose Car A has just enough battery to coast over the top of Angeles Forrest Highway and then regen down the other side. Car B rolls to a stop 1 ft from the top.

Most people would say these two cars have virtually the same range, but a test on hilly public roads could turn that into a 20+ miles difference in range.

They tested all of the cars in the same geographic location, following the same course on roads, highways, and freeways, ever time, during the same month. No other brand had the same systematic issue with overstated mileage despite the temperature differences, which affected all of the Teslas, not just the one driven during the coldest days (which were still no lower than 45 degrees F).

Most people would say these two cars have virtually the same range, but a test on hilly public roads could turn that into a 20+ miles difference in range.

That would have benefited Tesla, as their vehicles have larger batteries than their competitors. And yet...they systemically underperformed their EPA ratings.

Other comments mention that abusing the "zero remaining" buffer could damage the battery. On the one hand that doesn't sound ideal for a range someone may want to use more often, on the other hand the screen issue older Teslas had came with a maximum expected life of only seven years which software updates exaggerated down to three. So you can probably expect the car to fall apart long before the battery does.

The problem with these range tests is no one is driving 100% to 0%. People are more likely going 80% - 15% on road trips. Charging to 100% takes forever past 80% and getting under 20% starts making people nervous. Under 10% the car starts to cut power.

I'd like ranges also listed as the common range you'd expect. Maybe it can be standardized as an 80-20 test. For those on freeway driving I'm getting more like 150-200 miles on a M3LR.

Quick edit: I know some people might think 150-200 is low. Really could get more but I drive like it's a normal car and stopped caring about trying to optimize my efficiency. As long as I can get to the next supercharger that's all I care about.

Using the full battery (below zero) the 2021 Model 3 went 345+26 = 371 miles. Base price plus long range battery: $46,500.

The 2nd highest range among the cars tested was the Taycan at 323 miles (no adjustment for past zero was offered in the article). Base price for Taycan 4S: $104,000.

The story I'd like to understand is: Does the EPA actually publish results with one car underrated by 60% (Taycan 4S) and another overrated by 2% (2021 Model 3 LR)?

> The story I'd like to understand is: Does the EPA actually publish results with one car underrated by 60% (Taycan 4S) and another overrated by 2% (2021 Model 3 LR)?

My understanding is that in the USA: the manufacturers do the test and then report them to the EPA.

So the Taycan numbers were tested and reported by Porche. In contrast, the Tesla numbers were tested and reported by Tesla.

https://www.fueleconomy.gov/feg/how_tested.shtml

> Fuel economy is measured under controlled conditions in a laboratory using a series of tests specified by federal law. Manufacturers test their own vehicles—usually pre-production prototypes—and report the results to EPA. EPA reviews the results and confirms about 15%–20% of them through their own tests at the National Vehicles and Fuel Emissions Laboratory.

Note: EPA only reviews the tests, not conduct them. The manufacturers self-test.

or has certified labs test it.

Interestingly, the tests are required to use a real human driver in a controlled test cell (a dynamometer). The driver has to keep the speed between two narrowly spaced lines ( a min and max - see their video) for something like 20 minutes. There is still a 3-5% fuel economy between the 'bad' and 'good' drivers or at least there was in 2003 when I toured one of these labs as an undergrad.

Any theory on why Porche would under-report by such a huge margin?

Maybe:

1) They don't want customers to be disappointed, even when the conditions are far from optimal.

2) Volkswgen AG... diesel...

My 2021 Model Y hasn’t hit 325 miles. Ever. I drove 306 miles yesterday and I needed 47 minutes of >60kW charging and 9 minutes of <60kW at superchargers. I left home at 90%, and returned home with 23% battery, so I could have cut it closer. I wish Tesla SuperCharger history would tell me exactly the kWh per charge. And honestly, I wish the on-board miles remaining estimate were based on the energy used over the last 30 miles and not EPA rating.

If you want to dig more: this is just a re-hash of Edmunds article describing their test: https://www.edmunds.com/car-news/testing-teslas-range-anxiet...

Here's a Tesla Daily podcast where Rob Mauer interviews head of Edmunds about the test: https://www.youtube.com/watch?v=rb1toreNTrU

You can form your own opinion but Edmunds is roasted in the comments of this interview.

If you ask me Edmunds goes to great length to not report actual numbers. In the article they only tell you how many miles Teslas have driven after the car shows 0 range. This is to account for safety buffer.

They were criticized for an earlier test where they didn't account for this.

But instead of just telling what are the new numbers, they apparently want you to add number from earlier test (which was done in different conditions like outside temperature and driving style) and this delta from new test to arrive at total number.

It's almost as if the new test mostly invalidated their previous result but they're going out of their way to not admit that. Reeks of dishonesty.

I’d be curious to know how “range” is computed for ICE vehicles. Do you end when the gas gauge reaches “e” or until the engine stops? Seems to be perfectly analogous to the two possibilities with the Tesla. Either when the cars says it’s out, or when it actually stops.

Granted folks don’t seem to obsess over the range metric when it comes to ICE cars so the point may be moot.

Most cars keep a little gas under 0 miles left in order to preserve the fuel pump. If you run a car out of gas completely in a modern car, you can cause a lot of damage to the fuel system.

I was surprised how much of a buffer modern Mazdas keep. The "range" is based on 2 gallons hidden. At 3 gallons remaining, it appears as if you have 1 gallon remaining (i.e. 25-35 miles depending on your average gas mileage since the last reset.) So "0 miles remaining" happens when you really could go another 50-70 miles.

This is based on the total gas tank size compared to how much it takes to fill, but also because I'm stubborn and have pushed it several miles past "0" out of curiosity (but not regularly, as I have no wish to overheat my fuel pump or run out of gas!)

I have a Honda Element. The fuel light comes on when it still has about 3 or 4 gallons left.

People don't obsess over ICE range because it is hard for the average person to get more than 20 miles from a gas pump which can then fill the car to full in 5 minutes. As such the only time you run out of fuel is when you don't refuel despite multiple opportunities.

Today if you buy an electric car you are mostly fine for in town - the 95% of trips. However that last 5% is a real problem: there are not many places to charge, not all of them actually work, and even if you find a place you need to wait a fairly long time someplace where you don't wish to be for it to recharge. Some of this will change as electric cars become more popular, but for now if you are looking at an electric car it either needs to be a second car (use the ICE for long trips), or you need to obsess about range.

For electric cars with the least range the average commute in the US couldn't make it there, eat lunch, and back home on a cold day.

There is a pretty humorous video of some guys testing how far past empty an ICE vehicle can go. They were guessing 10-15km. It ended up being over 100! [1]

[1] https://www.youtube.com/watch?v=CpU72eM8vCo

Yep, this was not an issue with free charging previously. But with free charging gone now, the range will be noticed more and will become an issue. I also see lot of Musk fans believing him over other agencies when the discussion eventually happens though.

Edmunds direct link with data: https://www.edmunds.com/car-news/electric-car-range-and-cons...

Sounds like they were driving at consistent 65 mph. Where-as optimal speed for maximizing range is likely closer to 45-55mph.

They also ran the AC/heat with cabin set at 72F. With them doing the driving not on a track but real life (stop & starting).

Technically speaking the EPA numbers aren't necessarily "real life". Its a repeatable test environment: https://cleantechnica.com/2020/08/18/how-does-epa-calculate-...

EPA testing is a specific, defined drive cycle that is meant to reflect actual real-world use cases.

The goal isn't to maximize range in artificial conditions. The goal is to estimate how it might perform in standardized scenarios that can be repeated across different cars.

Yup, it's just an objective way to get a single number you can put on a window.

So I didn't see any actual numbers or precentages about how far the Tesla fell short, did it fall short by like 2% of 25% because that would be a very big difference which the article seems to go to great lengths to avoid. In fact there is only one vehicle that we get what the EPA estimates were for it and it's not a Tesla.

This seems to me like someone is going to great lengths to avoid putting in actual numbers and is instead just using this as a hit piece on Teslas.

The data is linked on the page: https://www.edmunds.com/car-news/electric-car-range-and-cons...

It seems like it falls just short, not anything worth writing about imo

Model 3 Long Range: 353 claimed, 345 actual for an example

Before people called them out on it, Edmunds only measured to when the Teslas displayed "zero range." Later, it turned out that the Teslas went further that the other EVs while reading "zero range."

The edmunds article is probably the best source, but this is the article that I saw yesterday with some of the numbers. Some are quite noticeable in their drops like the 2018 S Performance went from 310 mi to 256 mi. All the non-Teslas got higher than their EPA rating. With the Porsche Taycan 203 mi to 323 mi.

https://insideevs.com/news/496740/tesla-disputed-edmunds-tes...

That's an article too early. (Though they seem to address later objections in their edit.) It turns out that Teslas display more of their capacity under "zero range," giving more of a warning/buffer to drivers at low battery. The Teslas turned out to drive farther than the other cars when displaying "zero range."

https://www.youtube.com/watch?v=Wa_wf4EzLgY

The "zero range" tends to be a safety buffer for gas engines and electric motors alike and should not be included in the official range statements, imo.

The way I use my Model 3, I intend to almost never go below 20% and almost never go above 80%. I'm just fine with including the buffer in the stated range. Given how batteries actually work, I can understand Tesla's approach to the buffer, and in the long run, that's what's actually better for vehicle longevity, fleet-wide.

Looks like the numbers are here:

https://www.edmunds.com/car-news/electric-car-range-and-cons...

They range from -2% for the S to -17% for the 2018 Model 3.

> So I didn't see any actual numbers or precentages about how far the Tesla fell short

> This seems to me like someone is going to great lengths to avoid putting in actual numbers

The first sentence of the article includes a link to one of the tests, with graphs and comparisons to other vehicles: https://www.edmunds.com/car-news/electric-car-range-and-cons...

The rest of the article has a lot of specific numbers:

> For the other Teslas tested, however, things didn't go as "well." The 2018 Model 3 Performance was off by 54 miles—this might not look that notable on a couple-year-old model—but when the testing is performed in such friendly conditions that the 2020 Mini Cooper SE, with an EPA range of 110 miles, is exceeding that by 40 miles, then it just doesn't paint a pretty picture for the Tesla cars.

Has any car ever produced lived up to its advertised EPA range under regular use?

Yes, the Taycan does. Some of the other EVs are closer too.

From the article, the Porsche EV was well above its EPA range.

Yes. Ever since they changed the design to include using the AC, etc, they have or have been better than it for gas-powered cars under conscientious driving.

depends how you drive. if I try, I find it pretty easy to beat the epa highway rating on most cars. if you're not paying attention, it's also really easy to get poor efficiency out of the turbocharged low-displacement engines that are common today. efficiency tends to drop off pretty rapidly when you get into the turbo.

Has anyone ever gotten the EPA mpg estimates for their vehicle? I don't drive very aggressively, and I've never gotten even 80% of the mileage that the dealership sticker advertises.

I exceed the EPA highway MPG when using cruise control for 100s of miles w/ little traffic. When driving manually or being a bit more aggressive with passing, MPG falls below the MPG estimates.

Yes

I get almost exactly the EPA estimate on fleet of ~25yo cars, trucks and minivans. The car is almost always on the high end. I'm supposed to get 22. I get 21.something or 22.something or 23.something from week to week with 22 being most common. The vans and the trucks I don't drive enough to have a good feel for if they're less than an mpg high or low but they are definitely within +/-1mpg.

My commute is pretty much exclusively 70-90mph highway driving with approx four pulls (stoplights) from 0 to 50/70 (traffic speed in that area, depending on conditions) and am not shy about using all of 2nd gear.

I only approach EPA estimates when the tank nears empty and my right foot changes its behavior dramatically due to anxiety. I don't doubt that it is possible, but I find the kind of driving it requires to be unenjoyable. I don't have a lead foot and I enjoy spending less money to travel a given number of miles, but I do not often reach the optimal MPGs.

You may need to drive less in the cities (where gas engines are weakest), and be more conscious of braking and accelerating.

From what I understand, the EPA changed their measuring method many, many years ago, to include using the air conditioner, etc; and, from what I can see, mine is well within that range. In fact, I believe it regularly does better than its estimated EPA or right on it. Subaru WRX.

I had to look up the EPA mpg for my car, but I routinely get over the highway mileage (most of my driving is 400mi+ highway trips)--I got about 42-ish mpg and the highway mileage is 38mpg.

Reply on news.ycombinator.com