Saturday, September 19, 2015

Volkswagen, Diesel Emissions, and Regulatory Failure

Yesterday came the news out of the blue that the Environmental Protection Agency and the California Air Resources Board are charging Volkswagen with incorporating software in 2009-2015 diesel cars that enabled cheating on emission tests, particularly the emission of nitrogen oxides. According to the EPA, VW has admitted the use of a so-called “defeat device” in these cars. The EPA letter to VW states: “It became clear that CARB and the EPA would not approve certificates of conformity for VW's 2016 model year diesel vehicles until VW could adequately explain the anomalous emissions and ensure the agencies that the 2016 model year vehicles would not have similar issues. Only then did VW admit it had designed and installed a defeat device in these vehicles in the form of a sophisticated software algorithm that detected when a vehicle was undergoing emissions testing.”
The defeat device software was designed to sense when a car was being tested for emissions and would reduce emissions in order to pass the test. However, the degree of emission control used during the test is not applied when the car is being driven, and emissions are then significantly higher for nitrogen oxides and not in conformity with EPA or California requirements.

Though this is not spelled out by the agencies, presumably turning off some of the emission controls enables better performance and better fuel mileage. Approximately half-a-million cars may have this defeat device and will have to be recalled for a fix yet to be devised. (Disclosure: I own one of these cars.) There may be some reluctance among some owners to bring their car into a dealer for the fix, if the fix for the emission issue results in less power and worse fuel mileage. How the cars will be affected is not clear at this point. VW is under orders to devise a fix, but has not yet done so.

The EPA has the power to fine VW; press reports indicate that VW’s potential fine could be up to $18 billion, but most observers think it will be substantially less than that. Meanwhile, while the EPA and the CARB continue to investigate, the Justice Department is also investigating. Justice is presumably investigating whether any criminal charges should be brought.

West Virginia University and International Council on Clean Transportation (“ICCT”) researchers initially discovered the discrepancy between emissions in real-world driving and test results. (Update: More information has since appeared about this, which I summarize in the next post. The ICCT was not initially suspicious of VW, but was looking at the differences in the emissions between European and U.S. versions of the same cars because of questions that had arisen about the emissions of these cars in Europe.)
The West Virginia researchers may not have looked at the responsible software code. An interesting article by Alex Davies on the Wired website, “The EPA Opposes Rules That Could’ve Exposed VW’s Cheating,” explains that this likely would have violated the 1998 Digital Millennium Copyright Act, which is administered by the Copyright Office of the Library of Congress. According to the article, in December 2014, the Copyright Office was asked to grant exemptions from certain provisions of the Act for software used in cars, trucks, and agricultural machinery. The article states: “Having access to car controls would allow for ‘good-faith testing, identifying, disclosing, and fixing of malfunctions, security flaws, or vulnerabilities,’ [the exemption proponents] argued, according to comments they submitted to the Federal Register.”

The Alliance of Automobile Manufacturers opposed granting the exemptions, and the EPA opposed all the requested exemptions, but one, on which it did not take a position. The EPA was concerned that granting exemptions from prohibitions from examining the computer code would enable consumers to change the code in order to boost performance of their vehicles at the expense of higher emissions. The Copyright Office has not yet made a decision. The Wired article concludes:
The irony of the EPA’s concern over owners altering their vehicle code in a way that would violate the Clean Air Act is that VW was allegedly using its surreptitious algorithm to do exactly this—that is, to favor performance over fuel economy in a way that violated the Clean Air Act. And legalizing public access to the software used in the 482,000 VW cars now being recalled could possibly have revealed the alleged “defeat device” code earlier. As noted on Twitter by Thomas Dullien, a prominent security researcher and reverse engineer who goes by the handle Halvar Flake: “The VW case is an example why we need more liberal reverse engineering regulation. In a world controlled by code, RE creates transparency.”
Meanwhile, in Europe, where about half the cars are diesel, there has been concern that lab testing of automobile emissions is not providing accurate results. The European Commission plans to impose real-world emission testing requirements in 2017. There is some skepticism about whether the new testing regime will close the gap enough between test results and emissions produced by cars on the road.  

EU requirements for nitrogen oxides emissions are not as stringent as those in the U.S. Nevertheless, according to a February article in The Guardian, there are suspicions that auto manufacturers may be using “tricks” to pass the emission tests. The article does not address whether any of these suspected “tricks” are violations of law. At least some of them may be permitted loopholes. Regarding the current emission tests, the article states:
But the current ‘New European Drive Cycle’ laboratory test for measuring these emissions is a quarter of a century old, and has been outpaced by technological developments in the car industry. Studies have shown that lab techniques to measure car emissions can easily be gamed with techniques such as taping up doors and windows to minimise air resistance, driving on unrealistically smooth roads, and testing at improbably high temperatures.
Campaigners say that car makers also use tricks such as programming vehicles to go into a low emissions mode when their front wheels are spinning and their back wheels are stationary, as happens in such lab experiments.
Note that the programing trick the article refers to is similar to what VW has been accused of and admitted to doing in the U.S.  

The concern in Europe has recently been increased by a recent report claiming that only ten percent of new diesel cars meet current requirements. The concern in Europe about the health hazards of diesel cars has been building for some time. There are proposals to ban diesel cars in Paris in 2020, and the French government wants to phase out diesel cars. London is also considering restricting diesel cars, and Mayor Boris Johnson plans to impose a surcharge in addition to the congestion charge on diesel cars in 2020.
Clearly, there have been policy and regulatory failures with respect to diesel cars on both sides of the Atlantic. Europeans are reconsidering their move to diesel cars, and, in the U.S., I would think that the tax credits for diesel cars that were in place for a while to encourage diesel as an alternative automotive fuel will be viewed as a mistake. (I benefitted from that tax credit.)

Opponents of government regulation will no doubt jump at this experience to demonstrate how the government, with even the best of intentions, manages to do the wrong thing. That government policy was not well thought out in this area is clear. (I posted a comment in 2012 about how volume illusion was exaggerating the greater efficiency of diesel engines. A given volume of diesel weighs more than an equal volume of gasoline.) The inadequacy of testing and VW and perhaps other car companies apparently manipulation of test results are real failures of both the public and private sector. However, I don’t think that one can make an argument that without government regulation, automobile emissions would be less than they are now. They almost certainly would have been worse.
What VW has apparently done is appalling and that it took U.S. regulators this long – the cars in question date back to the 2009 model year – is not encouraging. VW of course will pay a price. Its dreams of becoming a major player in the U.S. market would seem to be shattered, and its bet on diesel cars is in question. It is not clear, though, whether other technologies than that used in the VW cars that are likely to be recalled and that are in use in some diesel cars in the U.S. are adequately reducing emissions. It is also possible that new cheaper and effective technologies can be developed without too much sacrifice.

Note: This post was updated in light of new information about the ICCT and West Virginia University research into this issue, which is discussed in the next post.

1 comment:

  1. Definitely Volkswagen has betrayed the trust of so many of its users. They should do something different to restore their brand value live science