Wednesday, September 23, 2015

The Fallout from Volkswagen's "Defeat Device"

  • The repercussions from VW's error in judgment seem likely to extend beyond the hit to their reputation and stock price, and the unnecessary extra pollution from these cars.
  • This incident will make a useful, fuel-saving alternative to gasoline cars less attractive, at least for now, resulting in higher future oil consumption and CO2 emissions.

I find the revelations concerning Volkswagen's reported efforts to circumvent vehicle emissions rules disturbing, especially as a VW owner and someone who has advocated diesel technology as a tool for reducing oil consumption and greenhouse gas emissions. VW has apparently admitted its colossal error. However, I haven't seen anyone attempt to explore the implicit emissions tradeoffs involved. As bad as this decision was, did it at least, on balance, help the environment?

The details that have emerged so far have focused on a software routine that manipulated diesel engine performance to produce one level of emissions in regulatory testing, presumably on a dynamometer, and different, much less acceptable results in real-world driving. Aside from the obvious questions about ethics and compliance, what did this mean for actual emissions?

For many years regulators have been tightening restrictions on allowable emissions of so-called criteria pollutants from cars. These include oxides of sulfur and nitrogen, particulates, and hydrocarbons, but not CO2. A whole gamut of technology was developed to tackle these pollutants, starting with catalytic converters on cars and deep desulfurization of fuels in refineries. Today's cars are much cleaner than those of a generation ago.

Oxides of nitrogen, referred to as NOx, are combustion byproducts that don't originate in a car's fuel, but from the nitrogen and oxygen in the air in which it is burned. NOx emissions from diesel engines have always been challenging, because they operate at higher temperatures and compression ratios than gasoline engines. Manufacturers that produce diesel vehicles have deployed different technologies to control NOx. As far as I know the VW Group uses at least two, depending on model.

Larger (and more expensive) vehicles appear to use a process called Selective Catalytic Reduction (SCR), in which small amounts of a liquid chemical such as urea chemically react with the NOx. The liquid must be refilled at service intervals. The technical manual for VW's 2-liter diesel engine involved in the current fiasco indicates it uses EGR, or exhaust-gas recirculation, which reduces the oxygen in the engine available to form NOx .

If controlling emissions from diesels is so challenging, why bother with them? Well, a typical diesel car uses up to a third less fuel than a comparably equipped gasoline model. After adjusting for the carbon content of the fuels, the lifecycle CO2 emissions are around 20% lower than for gasoline. Given the shortcomings of similarly priced electric vehicles in range and convenience, diesel provides a useful option. That helps explain why roughly half of European cars today are diesels, in many cases promoted by national fuel- and/or engine-tax policies.

That leads us to the question of whether such a reduction in CO2 might be worthwhile, even if it came at a penalty in NOx emissions, which act locally, rather than globally. To arrive at a ballpark answer let's assume that the 482,000 affected diesel cars couldn't have been sold at all if their engine software didn't fool emissions testers, and that the buyers would have otherwise chosen a comparable gasoline car. For comparison, the EPA rated the 2015 Jetta diesel at 36 miles per gallon (mpg) overall, while the 1.8 L turbo gasoline Jetta gets 30 mpg. At an average of 12,000 miles per year each, the collective annual fuel savings of the cars involved would be 32 million gallons, resulting in avoided CO2 emissions of about 300,000 metric tons per year, or 0.005% of US annual CO2 emissions.

If the tradeoff in extra NOx emissions is based on the reported maximum estimate of 40 times the EPA's allowed level of 0.07 grams per mile, then the affected cars would collectively emit an extra 16,000 metric tons of NOx per year. That's roughly 1% of the annual US NOx emissions tracked under the Clean Air Interstate and Acid Rain Program cap-and-trade markets in 2012. Even recognizing that those programs don't count all US NOx pollution, and that NOx and CO2 are very much apples and oranges in their environmental and health impacts, the relative proportions I calculated don't make this seem like a tradeoff worth making.

Whoever made the decision to circumvent the pollution controls on these cars did enormous damage to VW's brand and reputation. Unfortunately, the response in Europe and Asia suggests that this event has also raised questions about the emissions testing and compliance of the entire car fleet. Resolving them will take time and money, and if they are not seen to be dealt with properly, the impact on the public's trust of these processes on both sides--manufacturers and regulators--could be long-lasting.

Unlike in Europe, diesels made up just under 1% of new cars sold in the US last year. However, the technology was finally shedding the poor reputation that low-quality diesel cars earned in the 1980s, and the "take rate" was growing, along with the number of models offered.  VW's diesels are among the most affordable in the market. The NOx reduction technologies they use have been proven to work, when they are not circumvented, but that is not the message that this debacle will leave with the average consumer. Carmakers will have to work harder to convince buyers that this driver-friendly alternative to gasoline cars is worth a look, and that has implications for future oil consumption and CO2 emissions.


PelinoC said...


I find this whole mess very interesting on several levels: the cars are clearly able to run in a way that meets emissions tests (since the software makes them do it), but in that mode, what would they be like on the road? Would they be satisfactory to drive, much less competitive with other options available at similar prices? Would they go from being zippy and fun to drive "turbos" to "sewing machines" with sluggish acceleration? The VW engineers must have known what they were doing, and the tradeoff they were making... Will anyone be given the option to even find out what the tradeoff meant in the real world?

Also, if there is a concentration pattern to the sales of these cars, the local pollution impact could be even more severe: for example, if most VW diesels were popular in a few cities or regions (which is likely, since many cars have specific regional target areas), then in those areas the pollution impact could be much more significant than is suggested by your overall 1% calculation. Will there be some studies measuring this, and what will be the impact of taking these vehicles off the road (if it comes to that!). It would be a fascinating opportunity to gather pollution data (e.g., a specific, measurable one time change in pollutants, has what impact exactly on morbidity rates around asthma, allergies, etc.?),

More broadly, actually measuring the total fallout for VW is going to be interesting in and of itself, as a business story, rather than an energy or climate story: in many places around the world, the cars were subject to incentives, which would not otherwise have been offered. Will these be repaid? Many owners are filing suits claiming false advertising, asking for their money back, plus damages. To avoid total recall orders, VW might have to fully retrofit these vehicles with SCR systems costing thousands of dollars each... For VW, the costs of this fiasco could actually exceed their profit margins on the cars sold by many times. A classic corporate disaster.

Geoffrey Styles said...

Thanks for your insightful comment. You raise several excellent points, particularly about whether incentives must be refunded, and how the cars would drive in the "dyno" mode that meets specs. I've been wondering about that, too, though it's not an issue with my own VW. After looking at the economics, I couldn't convince myself that the economics of diesel made sense for me when I drive less than 5,000 miles per year.

As you say, if it's just software in the engine control module, one would think that a patch would get the EGR operating as designed, with NOx below the EPA limit. But what's the price in "pep" and miles per gallon?

Concentration is a key question for local pollutants like NOx, and I wasn't trying to get at real-world effects with my back-of-the-envelope CO2 vs. NOx comparison. I do see a lot of Golf and Jetta diesels in the DC metro area and the corridor up to NYC. But then I also see a lot of "blues", larger VWs, Audis and Mercedes using the urea system that isn't affected by this problem. Can't imagine a blue retrofit, though. From my conversations with the folks from Bosch at the DC Auto Show some years back it seemed there were technical reasons that the smaller cars used EGR instead.

Kate Winslt said...

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

Geoffrey Styles said...

Thanks for your comment and the link to LiveScience's article on diesel engines. Unfortunately, the quotes from one of the experts they interviewed don't pass muster. Characterizing gasoline as a "fairly uniform mix" and diesel fuel as "barely refined" reflects either gross oversimplification or information that is outdated.

In fact, each gasoline batch is slightly different, as a result of variations in crude oil inputs, refinery operations and seasonal specifications, while today's low-sulfur diesel often includes components that have been "cracked" from even longer, more complex molecules and then processed with hydrogen to remove most impurities.

For that matter, gasoline can also ignite under pressure (before the spark) if the octane rating of the fuel isn't sufficient for the compression ratio and other aspects of the engine's design and settings. This is called "knock" but is actually a form of "dieseling." The article is correct that this process takes place--by design rather than accident--at higher temperatures and pressures in a diesel engine, and is a key factor in diesel's higher untreated NOx emissions.