When consumers buy a car, they take the EPA’s mileage sticker — such as 33 miles per gallon highway, 27 city — as a fair comparison with other vehicles.
But does the sticker always deserve that trust?
The United States is thought to have the toughest fuel economy standards and enforcement in the world. But that reputation suffered a black eye in November when the Environmental Protection Agency announced that Hyundai and Kia had overstated the fuel mileage for 900,000 vehicles sold in the U.S. by as much as 6 miles a gallon.
Hyundai Motor Group, which makes both brands, made the inflated mileage claims on 13 models of cars and SUVs for about two years, often boasting of the high mileage in ads. It since has apologized and decided on its own to compensate owners of the vehicles, setting aside more than $400 million to cover the cost.
The mileage restatement, the largest ever in the U.S., shocked observers and the auto industry. Other automakers haven’t been accused of making the same errors as Hyundai and Kia, but they’re worried that the overstatement could tarnish their mileage claims as well.
Environmental groups are concerned, too.
“You think the numbers on the mileage sticker are accurate,” said Jesse Prentice-Dunn, the author of a Sierra Club report on fuel economy standards, so the Hyundai-Kia errors were surprising.
The inflated claims raise questions about the country’s mileage standards at a time when they have never been more important. Because of higher gas prices, consumers are making fuel economy a priority. And the U.S. government is ramping up mandates for cars, SUVs, pickup trucks and vans that aim to nearly double their average fuel efficiency by 2025.
The overstatement has increased scrutiny of Hyundai and Kia, with some also left wondering about the diligence of the EPA. Some experts say a decision 20 years ago to stop checking one important test contributed to the problem. The EPA declined to discuss the matter.
EPA critics note the vehicles were sold for a year before the agency was prodded into retesting them. It took almost another year before the EPA’s November announcement.
Jamie Court, president of Consumer Watchdog, said the group asked the EPA to test the Hyundai Elantra in late 2011 because of consumer complaints about its mileage and eventually had to appeal to the White House to get action.
Hyundai’s ads had been taunting competitors about their supposedly better mileage — claiming, for instance, that it had four models getting an EPA-rated 40 miles per gallon on the highway. Hyundai reported record sales in 2011 and 2012.
Auto industry insiders say another automaker probably tipped off the EPA after testing some of the vehicles and finding the problem. Automakers routinely test competitors’ vehicles.
In 2011, Ford Motor Co. arranged a test of the company’s Focus and the Hyundai Elantra, challenging reporters to get the mileage on the window stickers. The Focus got better mileage in the test, even though its EPA-rated fuel economy was less than the Elantra’s.
John Mendel, executive vice president of sales for American Honda Motor Co., in remarks at the Automotive News World Congress last month, said he was concerned that the mileage overstatement would lower the trust consumers have in the industry.
“Competitive pressure should never ever lead us to betray the trust and faith of our customers or society at large,” he said.
What went wrong?
Although some details have been disclosed, exactly how Hyundai-Kia produced the inflated mileage claims is still basically a mystery.
Sung Hwan Cho, a Hyundai executive in the U.S., said in a recent conference call that honest procedural errors were made in what is called the “coastdown test.” To improve the efficiency of the test, Cho said, the automaker added a few steps and processes that were different from what the EPA recommended.
But Hyundai didn’t say what those steps were, and some fuel economy experts are skeptical that inadvertent mistakes caused the overstated mileage. Government requirements are specific about how to do the tests and are meant to prevent additional steps of an automaker’s choosing.
Critics want to know more about the mishandling of the test, in part to remove suspicions about Hyundai’s motives.
“It’s possible it was an honest error, but I don’t believe it,” said John German, a senior fellow at the International Council for Clean Transportation.
His career includes 13 years at the EPA and 20 years at Chrysler and American Honda. He is called to testify in Congress about fuel efficiency issues.
The coastdown plays a crucial role in what goes into fuel economy testing.
A test driver climbs into a vehicle and on a long, flat stretch of pavement hits the gas till the vehicle tops 70 miles an hour. Then he puts the vehicle in neutral and lets it coast. This is repeated several times in one direction and, to mitigate any effects of wind, several times in the other direction.
Equipment keeps track of the deceleration, providing data needed to help calculate the vehicle’s “road load.”
The road load is affected by the aerodynamics of the vehicle — how it slips through the air — and its ability to overcome the friction of components such as the transmission. It also includes the rolling resistance between tires and road.
The coastdown numbers are fed into a dynamometer, which simulates the vehicle’s road load when other fuel economy tests are run inside a laboratory.
Experts caution that the test can be manipulated. Parts such as outside mirrors can be removed to improve aerodynamics, the car’s weight can be lowered and tires can be overinflated. Those changes would reduce the road load and improve the fuel mileage.
Such tweaking is increasingly suspected in countries with weaker standards and enforcement than the U.S.
A study by the University of Graz in Austria found that the road load test was contributing to a growing disparity between the fuel economy results in the laboratory and what automobiles were actually getting in Europe, where consumer groups are starting to lobby for a change.
U.S. and European regulations differ substantially in regard to the coastdown test. In this country, the vehicles in the tests are required, down to the tires and other equipment, to be similar to what will be sold to the public.
But German, the fuel economy expert, said a decision made 20 years ago by the EPA might have finally caught up with it.
He said the EPA used to hire an independent laboratory to do coastdown retesting on up to 10 vehicles a year. That confirmed a sample of automaker results and sent a signal to the companies that there was oversight and a chance they would be caught if they did the tests incorrectly.
The agency, however, never found errors in its coastdown retestings, German said, so it stopped doing them. But once the EPA didn’t do any checking, it had to rely on the word of the car companies.
German said “even limited (retesting) would likely have kept the problem from happening” in the Hyundai tests.
The EPA turned down requests to discuss its monitoring of coastdown tests, even though it has been a year since it began to investigate the complaints about Hyundai. It also has rejected requests for documents, including one by The Star.
But there is a hint that the agency is concerned about the coastdown tests. At a conference in September, several months after an investigation into Hyundai began, the agency said it was evaluating the procedures for the test.
Consumer Watchdog is pressuring the EPA to say more about whether Hyundai and Kia made inadvertent mistakes or manipulated the test to give it bragging rights about selling more fuel efficient vehicles. They say disclosure would be important to maintaining trust in the EPA’s stickers.
“EPA needs to be more forthcoming,” Court said. “There is too much smoke for there not to be fire.”
Inside a testing lab
Government-required fuel economy tests began in the 1970s when the OPEC embargo spurred calls to reduce oil imports, in part by making motor vehicles use less fuel.
The tests are done in laboratories simulating highway and city driving under different conditions, including hot and cold weather. The automakers do their own tests, with Hyundai and Kia, for example, performing them in a lab in South Korea. The EPA audits, except for the coastdown test, 10 to 15 percent of the other tests in their laboratory in Ann Arbor, Mich., or contracts the retesting to independent laboratories.
The tests have been improved over the years, including adding one that looks at the effect of air conditioning. But the tests still fall short in some respects, experts say. For instance, the tests don’t use ethanol blended gas, which has less energy per gallon than pure gasoline, even though it’s the primary fuel sold in the U.S.
And no matter how good the tests, individual habits — from keeping tires inflated to going easy on the gas and brakes — still have a big effect on the actual mileage achieved outside the laboratory.
Nestled in the General Motors Proving Grounds, a 4,000-acre complex outside Detroit, sits an unusual room that helps the automaker figure its fuel mileage claims.
A 6-inch-thick insulated door closes behind the vehicle to be tested. Overhead, 22 heat lamps drop down to raise the temperature, and in front of the vehicle is a fan to simulate the airflow as if it were traveling on a road.
The idea is to re-create a summer day and measure the effect on fuel economy of the vehicle’s air conditioner. It’s one of five fuel economy tests performed at the lab to calculate a model’s fuel economy.
“It’s not a couple of people putting on an analyzer at the local gas station,” said Dave Garrett, director of emissions compliance and certification at GM.
GM agreed to give The Star a recent tour of its laboratory, which is housed in a building that does both emission and fuel economy testing. The same tests are used to reveal the amount of pollution emitted by a vehicle, as well as its fuel economy.
The testing protocol is precise, including how to fill the fuel tank. A hose sucks the fuel out, and the tank is filled to 40 percent of capacity. GM has a bay of fuel dispensers with the different formulas of gasoline and diesel used around the country. Two suppliers in the country provide the fuel, which has the same octane, sulfur and other components to ensure it’s the same as used by other automakers.
Temperature can affect fuel economy, so one test simulates city driving in cold weather. In the laboratory, that is done by putting a vehicle in a “soak” room for 12 to 36 hours at 20 degrees. A paper sign on the vehicle states when the soak began and warns not to start the engine.
The vehicle is then sent to a bay with large metal rollers in the floor that the tires go on. The dynamometer, which has been programmed with the road load, is connected.
A professional driver gets into the vehicle, and a video screen pivots in front of the windshield. In what looks like a trend line being drawn on the screen, the driver gets second-by-second instructions, including speed, acceleration and idling. The driver’s record in complying with the instructions is stored and available for an EPA audit.
The test instructions, again, are precise. In the cold-temperature city drive, the top speed is 56 mph and average speed is 21.2 mph. The car idles 18 percent of the time and stops 23 times. The simulated distance is 11 miles, and the time to do the test is 31.2 minutes.
“The tests provide a consistent basis for comparison,” Garrett said.
The U.S. Department of Energy says the tests are done in ways that can differ with how some vehicles are actually used. Cargo racks, which reduce fuel economy, aren’t used in the tests. The EPA test assumes vehicles operating on flat ground and with 350 pounds of cargo and passengers. If you’re towing a trailer, your mileage will be worse.
And numbers matter.
Ford’s C-Max Hybrid recently aced the test required by the EPA, getting 47 miles per gallon on the highway test. But Consumer Reports gave it 38 miles per gallon. The suspected issue is that the Ford battery can work up to 62 mph. But the magazine’s highway test is a steady 65 mph, which meant the gasoline engine was powering the car. The EPA’s highway tests go up to 80 mph, but they average under 62 mph. That meant the battery carried more of the load and improved the mileage in the EPA test.
Keeping consumer trust
It has been three months since Hyundai and Kia admitted to the inflated mileage claims.
John Krafcik, head of Hyundai Motor America, said in a statement: “Given the importance of fuel efficiency for all of us, we are extremely sorry for these errors. When we say to Hyundai owners, ‘We’ve got your back,’ that’s an assurance we don’t take lightly.”
Both Hyundai and Kia reported record January sales, suggesting plenty of consumers are willing to move on and forgive the company.
But disclosing exactly what went wrong in the Hyundai case would help keep consumers’ trust in the EPA mileage stickers, critics say. And there are lessons to learn, including whether there was a regulatory lapse.
“We should worry about that,” said German, the fuel economy expert.
Prentice-Dunn, the author of the Sierra Club report, said it’s important that the fuel economy tests remain prized by automakers and consumers, and that means automakers playing by the same rules.
“It’s got to be a level playing field,” he said.