How does EPA test for fuel economy?

[PriusRos]

Can somebody please provide a simple description of the criteria that EPA uses for "highway" vs. "city"?

I went to the EPA website and all I could get out of it was that the new tests more closely represent real-world driving, using higher speeds and AC, etc. -- however, I could not find anything that describes the actual tests.

For example, how long are the routes, how long do they drive for, and at what speeds?

I'm particularly interested in understanding how they measure city driving. I don't know how they come up with higher numbers for the city than for highway. In my experience, city driving involves lower speeds, yes, but also shorter trips (sometimes barely enough for the car to warm up), stop and go and traffic lights and stop signs, stops and starts on hills, multiple short trips to run a series of errands -- such as driving for 5 minutes to the next store, stopping the car for 30-60 minutes, then starting all over again. None of this is condusive to high mpg.

[a priori]

Here's a great source of information, found in a thread started by cwerdna: Car and Driver: The Truth About EPA City / Highway MPG Estimates

[wvgasguy]

Originally Posted by PriusRos

Can somebody please provide a simple description of the criteria that EPA uses for "highway" vs. "city"?.

You definately want to read the Car and Driver article to see how it's changed and where it's headed in the future.

Don't think for a minute that manufacturers don't know whow to game the system. It's not cheating but it might be somewhat deceptive.

One classical example is that in the US we ALL want to know the 0 to 60 mph time of a car to determine it's quickness. If a manufacturer sets up the gearing and shift points to shift just after 60 mph they can pick up some pretty impressive time in a 0 to 60 test. However their 0 to 100 may be slower than a car they beat in the other test. Only a speed freak looks at the 0-100.

Additionally the car magazines sometimes test a car like you would not want to do if you owned it. When the Honda s2000 first came out it was considered a slug in early reviews. If you simply drop the clutch and floor it, a lot of non-sporty cars would beat it to 60, maybe even my Prius. However if you reved the engine to 6K and then dropped the clutch, as expected the results were significantly different. Do you drive like that? Probably not unless you were drag racing.

So if you want to know how it does when pulling out from a red light , you have to ask yourself, do you usually rev the engine and drop the clutch or do you try to do a gradual release while gradually increasing the throttle. Test drives can tell you how the car will respond to your driving style. Tests only give you a base line for comparisons between various cars when all others things are kept equal.

The EPA is simply a test with specified variables. Your driving style or normal route may have nothing in common with their test.

I'm averaging 46mpg and not pushing my car hard.

[Politburo]

This is what you are looking for.

Fuel Economy Test Schedules

[PriusRos]

Originally Posted by Politburo

This is what you are looking for.

Fuel Economy Test Schedules

Thank you -- this is exactly what I'm looking for. What I find most unrealistic about these tests is that they use approximately the same simulated distances for both the highway and city tests -- 10 miles and 12 minutes is not a very long test for highway but is 11 miles and 31 minutes is quite long for city -- giving it plenty of time to warm up. Also, I assume that the simulated drives are over flat terrain so there are no hills to deal with.

[bwilson4web]

The Europeans have a similar test although the speed protocol is different. But what I find interesting is the 10-11 mile duration used for these tests. This corresponds to what I've found to be the minimum distance for credible field data. This also means anything beyond this distance is nice but not really necessary. <grins>

Bob Wilson