EPA to change diesel tests to thwart VW-like cheating
Sep. 25, 2015
WASHINGTON (AP) — The Environmental Protection Agency said Friday that it will launch sweeping changes to the way it tests for diesel emissions after getting duped by clandestine software in Volkswagen cars for seven years.
In a letter to car manufacturers, the EPA said it will add on-road testing to its regimen, "using driving cycles and conditions that may reasonably be expected to be encountered in normal operation and use, for the purposes of investigating a potential defeat device" similar to the one used by Volkswagen.
The testing would be in addition to the standard emissions test cycles already in place, the EPA said.
VW's sophisticated software allowed its cars to pass tests in the lab and then spew pollution into the atmosphere while on the highway. The changes announced Friday are designed to detect software and other methods automakers might use to rig a test.
"We're actually making sure that this is a one-off," EPA Administrator Gina McCarthy said Friday.
The agency is going to "look at all of the other models aggressively and do the testing we need to make sure there aren't any hidden software devices or other ways they could defeat the emission system," McCarthy said.
The revelations about VW led to unwanted scrutiny for the EPA. Its testing procedures have been criticized for being predictable and outdated, making it relatively easy for VW to cheat.
EPA did not initially uncover the problem; researchers at West Virginia University did, using on-road testing that EPA did not.
Sen. Bill Nelson, D-Fla., said he was frustrated that regulatory agencies such as the EPA are failing to protect the public. "Seven years is way too long a time that the EPA has been asleep at the switch," he said.
The VW case has similarities to those involving General Motors ' defective ignition switches and Takata Corp.'s exploding air bag inflators, where it also took years before those problems were disclosed to consumers, Nelson said.
"When there is this kind of deception, we've got to get these agencies to be able to cut through it and catch it," he said.
Chris Grundler, head of the EPA's office of transportation and air quality, defended the agency's testing procedures, noting that passenger vehicles with diesel engines account for far less than 1 percent of overall vehicle emissions of nitrogen oxides and other pollutants.
"It's not a question of equipment or technology or capability. It's a question of where we deploy those resources," Grundler told reporters Friday.
The EPA has conducted on-road testing on heavy duty trucks, rather than passenger cars, "because that's where the emissions are," he said. The additional testing announced Friday is part of a "continuous evolution of our oversight" of new and used cars and trucks, Grundler said.
VW has admitted to installing so-called defeat devices on Volkswagen and Audi cars with four-cylinder diesel engines. The devices switch on pollution controls when the cars are being tested, but turn off the controls when the software determines that the cars are back on real roads. The EPA says about 500,000 U.S. cars including the Jetta, Golf, Beetle, Passat and Audi A3 have the cheating software, and VW says a total of 11 million cars have it worldwide.
VW was able to fool the EPA because the agency only tested the cars on treadmill-like devices called dynamometers and didn't use portable test equipment on real roads. The software in the cars' engine-control computers checked the speed, steering wheel position, air pressure and other factors to determine when dynamometer tests were under way. It then turned on pollution controls that reduced the output of nitrogen oxides that contribute to smog and other pollution, the EPA has said.
VW started the scheme with the 2009 model year, and may not have been caught without testing performed at West Virginia University on behalf of the International Council on Clean Transportation, a nonprofit group that advises governments on regulations. EPA and California regulators confronted VW with those findings to VW in May 2014. The automaker eventually did a recall late last year, without much improvement, the EPA says.
Only when the EPA and the California Air Resources Board refused to approve VW's 2016 diesel models for sale did the company admit earlier this month what it had done.
The EPA said the cars are safe to drive but VW will have to pay to recall and fix them. VW also faces billions in fines.
The European Union became aware in 2010 that lab testing was not as accurate as on-road tests might be. In the case of carbon dioxide emissions tests — not nitrogen oxide — the discrepancy was as much as 20 percent.
The EU will introduce new tests from next year, with on-road testing complementing the laboratory work that is currently being done. Lab tests will also be refined to try to reduce the margin of error.
It's not the first time the EPA has had to change testing to make sure automakers are playing by the rules. Earlier this year the agency updated gas mileage tests after some automakers were caught with inflated window sticker estimates.
Grundler disputes the notion that EPA would never have caught VW without the outside help. European regulators were looking into VW's on-road diesel emissions as far back as 2012, and since diesels make up half the cars there, the EPA decided to let Europe take the lead, he says.
"I don't think it's fair to say that this would never have been uncovered," he says. "Do I wish we had uncovered it sooner? Absolutely."
Krisher reported from Detroit. Lorne Cooks in Brussels contributed to this report.