4 ms·

It's great that Tesla keeps increasing the range, but the EPA numbers are ridiculous. I would love to see the EPA range claims for electric vehicles re-standard

It's great that Tesla keeps increasing the range, but the EPA numbers are ridiculous. I would love to see the EPA range claims for electric vehicles re-standardized to measure: * at the median driving speed on free-flowing US interstate highways * during cold weather * with some hills * using a medium level typical of drivers of cars of the price range A car manufacturer particularly focused on realism could publish a two range values; a lower number as described above, an upper EPA number (perfect conditions, low speed).

burlesona 6y ago
The EPA tests are interesting. They do simulate stop and go driving, and that’s why on a gas car the city number is worse, but the steadier highway test shows better numbers. One of the weird things with an EV is that the highway numbers are actually worse (as other posters have commented). The tests definitely trade off realism for consistent reproducibility though. This article describes it in detail, it’s really interesting: https://www.caranddriver.com/features/a15388892/the-truth-about-epa-city-highway-mpg-estimates/ https://www.caranddriver.com/features/a15388892/the-truth-ab...
- kenhwang 6y ago
  The test is absolutely insane and it totally makes sense why EPA estimates can vary so much from real world numbers. It would practically ignore the effects of aerodynamics and I'm not even sure weight plays a role given both of those are supposedly accounted for by a simulated drag on the rollers.
  - avs733 6y ago
    weight is inertia and aero drag is a speed dependant force...both are just forces that can be simulated really easily on a dyno.
- avs733 6y ago
  I was going to write something longer and then got distracted and this thread went bonkers, so sadly your post is not going to get the attention it deserves. I've actually toured a fuel economy lab for a Tier I/Big 3 auto manufacturer. The testing procedures described there are correct...account for as many variables as possible. The lab I toured had the dynos in temperature and humidity controlled rooms. Exhaust vents were standardized in length to ensure consistency, etc. MPG is serious business with millions on the line. Most interesting story I had was that they have several drivers they use for different purposes. Some are super consistent and used for repeatability tests. Others specialize in maximizing mileage. They quoted a 5% improvement by certain drivers over others on the SAME TEST. which already has incredibly precise time/speed windows that the drivers must stay in between. Every time, someone is manually controlling the throttle and brake pedal. Although they can use a computerized control for other testing. They will run the test until they get the best possible test and then ship that to the EPA.
octorian 6y ago
> the EPA numbers are ridiculous If you think that, you should look at the NEDC numbers they post in Europe :-) Okay, tried to dig those up and they've apparently switched to the WLTP standard there. That's good, because the NEDC numbers tended to be massively inflated.
gok 6y ago
The fundamental problem is that the EPA tests (and tests in the rest of the world) were really designed to give an estimate of efficiency, not range. For efficiency, they are actually pretty reasonable; the MPGe and watt-hours-per-kilometer actually give a reasonably good picture of typical driving, with a mix of high-speed/freeway and stop-n-go/city driving. But the range you care about in a battery EV is more about how much anxiety road trips induce, which is a much more complicated problem, involving variables like high speed charger availability.
alex_duf 6y ago
Renault does that, they publish winter and summer realistic range. They're pretty spot on.
gameswithgo 6y ago
That won't do any good, then pedantic people will complain that they can better the numbers all the time, and thus EPA numbers are ridiculous, just like you are now. People need to accept that these kinds of measures are useful approximations, and that there can be no perfect single number, and that this is ok.

burlesona 6y ago

The EPA tests are interesting. They do simulate stop and go driving, and that’s why on a gas car the city number is worse, but the steadier highway test shows better numbers. One of the weird things with an EV is that the highway numbers are actually worse (as other posters have commented). The tests definitely trade off realism for consistent reproducibility though. This article describes it in detail, it’s really interesting: https://www.caranddriver.com/features/a15388892/the-truth-about-epa-city-highway-mpg-estimates/ https://www.caranddriver.com/features/a15388892/the-truth-ab...

kenhwang 6y ago
The test is absolutely insane and it totally makes sense why EPA estimates can vary so much from real world numbers. It would practically ignore the effects of aerodynamics and I'm not even sure weight plays a role given both of those are supposedly accounted for by a simulated drag on the rollers.
- avs733 6y ago
  weight is inertia and aero drag is a speed dependant force...both are just forces that can be simulated really easily on a dyno.
avs733 6y ago
I was going to write something longer and then got distracted and this thread went bonkers, so sadly your post is not going to get the attention it deserves. I've actually toured a fuel economy lab for a Tier I/Big 3 auto manufacturer. The testing procedures described there are correct...account for as many variables as possible. The lab I toured had the dynos in temperature and humidity controlled rooms. Exhaust vents were standardized in length to ensure consistency, etc. MPG is serious business with millions on the line. Most interesting story I had was that they have several drivers they use for different purposes. Some are super consistent and used for repeatability tests. Others specialize in maximizing mileage. They quoted a 5% improvement by certain drivers over others on the SAME TEST. which already has incredibly precise time/speed windows that the drivers must stay in between. Every time, someone is manually controlling the throttle and brake pedal. Although they can use a computerized control for other testing. They will run the test until they get the best possible test and then ship that to the EPA.

octorian 6y ago

> the EPA numbers are ridiculous If you think that, you should look at the NEDC numbers they post in Europe :-) Okay, tried to dig those up and they've apparently switched to the WLTP standard there. That's good, because the NEDC numbers tended to be massively inflated.

gok 6y ago

The fundamental problem is that the EPA tests (and tests in the rest of the world) were really designed to give an estimate of efficiency, not range. For efficiency, they are actually pretty reasonable; the MPGe and watt-hours-per-kilometer actually give a reasonably good picture of typical driving, with a mix of high-speed/freeway and stop-n-go/city driving. But the range you care about in a battery EV is more about how much anxiety road trips induce, which is a much more complicated problem, involving variables like high speed charger availability.

alex_duf 6y ago

Renault does that, they publish winter and summer realistic range. They're pretty spot on.

gameswithgo 6y ago

That won't do any good, then pedantic people will complain that they can better the numbers all the time, and thus EPA numbers are ridiculous, just like you are now. People need to accept that these kinds of measures are useful approximations, and that there can be no perfect single number, and that this is ok.