Edmunds: Here's Why Real-World MPG Doesn't Match EPA Ratings

My Prius generally gets a little worse than EPA. My V6 sedan maybe struggles to match the city numbers but beats the highway EPA by at least 10%.

It was funny reading the pass through article on the Elantra. One person said they get 18-19 mpg city, they must spend all darn day at the lights.

 

I average about 50% over EPA, so AFAIAC C&D are not telling the truth, and the EPA are fudging tests left and right.

 

I have a question: Regen braking, which the Prius, Leaf, and Volt use in moderation, and which the Tesla uses in large amounts, relies on the energy in the moving mass of the car itself. On a set of rollers, where the rollers mimic such factors as wind resistance, there is virtually no energy to recapture. How do they take that into account? Just by a formula? Or is it ignored?

(I apologize if this was covered in the article. I only read the first page.)

 

Bottom line, the EPA tests are a good estimate of fuel efficiency. I think people complain about them like they complain about taxes or the weather.

 

It is my understanding that the rollers on the dynamometer - which the vehicle wheels turn - have mass/inertia associated with them. Hence, when the vehicle "slows down" the inertia of the spinning rollers drive the wheels, instead of the opposite. So the moving mass is not the vehicle, but the spinning rollers.

 

TO clarify that a bit, the dynamometer are programmed to account for mass and wind based on the "coastdown" procedure described in
40 CFR 1066 : VEHICLE-TESTING PROCEDURES [ 40 CFR 1066 ] : (PROTECTION OF ENVIRONMENT [ 40 CFR ])

So there are at least coefficients estimated from vehicle data that go into the dynamometer setup. So the "slow down" is driving the wheels in a way that is consistent with the wind/drag/mass of the vehicle.

I will add, however, that the regen ability is probably only weakly modeled and maybe why there is such a significant variation in drivers for EV mode or maybe even PriusHV, compared to EPA. The tests have rather short/agressive stops that require more friction than regen. But a real work prius or EV driver can learn to better anticipate the stops. (An EPA driver must match the speed so cannot coast to a stop). I don't think EPA is being disingenuous, just basing the test on years of ICE data which does not translate as well to regen.


And the Volt, in L, does much more agressive regen than in D. I wonder which is used in EPA testing?

 

That makes sense to me, but I do not see why it is wrong. I just view the inertia as simulating the kinetic energy of the car.

More clearly, wind is not simulated, and cabin heating not at all.

 

Whether regen or friction, brake use does not increase the friction between the road and the tyre (for the most part.) The dyno is blissfully unaware, but the traction battery knows which was used.

 

Not sure what you mean by "wind". If you mean the flow of the air over the car as it moves, then that is address in the formula's for calibrating the dyno for that particular car. if you mean the car having a head-wind or tail wind then that is correct, the measurements presume relatively calm air. (Calibration must be done on days with forecast winds < 6mph).

 

Yep

 

I have had the following mpgs compared to EPAs:
2011 Prius 51 EPA city/ 48 highway - I get 50+ city/62.1 highway
2005 Dodge Neon SRT4 30 highway - I get 33 highway
1995 SOHC Plymouth Neon 34 highway - I get 46 after mods, 38 before
1984 Plymouth Horizon 24 highway - 24 (gear changed for drag racing)
All of the EPA ratings are according to their web site with older ratings reduced to modern equivilants.
The last two cars I had:
2001 Chrysler Voyager 22 highway - I got 24 bone stock and 26 with a homemade CAI
1996 Geo Metro 3 banger 44 highway - I got 46, then 48 after a tune up...a decent 60.1 after wt reduction of 140 lbs, headlight covers, and a grill block.

Of course all cars have the max air pressure in their tires.