Industry News: EPA: More Gas Mileage Audits Are Coming

[William Maley]

By William Maley

Staff Writer - CheersandGears.com

February 14, 2013

After the whole Hyundai-Kia mileage fiasco, the EPA says it will be conducting more fuel economy audits.

Jeff Alson, a senior EPA engineer said at a conference about the government’s 2025 fuel economy standard that the agency performs “routine audits” of manufacturer's mileage claims, but does admits more can be done.

“We’ll probably do more of that in the future than we did in the past,” Alson said.

Alson also told the conference that the EPA also needs to do a better job of communicating real-world mileage with customers.

“Everybody wants a label that tells you exactly what you’re going to get, but obviously that’s not possible. A good general rule of thumb is that real-world fuel economy is about 20 percent lower than the lab numbers.”

Source: Automotive News (Subscription Required)

William Maley is a staff writer for Cheers & Gears. He can be reached at [email protected] or you can follow him on twitter at @realmudmonster.

View full article

[motorman]

do they do the tests with E-10 as not using E-10 will give you more MPG.

[G. David Felt]

How about they actually do real work and testing rather than just write up documents and assume the companies actually test.

[dwightlooi]

Well, most drivers don't put E-10 in their car so why should anyone test with E-10 as a "standard"? As a matter of fact, using this fuel or that fuel is not particularly important. What is important is that everybody is using the same fuel during their tests.

The problem with EPA numbers is two fold.

The first being that EPA's defined cycles are not particularly reflective of typical driving habits of Americans (or any other market group for that matter). It is also completely separate from the standard used to determine CAFE contributions and/or European fuel consumption numbers. This has gotten slightly better over the years, but overall EPA numbers overstate mileage. IMHO, that is actually just fine as long as it is consistent. But it is frequently not.

The second being that the EPA does not actually test the vehicles. Many people think that manufacturers send their cars to the EPA and they get official mpg numbers in return for their window stickers. That cannot be further from the truth. The EPA defines the test cycles and does not actually do any testing. There isn't even an independent 3rd party (like a UL or whatever) doing the testing. The manufacturers do their own testing and report their own numbers. They are supposed to base their testing strictly on EPA's test guidelines. But as it turns out, some flat of do not and others "game" the system by testing with with environmental preconditions that are not representative of typical situations.

[dwightlooi]

The silly thing about EPA tests really, a very average dude can probably write a better test schedule that the EPA. I mean really... do you know just how stupid the procedures are?

Let me give you a few facts that you may not be aware of. Do you know that...

• EPA City and Highway test course requirements do not exceed 11 miles in TOTAL distance driven?
• EPA Highway test never took a vehicle above 60 mph at an average of speed of ~48 mph? (Who drives like that?)*
• EPA City and Hwy Fuel Economy test DO NOT require that the HVAC system be functional
• EPA tests do not specify how the transmission needs to behave or be operated during the test?

* Approx. 1/3 of the Highway test is driven between 40~50 mph, another 2/3rds is driven at 50~60 mph. You only kiss 60 mph three times during the entire 12~13 mins test and only for a couple of seconds. Not even the granny who always drive 5 mph under the speed limit on the right most lane on the freeway drives like that.

I can go on & on. But you get the picture. Let me give you an example of a "common" trick in gaming the system. You gear the vehicle to deliver the lowest brake specific fuel consumption in 6th gear at 48 mph. 48 is the magic number because it is the average speed of the highway test. The car never really has "acceptable driveability" in 6th at 48mph, and the transmission programming never really shifts into 6th at that speed. But, but the purpose of the test, you manually force it into 6th. Nevermind that in real life the transmission will be in 5th at that speed and does not go into 6th until 63 mph and the test never takes you above 60.

[ZL-1]

Independent 3rd party confimration should always be had with the EPA certifying the results, or the EPA should do the testing and certification of results itself.

[G. David Felt]

3rd party verification and real world driving should be required.

Highway, average speed for 30 min = 65mph

City driving, must include stop and go for 30 min, average 15mph for 15 min then 25mph for 15 min, then 10mph for 15 min.

This is just a quick example of what the real world testing should be on OEM programming. I know from chatting with a few friends, they do game the code that the engine,tranny run on to maximize fuel economy and then take the programming to the best average for reliability, etc. It is getting better that with modern electronics that the coding is not to far off between fuel testing and real world car, but then we would not have 3rd party programs that can give better HP/Torque if the OEM was not so wide spread in their coding of how the engine and tranny work.

I think we all agree that EPA needs to either start using 3rd party verification, standardize on fuel economy testing for world wide and if they do not want a 3rd party then test it themselves and stand behind their testing.

[dwightlooi]

3rd party verification and real world driving should be required.

Highway, average speed for 30 min = 65mph

City driving, must include stop and go for 30 min, average 15mph for 15 min then 25mph for 15 min, then 10mph for 15 min.

This is just a quick example of what the real world testing should be on OEM programming. I know from chatting with a few friends, they do game the code that the engine,tranny run on to maximize fuel economy and then take the programming to the best average for reliability, etc. It is getting better that with modern electronics that the coding is not to far off between fuel testing and real world car, but then we would not have 3rd party programs that can give better HP/Torque if the OEM was not so wide spread in their coding of how the engine and tranny work.

I think we all agree that EPA needs to either start using 3rd party verification, standardize on fuel economy testing for world wide and if they do not want a 3rd party then test it themselves and stand behind their testing.

LOL... the thing is that the programming is usually stock and the same as the delivered vehicles will have. However, the guidelines allow the operator to drive the vehicle as he deems safe and appropriate as long as he hits all the speed and acceleration benchmarks. Hence, it is for instance not a procedural violation to shift an auto manually for instance to put it in 6 when it normally is in 5, or to put it in M so as to prevent a kick down, as long as the consumer can also reproduce that series of actions. Whether they normally would is completely irrelevant. That's almost SOP!

More sophisticated "cheating" involves having programming that recognizes a certain sequence of events and "adapt" itself to an upcoming EPA test. For instance, you may program the transmission to go into EPA gaming mode if the car is driven for 15 mph for 30 secs, shut off for exactly 75 secs, started and driven at a constant 5 mph for 2 minutes, shut off for 10 secs then started for 60 secs and off again. The car then goes into mpg gaming mode for a given amount of time before reverting to normal again. The chances of that happening in exactly that sequence in real life is essentially like the odds of hitting the lottery jackpot. And it is not exactly catastrophic or even particularly unsafe if it did.

[Drew Dowdell]

Well, most drivers don't put E-10 in their car so why should anyone test with E-10 as a "standard"? As a matter of fact, using this fuel or that fuel is not particularly important. What is important is that everybody is using the same fuel during their tests.

The problem with EPA numbers is two fold.

The first being that EPA's defined cycles are not particularly reflective of typical driving habits of Americans (or any other market group for that matter). It is also completely separate from the standard used to determine CAFE contributions and/or European fuel consumption numbers. This has gotten slightly better over the years, but overall EPA numbers overstate mileage. IMHO, that is actually just fine as long as it is consistent. But it is frequently not.

The second being that the EPA does not actually test the vehicles. Many people think that manufacturers send their cars to the EPA and they get official mpg numbers in return for their window stickers. That cannot be further from the truth. The EPA defines the test cycles and does not actually do any testing. There isn't even an independent 3rd party (like a UL or whatever) doing the testing. The manufacturers do their own testing and report their own numbers. They are supposed to base their testing strictly on EPA's test guidelines. But as it turns out, some flat of do not and others "game" the system by testing with with environmental preconditions that are not representative of typical situations.

Beyond the fact that the defined cycles are not how anyone in the world drives, the problem isn't just that the EPA mileage is overstated (if often is), the problem is that it is frequently understated as well. There are very few cars that I've driven that have been exactly on the EPA mark (my '04 CTS was one of them, it got the EPA highway economy reliably on every trip and not an MPG more). Others have been well under (Our '04 CR-V that only meets its highway EPA rating going down hill, A/C off, with a tail wind). Others have been way over (300C regularly gets me 30mpg highway, Cruze Eco getting me 51mpg through the mountains, New New Beetle Turbo pulling into the mid-30s, Charger V6 pulling into the low-mid-30s, Passat TDI pulling well over 50+ mpg)

Some manufactures play to the test and are only that efficient at those testing conditions (I'm looking at you Hyundai and Honda), but when people drive them for real, they make the Vtec kicketh in yo a lot more often. Others, notably the domestics, seem to be more willing to give up a few ticks on the EPA sticker to have overall better real world driving results (Ford hybrid flap not withstanding).

[Drew Dowdell]

The silly thing about EPA tests really, a very average dude can probably write a better test schedule that the EPA. I mean really... do you know just how stupid the procedures are?

Let me give you a few facts that you may not be aware of. Do you know that...

• EPA City and Highway test course requirements do not exceed 11 miles in TOTAL distance driven?
• EPA Highway test never took a vehicle above 60 mph at an average of speed of ~48 mph? (Who drives like that?)*
• EPA City and Hwy Fuel Economy test DO NOT require that the HVAC system be functional
• EPA tests do not specify how the transmission needs to behave or be operated during the test?

* Approx. 1/3 of the Highway test is driven between 40~50 mph, another 2/3rds is driven at 50~60 mph. You only kiss 60 mph three times during the entire 12~13 mins test and only for a couple of seconds. Not even the granny who always drive 5 mph under the speed limit on the right most lane on the freeway drives like that.

I can go on & on. But you get the picture. Let me give you an example of a "common" trick in gaming the system. You gear the vehicle to deliver the lowest brake specific fuel consumption in 6th gear at 48 mph. 48 is the magic number because it is the average speed of the highway test. The car never really has "acceptable driveability" in 6th at 48mph, and the transmission programming never really shifts into 6th at that speed. But, but the purpose of the test, you manually force it into 6th. Nevermind that in real life the transmission will be in 5th at that speed and does not go into 6th until 63 mph and the test never takes you above 60.

Really? A number of vehicles I've driven go into 6th well below 63mph... the VW DSGs especially. Driven normally, the DSG won't let you go above 2,000 rpm until you've hit top gear, which puts the VWs into 6th somewhere in the high 30mph to low 40mph range.

[pow]

In addition to the city and highway tests, the EPA test cycles also include a high speed test (e.g., multi-lane freeway simulation with a top speed of 80 mph), as well as A/C use and cold temperatures.

The truth is that your mileage will vary. The amount of fuel consumed is determined by the amount of work your car does. Everyone drives differently on different routes in different conditions. You can't expect one number to represent all city driving, and one number to represent all highway driving. IMO, the point of these tests is for comparative purposes... to see how cars do relative to others under the same circumstances. Of course, designing for the test might not be helpful if your "real world" driving patterns are different from the test.

As an example of the wide variability in fuel consumption, the trip computer in my Focus is telling me that I'm averaging 17 MPG so far on this tank. This is because I've been doing purely city driving in the peak hour, where I spend more time idling than moving. Yet if I reset the trip computer on a flat freeway going 55 mph, I've been able to sustain 45+ MPG over several miles. The car is rated 26/36 MPG.

[G. David Felt]

Very True, my 2008 trailblazer SS averages 12mpg well below the estimated city/highway as it gets 75% of the time city and 25% highway since the wife drives it daily to college.

Yet on a road trip I took, I reset the MPG and averaged 28 with my Corvette V8.

[dwightlooi]

In addition to the city and highway tests, the EPA test cycles also include a high speed test (e.g., multi-lane freeway simulation with a top speed of 80 mph), as well as A/C use and cold temperatures.

But, that's not used for generating the EPA Highway mpg numbers. And, when was the last time you saw EPA "High Speed" mpg on the window?

[pow]

In addition to the city and highway tests, the EPA test cycles also include a high speed test (e.g., multi-lane freeway simulation with a top speed of 80 mph), as well as A/C use and cold temperatures.

But, that's not used for generating the EPA Highway mpg numbers. And, when was the last time you saw EPA "High Speed" mpg on the window?

It is. That's why the 2008+ EPA numbers are lower.

[regfootball]

Well, most drivers don't put E-10 in their car so why should anyone test with E-10 as a "standard"? As a matter of fact, using this fuel or that fuel is not particularly important. What is important is that everybody is using the same fuel during their tests.

The problem with EPA numbers is two fold.

The first being that EPA's defined cycles are not particularly reflective of typical driving habits of Americans (or any other market group for that matter). It is also completely separate from the standard used to determine CAFE contributions and/or European fuel consumption numbers. This has gotten slightly better over the years, but overall EPA numbers overstate mileage. IMHO, that is actually just fine as long as it is consistent. But it is frequently not.

The second being that the EPA does not actually test the vehicles. Many people think that manufacturers send their cars to the EPA and they get official mpg numbers in return for their window stickers. That cannot be further from the truth. The EPA defines the test cycles and does not actually do any testing. There isn't even an independent 3rd party (like a UL or whatever) doing the testing. The manufacturers do their own testing and report their own numbers. They are supposed to base their testing strictly on EPA's test guidelines. But as it turns out, some flat of do not and others "game" the system by testing with with environmental preconditions that are not representative of typical situations.

Here in MN e-10 is mandated in the mix, and they are trying to up that to E-15. All it means is for E-10, you get 10% less mpg. You basically are paying a 10% tax to fund the ethanol industry.

I would not mind having an E-85 car, i would occasionally buy E-85, but then it would be my choice. You can buy 'small engine gas' here in like 3% of the pumps, whenever I run that in my car, I get 10-15% better mpg instantaneously.

[regfootball]

maybe there would be a market for a company to do their own testing just to publish real world numbers and slander the epa numbers.....i bet that website would get a ton of hits and advertising.

The silly thing about EPA tests really, a very average dude can probably write a better test schedule that the EPA. I mean really... do you know just how stupid the procedures are?

Let me give you a few facts that you may not be aware of. Do you know that...

• EPA City and Highway test course requirements do not exceed 11 miles in TOTAL distance driven?
• EPA Highway test never took a vehicle above 60 mph at an average of speed of ~48 mph? (Who drives like that?)*
• EPA City and Hwy Fuel Economy test DO NOT require that the HVAC system be functional
• EPA tests do not specify how the transmission needs to behave or be operated during the test?

* Approx. 1/3 of the Highway test is driven between 40~50 mph, another 2/3rds is driven at 50~60 mph. You only kiss 60 mph three times during the entire 12~13 mins test and only for a couple of seconds. Not even the granny who always drive 5 mph under the speed limit on the right most lane on the freeway drives like that.

I can go on & on. But you get the picture. Let me give you an example of a "common" trick in gaming the system. You gear the vehicle to deliver the lowest brake specific fuel consumption in 6th gear at 48 mph. 48 is the magic number because it is the average speed of the highway test. The car never really has "acceptable driveability" in 6th at 48mph, and the transmission programming never really shifts into 6th at that speed. But, but the purpose of the test, you manually force it into 6th. Nevermind that in real life the transmission will be in 5th at that speed and does not go into 6th until 63 mph and the test never takes you above 60.

[motorman]

Well, most drivers don't put E-10 in their car so why should anyone test with E-10 as a "standard"? As a matter of fact, using this fuel or that fuel is not particularly important. What is important is that everybody is using the same fuel during their tests.

The problem with EPA numbers is two fold.

The first being that EPA's defined cycles are not particularly reflective of typical driving habits of Americans (or any other market group for that matter). It is also completely separate from the standard used to determine CAFE contributions and/or European fuel consumption numbers. This has gotten slightly better over the years, but overall EPA numbers overstate mileage. IMHO, that is actually just fine as long as it is consistent. But it is frequently not.

The second being that the EPA does not actually test the vehicles. Many people think that manufacturers send their cars to the EPA and they get official mpg numbers in return for their window stickers. That cannot be further from the truth. The EPA defines the test cycles and does not actually do any testing. There isn't even an independent 3rd party (like a UL or whatever) doing the testing. The manufacturers do their own testing and report their own numbers. They are supposed to base their testing strictly on EPA's test guidelines. But as it turns out, some flat of do not and others "game" the system by testing with with environmental preconditions that are not representative of typical situations.

most states have to use E-10 gsoline to meet the federal mandated emissions for their area. i know from personal experience that E-10 will drop the MPG about 2 MPG in the same car on the same trips

[dwightlooi]

most states have to use E-10 gsoline to meet the federal mandated emissions for their area. i know from personal experience that E-10 will drop the MPG about 2 MPG in the same car on the same trips

I was under the impression that most gasoline blends contain less than 10% ethanol and that 10% is the absolute limit permitted by the EPA without labeling it as an ethanol blend fuel. Typically, the actual blend is between 3% and 8% mostly as a convenient way to meet Octane ratings and oxygenated fuel requirements at the same time.

In any case, I find a reduction in fuel economy by a whopping 2 mpg highly dubious. The reason being that Ethanol has roughly 67% (2/3rds) the energy content of typical gasoline class distillates. A blend of 90% gasoline with 10% ethanol will yield a fuel with 96.7% the calorific value. If you average 30 mpg, the maximum you are going to lose due to the fuel blending is 1 mpg (3.3%). In reality it is probably less because fuels have never been 100% gasoline since it existed. Anti-detonants, detergents, etc typically make up about 1/30th of the volume anyway and these stuff typically don't burn to produce energy even as well as ethanol.

[regfootball]

your theory is nice, its not reality. my cobalt struggles to hit 30 average with the E-10 blend. the moment i fill an empty tank with pure gas i am pulling 32.5-33 no problem on average. my best readings in optimal scenarios with E-10 is about 37. I've reached 40 on my 43 mile jaunt (most at 70+ mph) with non E-10.

My taurusx same deal. Best i can usually pull on a trip is low twenties. With non E-10 in the tank i have been in the 24.5-25 range.

Other friends etc. have confirmed it (one friend pulls about 40k miles a year or more in company cars). 10-15% better mpg with non e-10 mandated swill crap.

Edited February 18, 2013 by regfootball