Article: EPA to Fortify Certification Oversight in Wake of Mileage Controversies
 

It looks like the automakers are going to have fewer opportunities to make "mistakes" (Hyundai/Kia) or "game" (?) the tests (Ford hybrids) to give their cars better EPA fuel economy ratings than they deserve...

Source: EPA to Fortify Certification Oversight in Wake of Mileage Controversies | Politics content from WardsAuto

Update -- see also: http://ecomodder.com/forum/showthrea...ims-27224.html

Its nice to see the EPA addressing these issues and doing something about it. Wonder what those 30 people were doing that they won't be anymore though...

God forbid they just lay out an on road real world test track like NHTSA uses to for calculating tire tread wear and gather an actual sample size with multiple test subjects like real scientists do.

Even mythbusters was able to figure out 2 new identical vehicles don't get identical fuel economy even when driven together on the same road at the same time.

Nothing wrong with proper lab tests. The open road- as you ought to know- has so many variables it is nearly impossible to test with consistency. I thought the last generation of EPA test protocol was spot-on. Yes there is individual variation between "identical" vehicles; is it large enough to worry about and/or do you think a big enough number of each model should be tested?

The numbers they come up with are completely useless anyway. Fuelly is far more accurate at predicting real world fuel economy.

It is a standardized test.
 

The EPA testing protocol is to compare the economy of models to each other. We all know it has little to do with what each of us will actually get out in the real world. However, some manufacturers have been extensively gaming the tests to gain more marketing ammunition. Now it has come back to bite them.

Random auditing should be sufficient after the first test.

Real world is simply an unrealistic way to test with any consistency.

The current EPA may not be perfect, but it's the best "official" measure available at the moment.

Not, when you consider these two points: (1) all EPA tests are performend using 'neat' Indolene 91 octane gasoline, NOT the ethanol-diluted E10 stuff everybody is forced to use; and (2) some manufacturers stipulate milage tests be performed with daytime running lamps (DRL) and automatic headlamps and other "normal" electrical loads disconnected (which can only be done by pulling fuses!). So, how often do you pull fuses before you drive to work?

Also, up to now, EPA has NOT had a standard E85 fuel specified, so they relied upon numbers provided from manufacturers tests. They, now, however, have proposed an E85 standard fuel...but, it's not (yet) been approved.

Kia gives my wife about $90 a year since they had to eat crow and drop her Sorento's highway rating from 32 to 29. I still think they should give her more to cover the difference in used value since the lower mileage rating is a bargaining chip in buying-selling.

regards
Mech

How Vehicles Are Tested

Accounts vary, but links here https://www.google.com/#bav=on.2,or....ends+are+there claim up to 59 different blends of gasoline sold in the U.S.:

http://www.gao.gov/products/GAO-05-421

Several billion real world miles logged by owners would beg to differ.
Browse All Cars | Fuelly

I beg to differ.
 

If I am to make a purchase based on a baseline test, I am not going to wait for the manufacturer to be able to test for billions of miles. It is unreasonable and illogical based on the design and development time cycle. The original post was to an article that hoped to address the issues found when Brand X games their test so they can out do Brand Y by a great margin to provide marketing clout only to find the tested numbers are far outside the bell curve of normal driving resulting in "poor" consumer satisfaction as few can attain the test numbers. The test numbers should fall reasonably within the middle ground of real world numbers. Unfortunately, as we all know, the average driver has driving habits that return poor numbers in comparison to the EPA numbers. The problem becomes highlighted when few can even attain the EPA numbers.

The EPA standardized tests are fine just as long as the manufacturers adhere to guidelines. Ecomodder tricks such as special lubricant packages as well as higher coolant temperature levels help a "standard" vehicle do much better on the EPA cycle tests. More subtle tricks are re-flashed engine and drive-train ECU programs. The standard fuel is just that - a standard. The tests allow someone to compare vehicles. Fuelly.com does give me real world comparisons but not for the new model on my dealer lot.

It looks like the EPA will allot an extra number of people who will hopefully have the training to make sure the models of cars we buy are as reasonably represented by their tested counterparts, to a standard.

There are ways to game other tests. The NEDC is woeful in the way they blend electric and gas. The JC08 tests, I'm not sure, but I think you can choose your octane, which, if the vehicles are programmed for it, can make for better economy than with a set octane.

As for electrical loads... are DRLs required in all states? If not, then it's a valid reason to disconnect them for the tests, though if they're standard, they shouldn't be.

If all owners were the same and had the same routes, yes... the averages would average out. But for some vehicles, there is a self-selection bias in the kind of driver who will report on Fuelly. Not discounting the importance of the place, though. It's a very good resource.

A standardized test, if you close the loopholes and test over a wide enough set of driving conditions, should give you a fair representation of the difference between two cars in comparable conditions.

Mercedes caught by EPA audit - revises down 2 vehicles' ratings
 

Don't think FE is high on the list for Benz owners, but good to see EPA is checking.

BMW / Mini too...
 

BMW is the latest to be caught:

BMW dropping mileage ratings on ’14 Mini Cooper

Source: BMW dropping mileage ratings on

They're mostly being reduced by 1 mpg (eg. combined rating of the 3-door manual is going down from 34 to 33 mpg). The biggest change is the Cooper S 3-door manual, which is going from 38 to 34.

Ouch, 4 mpg drop!

If I'm honest, I'm having a real hard time matching the EPA numbers in the new car, versus the old car where 50% over EPA was an obtainable number. I don't know if my hypermiling skills has eased up or if the EPA Cycle #s are a little optimistic to begin with.

Still satisfied with the mileage, but just not blown away I guess. I've also read that the mileage improves as the engine/car breaks in as well. Just seems odd still.

My guess is that it will become increasingly difficult to obtain MPG much higher than EPA as cars become more efficient in general.

In an older manual gearbox car, I can select efficient gears, DFCO, cut the engine at stoplights, etc. A newer car might come with engine stop/start, CVT, and other efficiency improvements that allow less skilled drivers to get good economy.

With my 1997 5-speed Subaru Legacy, I was able to get about 30 MPG compared to the 24 MPG EPA rating. When I bought a 2007 Outback with automatic tranny, my hypermiling couldn't get it above 25 MPG.

I think some manufactures are trimming the fat in the EPA ratings they submit. Looked at the 2015 Subaru Impreza, AWD rated at 37 mpg HWY, pretty impressive. One Fuelly member of 63 is making it. Manual tranny is rated 33, 9 of 63 (14%) are matching or beating that.

By comparison 2014 Mirage 9/26 (35%) are getting the 44 hwy mpg the CVT is rated for or better, 11/26 (42%) are beating the 42 mpg hwy of the manual.

Not the best cars to choose to compare, I'd like to see how Outlander sport AWD with 2.0 would compare, I think it's rated 30 with CVT, but not much of a sample and hard to tell which are FWD or AWD. Be interesting even to compare Subaru's ratings of today to 4-5 years ago.

"Gaming" the EPA fuel economy tests will become all the more 'invisible' as we get closer to the EPA-CAFE-mandadted 55 MPG numbers.

real world
 

I think that the premise for the EPA testing comes down to repeatability.And a laboratory environment is probably the only environment in which one has a chance to cover every conceivable variable,really locking down the numbers.
If the car maker's testing follows the protocols,their results would be the same as if EPA Mobile Sources did the testing themselves.
And EPA has never suggested that what they put on a window sticker is what you'll realize in real world driving.It's just a guide.

I cannot imagine them doing this, but I wonder if they could make cars perform better on the EPA tests and worse when hypermiled.

Can you imagine the nerd rage that would ensue? :)

Oh, it looks like it has already started...

Take your pick:

EPA lies

EPA damned lies

Manufacturer lies (errors).

All the above.

You forgot statistics.

It's all about cycle-beating.

As greater and greater electronic control allow ever more specific engine tuning, manufacturers are gaming the test by making cars with very specific gear ratios, shift points and engine tuning, all designed to perform very well within the narrow performance window the EPA is test in.

Which means, drop your revs/speed or raise your revs/speed either way and you're out of that programmed "EPA Window" and into "regular driver" territory... where the engine is tuned more for power and reliability than ultra-lean cruising.

It's why more and more naturally aspirated cars are coming with a stupidly U-shaped torque curve. Gaming the EPA. The bottom of the U fits neatly into their planned acceleration profile.

Be nice if the EPA mandates economy over a wider range of situations. If you have to game the tests at ten different set speeds (say, from 30 mph out to 70 mph) and over three different acceleration rates... then the car is programmed to perform well for 90% of the people out there. That would be enough.

They should redesign the entire thing and just do a point system. You got a low .cd ? Have some points. A nice cvt in this year's model vs the old 4 speed auto? Points. Cut weight, redesigned mirrors. Ect.

This needs to stop my grandfather has a brand new Malibu a 4 cylinder and gets 18mpg. Epa says 23-29 he doesn't speed or drive crazy just normal lightfoot driving. Not a hypermiler but he should get decent mileage. He will upshift the automatic into 6th at 45mph. Still 18 mpg. Tires are good. Everything is good. The epa allowing cheating is not.

There has to be something wrong with that car.

I think dumbing it down to two numbers makes the whole system broken by design.

Oh, I don't know.

My mom, who literally drives like a grandma (because she is one), would get ~20 mpg US from the last 4 cyl Camrolla she had. Usage? Almost all short trips. The car probably never reached full operating temp the whole time she had it. I shudder to think what she got after Dad replaced it with the V6 model.

There are some cars that just can't be hypermiled. Just look at my sig, the Fiat is way off the EPA estimates despite being a bit of a Sunday driver and being driven in more favorable conditions than the rest of the fleet.

I don't hypermile my F150 and it beats that lil box?!?

I asked my grandfather and he said he does take short trips but took a trip out to west Virginia at 55 mph 75*F outside 42psi in the tires and only got 19.5mpg. And that's the best tank. Nothing is wrong with the car except it is thirsty and has strange shift points but like I said he will put it into manual mode and Upshift. I don't know what's wrong with it the dealer just tells him to drive efficient. And he does. .

It must be leaving a trail of raw fuel out the tailpipe. :eek:

Years ago Ex-Gf's mom had a V8 Dodge that ran poorly. I mean, she'd let it warm up for at least 20 minutes before going anywhere and I observed huge wet spots on the ground under the tailpipe after such starts. She'd brought the car to a local garage for service multiple times for the problem yet after what had to have been over $1000 worth of parts and labor, nothing changed. The entire ignition system had been replaced including computer. This went on for- I don't know- a year? It was a long time, part of it prior to my knowing them.

GF inherits the car and I finally decide to get to the bottom of this nonsense. Took all of an hour or so to deduce that it was flooding badly, to pop the lid off that nasty '80s 4bbl, pull the pair of floats out and chuck 'em into a pail of water, and watch one of them sink to the bottom like an anchor. Run to the parts store and I don't know- $7 later?- had the float replaced and the car running like a champ.

To this day I rarely miss an opportunity to badmouth that shop. What an elementary FAIL. It's such a ginormous fail that I think they were screwing their naive female customer on purpose. :mad:

The point? Oh yeah- gallons and gallons of gas straight out the tailpipe; must have been some into the crankcase too. Stupid thing must have gotten single-digit fe. :rolleyes:

.

Jack the car up and check if a brake may be dragging.

If you have a Ultragauge, check and see if it is coming out of open loop enrichment. The loop in the top right corner will be open when cold and close after a few miles of warming up. Using a ultra or scangauge will provide a lot of other info that may help identify the problem.

Other than that, I would keep taking it back to the dealer. Document everything. If the problem persists, tell them your going to get a attorney. In light of the EPA cracking down on manufactures false claims of mpg, they may offer some sort of compensation.

Remember, a squeaky wheel gets the oil... ;)

>

2015 Malibu is rated 36 hwy, I'd have to try hard and be burning E85 to get 19 mpg with my 3.5 Impala. Most people brag high about MPG, maybe he's the odd duck that goes the other way. Borrow it for a few hours and see how does for you.

Do these grandparents still start it 5 minutes ahead of taking a 2 mile trip to "let it warm up?" Something they probably got used to doing in their early carb days. Also the driving like a grandma, doesn't always turn out to be the best for economy. Taking 40 secs to accelerate slowly to 60 MPH doesn't get you into the most efficient acceleration. Hanging 8 seconds back from the car ahead doesn't give you any benefit of a draft. etc. etc.

Sub 20's doesn't sound right to me either. Every Malibu I have rented would easily do 30+ MPG in the hills of WV. Take the car out for a spin & see how it does for you.

Math is hard. Inventing statistics that support their argument is easier.

100% of all librarians think that I am a sentient macaroon.

100% of all librarians or 100% of sampled librarians?