EPA Expanded Testing: the 2012 Toyota Fleet

"CAFE the epitome of reasonable driving?" Hardly ! More like the epitome of carefully administered lab testing.
According to Aaron Robinson, writing for Car and Driver, only about 15% of cars sold are actually
verified by EPA testing. For the rest, the government relies on the automakers own testing.

"Sure, fuel economy labels bear the seal and certificate of the Environmental Protection Agency
and are based on a set of test criteria that runs to the hundreds of pages. New models are regularly
wheeled into the EPA's Michigan testing lab and run on dynamometers to hairsplitting regimens
of idling, acceleration and cruising meant to mimic the public's driving patterns."

"Complex math compensates for factors that are difficult to replicate on a laboratory dyno, such as
aerodynamic drag and rolling resistance. Exhaust gases are collected and analyzed, then a blanket reduction
is applied ( greater since revisions took effect in 2008, which include a new high speed, air conditioning-on,
and low temperature tests ), and pop out the mpg numbers."

The reality is that one set of window sticker numbers can't possibly account for all the real world variables
and be accurate for everyone. Can they be a guide as to what may be possible mileage wise ? Sure ! But
the EPA mileage ratings are far from the "epitome" of anything, much less reasonable driving.

David Friedman, deputy director of the Clean Vehicles Program at the Union of Concerned Scientists
has stated that a push to ditch today's testing and base labels on actual driving behavior gleaned from
GPS tracking and data recorders plugged into onboard diagnostic ports is still in it's infancy.
"Such technology promises to sharpen the relevancy of the labels. This is about getting good data
so we don;t have to debate the arcane intricacies of the testing system>" What might happen ?
"My gut feeling is that ( fuel economy on ) window stickers would go down , but I can't tell you
because we dont have the data to prove it yet."

CAFE might be a "target to shoot for"......for some....but it can also be a lot like chasing rainbows
in the real world considering the daily realities of real world driving for others.

 

Then provide it instead of a total nonsense rant !

There were no statements about "speeding to a red light"....or doing 65 in a 75........and the last time I looked.....
I sure didn't see the majority of drivers driving like "drunk idiots"......unlike the many Prius drivers I see
who are rolling roadblocks acting "holier than thou" attempting to navigate a vehicle when they should probably
get a horse and buggy instead. And.....you sound like one of them !

And yes....we have a Prius in the fleet....love it and it's mileage capabilities, but we sure don't drive to in a way
that requires other drivers to use their brake, accelerator, or steering wheel to avoid us.

Couldn't agree more about driving efficiently and safely along with consideration for others, especially
while behind the wheel. Your "rant' however, has absolutely nothing to do with CAFE, the EPA or any
provision of "data."

 

Maybe the sticker should also include a range for MPG's, given that driver and driving conditions have such a big effect. From posts on here, it appears that "real world" MPG's for the Gen III range from mid/low 40's to high 60's/low 70's (which as you point out is close to the CAFE numbers). That's from say 10% under the 50mpg combined EPA to roughly 40% above.

But the EPA and CAFE ratings were never intended to be "real world". Given that there is so much variance in "real world" numbers, that would be hopeless. They were solely for comparing different cars under the same defined lab conditions, to stamp out the wild mpg claims manufacturers were making.

Unfortunately they have come to be "what you will get" - an expectation - and have de-emphasized the impact of how you drive. People think that EPA is guaranteed and is what you will get no matter what you do.

Adding a range would emphasize that how and where you drive has a huge effect.

Personally the upper limit gives me something to shoot for. But it's really up to the driver to decide how far they want to go.

 

Gad! 72 mpg City and 70 mpg highway (unadjusted) What is it unadjusted for? Warm up time? Wind resistance? Gravitational resistance (hills)? I try my damndest to get good MPG's (drive 60 mph or slower in the right lane, don't tailgate, have a longish commute, little headwind on the way to work and something of a tailwind going home) and I can manage 75 mpg displayed on a round trip (71 mpg calculated). That is with a nearly ideal commute, in the warm weather months on 10% ethanol gasoline. Cold weather, winterblend gas, etc and you can neatly slice of 15% from that number....easy. Deviations from my commute (weekend driving, visiting family, errands) typically take my lovely displayed 72+ mpg into the mid to high 60's.

Is CAFE meant to be a theoretical value (like doing a physics problem where you ignore friction in your calculation)?

 

I'd call CAFE a practical near maximum when enviro conditions are benign.

OK, I think I can accept that as a reasonable definition.

 

You're right, the current one gives a MPG range, I never noticed that!
EPA Fuel Economy Ratings

Next year's will not show the MPG range to be expected:
New Fuel Economy and Environment Label

 

Theoretically the EPA mileage figure should be a tool for comparing vehicles, rather than considered an actual real-world number.
Witness the recent court case against Honda. Personally I think that finding against Honda because a particular woman did not get the same numbers is a farce. But then again, I am often amused by reality.

The real problem with the EPA numbers is that cars are so smart. "Some" manufacturers build in a program that recognizes when the car is being driven on the EPA cycle and optimizes the engine differently than when it is being driven normally. Naturally, the government regards this as cheating and has forbidden this behavior.

 

It was a farce, but the ruling was reversed on appeal. Sanity prevailed.

I'm still curious how she got those numbers. Her pattern of usage? Driving style? High rolling resistance replacement tires? I'd like to drive that car to see what is going on with it and what I would get.

 

CAFE is a farce, and more about the car manufactures lobbyists in helping them to continue selling vehicles that don't make sense in the "new world" of needed fuel efficiency. The EPA mpg ratings are the one to be used IMO. Better yet is to use the numbers that Fuelly tracks on all sorts of cars and trucks. These are real world examples...

 

The problems with the Fuelly numbers are:
- They are a selected sample of people who (a) actually care about mpg and (b) bother to post their data online. If we want numbers from the upper end of "real world" use they might be helpful.
- Fuelly is only one of many sites where people post mpg data (fueleconomy, cleanmpg, ecomodder, greenhybrid). Personally, I keep a spreadsheet on my computer (and also post the data to cleanmpg) because storing my data only on a private site could mean I would lose my data if the site closed down.

The problem with "real world" is that it encompasses a huge set of variables which gives a wide range of mpg results. So we can take the scientific approach, try to pin down as many variables as possible, and come up with a few numbers under defined conditions, as the EPA does. Which raises the question of how "relevant" those conditions are, but at least enables different vehicles to be compared under identical but possibly irrelevant conditions.

Or we can gather data from usage in many different conditions, and hope to come up with a range for each vehicle. But this data may be skewed. Usage patterns may differ between vehicles (convertibles might be used only in warmer weather, CUV's might be used for short kid-hauling trips, hybrids by people who care at least a bit about mpg, etc). And do we include outliers: the Prius being used in winter in Alaska on dirt roads, getting 40mpg, and the hypermiled Prius driven on long distance courier trips, never cooling off, getting say 75mpg? Plus gathering and collating all this data seems like as much work as running EPA tests, unless we simply accept the selected data from self-reporting on online sites.

 

True, but then the assumption must be made that this is equal across all makes and models posted on these online mpg tracking sites. With that being said, the data is valid to compare side by side. Perhaps it is not best to compare directly to the EPA since folks not driving for fuel economy are not posting numbers...but it is valid to compare car to car / make to make and to be quite honest that is what we are after anyway.

We all know what the Pri can do. That data just tells us what others with non-Pri's are achieving with the same general mindset which to me means the data is very valid.

 

I've very recently seen new cars sitting on the lot with the expected range and others that don't (w/the latter emphasizing combined mileage, just like the "new" sticker). I'm not sure what the rhyme or reason is.

Also, for those who haven't seen it yet, Car and Driver: The Truth About EPA City / Highway MPG Estimates | PriusChat should be insightful. Also, the EPA data can be downloaded from Download Fuel Economy Data.

 

Why? That assumption seems completely unjustifiable to me, absent data to support it.

I don't know about scientific illiteracy, maybe more like people demanding more of science than science can deliver. As I said, fuel economy is a complex function of a lot of factors. What science can do is to identify some of those factors and give some idea of their relative importance. But at the end of the day you're still left with a wide range of "real world" numbers that vary by at least 40% for a given vehicle (for the Prius, mid 40's to mid 60's). To make it worse, Vehicle A may do worse than vehicle B under some conditions but better under others (think diesel vs gas vs gas hybrid).

When people start talking about "real world" numbers, I wonder which of the many "real worlds" they are talking about.

 

On this note, I finally found out where the (now defunct?) "expected range" values come from/how they were derived, after an argument on another board and me doing some digging.

From page 80 of http://www.epa.gov/fedrgstr/EPA-AIR/2006/December/Day-27/a9749.pdf which comes from Fuel Economy and Environment Labels Regulations | OTAQ | US EPA.

 

EPA or not, two drivers driving the same vehicle will get different results. BUT, there is enough data from real world drivers in the case of the Prius and many other vehicles, that you can easily get EPA mileage without trying hard. Case in point. My good friend bought an '11 Prius III, same color as my '10 Prius IV - you would think we would get close to the same mileage since we live in the same area (1 mile apart actually) and we're in the same profession. As you can see from Fuelly label, I get slightly better than EPA rating, and I drive like I want to, which means I'm not a "rolling roadblock" as someone else put's it, I go with the flow of traffic, and frequently go 5-15 mph faster than the speed limit on local roads, mostly going with the flow. My avg speed ends up between 34-36 mph each tank, and my typical commute is about 22 miles with 16-18 stop signs/lights on 55 mph roads. But just drive around the block with my friend, and you know exactly why her mileage is never out of the mid-40's. She only knows "go" and "stop". She will never get EPA mileage in any car that she drives. Anyone who says "their" driving is real world and yet does not get close to EPA - what they are really saying is they won't, or refuse to drive properly, plain and simple. I have always averaged at the upper end of EPA mileage in every vehicle I've owned. It's not hard to do!