THE uber-cool world of German automobile engineering suffered a rude shock on September 18 when Volkswagen conceded that it had used software patches in “defeat devices” to cheat on emissions by its diesel vehicles. More than 11 million of its cars are in jeopardy following the revelations, and the company, emblematic of German engineering excellence, faces up to $18 billion in fines by United States regulators for the violations; according to one estimate, the cost of the recall and fines could total €32 billion. As regulators across the world scramble to initiate action against the iconic brand and class action lawsuits loom large in several markets, the company faces its biggest crisis since its inception. Volkswagen has already sacrificed its CEO, Martin Winterkorn. An indication of the task at hand for the company is revealed by the new CEO Matthias Mueller’s admission that the recall process will be completed only by the end of 2016.
Two young researchers at the Center for Alternative Fuels, Engines and Emissions (CAFEE) at West Virginia University (WVU) conducted the study that has brought the world’s biggest car manufacturer to its knees. Chennai-born Arvind Thiruvengadam and his colleague Marc Besch, in association with CAFEE Director Dan Carder, conducted the tests on the vehicles in early 2013. After the team finalised its report in 2014, it drew the attention of the California Air Resources Board (CARB) and the U.S. Environmental Protection Agency (EPA) about the elevated levels of nitrogen oxide emissions from Volkswagen vehicles in actual on-road conditions.
Documents available in the public domain show that the CARB initiated moves with Volkswagen and asked it to replicate the tests done by the team at WVU and that Volkswagen failed to meet the norms. In December 2014, Volkswagen “voluntarily” recalled half a million cars in the U.S. and claimed to have fixed the problem.
However, the documents reveal that the CARB’s “confirmatory testing” in May 2015 proved that NOx (nitrogen oxides) emissions in on-road conditions were still at elevated levels. Further testing by the regulator revealed “uncontrolled emissions” by the Volkswagen cars. The scandal has revealed that Volkswagen is by no means the only manufacturer to cheat on emission controls. They have also revealed the laxity of standards in Europe, compared with those in the U.S., although cynics wonder whether the regulatory systems in the U.S. are designed to keep European competitors at bay.
Speaking online to Frontline from Sacramento, California, 32-year-old Arvind Thiruvengadam, who did his bachelor’s degree in mechanical engineering from the University of Madras, talked about how the team got interested in the project, the possible motivation for car manufacturers to cheat, the need for India to upgrade environmental standards and several other issues. Excerpts.
You have taken great pains to stress in the numerous interviews you have given the media that your findings were never meant to nail Volkswagen or any other specific manufacturer. This project was a relatively small one. What were the motivations for undertaking it?
We are a completely neutral research group. We are known to never take sides. Our conclusions are meant to fall out of the research we do. When this project came through, it was very interesting for three reasons. One, we were asked to test European diesel cars, which were just coming into the market, marketed as “clean diesel cars” and promising much better mileage. Volkswagen was one of them. The second reason was that we had always worked with trucks and heavy-duty vehicles. We thought this project would give us an opportunity to work with passenger cars, especially [those that are] diesel-run. Diesel-run passenger cars have not been popular in the U.S. Although the funding agency for the project, the International Council for Clean Transportation [ICCT], said all they could spare was $50,000, we realised the enormous value of the data that would come out of the project. The ICCT’s hypothesis was that these vehicles would come out extremely clean because they had been certified to meet the most stringent emission regulations in the U.S.. The ICCT was hoping to use the results of our study to put pressure on European regulators to enhance their standards. They would have argued thus: “If these companies can make cars cleaner for the U.S. market, why can't they do the same in Europe?”
The Request for Proposal was floated in December 2012. We received funding the following month. We first thought we would choose one car and drive across the U.S., from Morgantown to Los Angeles. We were interested in it purely as an academic subject. We hoped to climb over the Rockies, cover as many cities as possible, and test the vehicle in a diverse range of real-life situations. But we could not find a suitable vehicle and therefore decided to focus completely on California on the other side of the U.S. California is completely unique in the U.S. The vehicle population is very high; the rush hour traffic is world-renowned; diversity of terrain is available in the State; and it has the highest truck population in the U.S.
But, above all, California is one of the “non-attainment” States, those States that fail to meet the U.S. EPA’s air quality standards. These States have to meet the standards within a stipulated time or face economic sanctions. This is the reason why the emission regulations are the strictest, more stringent than those laid down for the U.S. as a whole. In fact, we often joke that California is a separate country!
How did you choose the three diesel cars manufactured in Europe for the study?
We actually ended up testing three vehicles—two Volkswagens, a Jetta and a Passat, and a BMW. But mind you, we did not choose the vehicles by the manufacturer but by the technology that was used in the vehicles. There are basically two technologies for emission control in diesel vehicles—one is the Lean NOx Trap [LNT], and the other is called Selective Catalytic Reduction [SCR]. These are also the most advanced technologies available for diesel cars. The Jetta was chosen for its use of LNT technology, while the other two use SCR. Mercedes was the first to start using the SCR technology. Many of the luxury vehicles, such as Audi and BMW, use SCR. We managed to hire the BMW from a car rental company that hires out luxury cars; we hired the car for $600 a day.
My Swiss colleague Marc Besch and I finally decided that we would test the cars from Los Angeles to Seattle and back, a distance of about 2,400 miles [3,862 kilometres], between February and May 2013. The most important aspect of the project was that the CARB said it was interested in baseline testing for these vehicles. That was probably the defining moment in this whole Volkswagen saga. The Board went out of its way, it let use its lab and other facilities. Every manufacturer wanting to sell a vehicle in the U.S. or in California has to send a vehicle to this same lab [located near Los Angeles] for testing. These three vehicles, too, went to the lab for certification, and they turned out all fine. But when we tested the same vehicles on the road, the results were completely different.
In real-world testing, there is never a right or a wrong because there are so many variables in real life. My own experience tells me that I can never, by merely looking at the data, conclude that the manufacturer is doing something wrong. It is also important to understand that manufacturers are not liable to meet emission norms in all conditions. For example, if a car is stuck in traffic for an hour and then starts moving, you cannot expect it to meet emission standards.
But what we found in the Volkswagens was astounding. In certification tests emission was really low, but on the road the cars emitted 30-40 times the maximum prescribed levels, which was alarming. We knew that driving at about 70 miles per hour would be the most optimal for the vehicle, in terms of both performance and emissions. But what we saw in these two cars was very different. We asked: “Why is the emission level never going down?”
What happens in a lab environment? I ask this because it is critical to understand how Volkswagen deployed its “defeat devices”.
In the lab, the vehicle is always driven on a set of rollers. In the lab certain aspects of the vehicle are to be turned off or on in order to ensure that the vehicle performs well on the rollers. For example, the anti-lock braking system and traction controls are disabled during a test. Steering controls may also be disabled during testing. Above all, the duration of the test itself is of a finite length of time, say 20 minutes. Maybe if you ran these vehicles in test conditions for an hour you would have seen emission levels shoot up. It is likely that defeat devices were triggered by the behaviour of these parameters.
What surprised us the most was the fact that a light-duty manufacturer was doing this. In 1998 seven heavy-duty truck manufacturers were caught doing the same thing and were fined $83.4 million by the EPA. [Seven major diesel engine manufacturers—Caterpillar Inc., Cummins Engine Company, Detroit Diesel Corporation, Mack Trucks, Renault, Navistar International Transportation Corporation and Volvo Truck Corporation—were caught cheating. The EPA found that these companies had sold 1.3 million heavy-duty diesel engines containing “defeat devices,” which allow an engine to pass the EPA emissions test, but then turn off emission controls during highway driving. The EPA found that these engines’ NOx emissions were up to three times the then prescribed limits].
Ironically, these manufacturers funded the WVU to the tune of $6-7 million to help design and develop the next generation portable emission systems. The greater irony lies in the fact that we used these systems on the Volkswagen cars.
Can you tell us something about the state of the art in your line of business—pollution caused by automobiles? How much of the issue is a result of the nature of the fuel, particularly diesel versus petrol? In the wake of the Volkswagen scandal, many people have described diesel as the villain of the piece. Or, do you think the issue is more complex than that?
If you consider the nature of the fuel, diesel is, relatively speaking, not as clean as petrol. This is because diesel is not as volatile as petrol is. Unlike in the case of a petrol engine, there is more air than fuel in the chamber of a diesel engine. Since air has about 75 per cent nitrogen, there is more NOx formation. There is also soot formation. That is the challenge in a diesel vehicle. But the state of the art, primarily developed because of the environmental regulations that have been pushed in the U.S., has resulted in diesel vehicles becoming extremely clean. Particulate filters trap the soot, store it and then burn it away when the chamber is full, which generates only carbon dioxide. The LNT system, which is of earlier vintage, is no longer considered as being effective in meeting the lower emission norms of today. The SCR uses ammonia [in the form of urea] to reduce NOx emissions. So vehicles using this technology have a separate tank for the liquid urea. The urea and the catalyst react with the exhaust gases to produce nitrogen and water.
The notion that diesel is bad is from a legacy technology point of view. A few years ago, the World Health Organisation [WHO] deemed diesel to be carcinogenic. What it failed to mention was the fact that it is the uncontrolled emissions of diesel vehicles that are truly carcinogenic. Thus, the WHO statement did not reflect the current state of technology. What happened then was very interesting. Subsequently, the Health Effects Institute in the U.S. tested the new-generation diesel technology and found there were absolutely no carcinogenic emissions from these vehicles.
But the WHO may be right if, as the Volkswagen episode has shown, companies may well be using outdated technology all over the world, except perhaps in the U.S.…
Exactly. My point was only that had the WHO made a more qualified statement about emissions by uncontrolled diesel engines, it would have pushed manufacturers to address the issue. Regulations ought to have been upgraded, which would have motivated the adoption of better technology. There is no point in making a fuel-based attack. That would not be a progressive approach towards the adoption of improved technologies, not a mindset that fosters innovation.
The Volkswagen Passat had an SCR system. What would have been the motivation to use a defeat device in this vehicle since it is equipped to meet low emission norms?
Every vehicle system involves a trade-off between performance in terms of what a car buyer expects from a car and emission controls. But my own experience shows that the average user would never be able to find the difference in performance between a car that is meeting lower emission targets and a car that is not. We never really felt any difference between when the Passat was working fine and when it was not working fine.
But when a car manufacturer develops a car, emissions are probably the lowest in the list of priorities. Customer perception, in terms of the vehicle’s fuel economy and durability, are far more important objectives for the manufacturer. They do not want to be facing repeated warranty claims. Their design constraints are based on the priorities they have in mind for the market. The rule of thumb in my line of business is that whenever you try to go for better emission controls, you are likely to take a hit on fuel economy in a vehicle. Most of the strategies to reduce emissions also require a higher fuel input into the vehicle. This is because every part of the system needs heat energy to work with. Without more heat none of the catalysts [used in emission control systems] is going to work. The only way to make them work is to put more fuel into the system.
In the case of SCR technology, urea is a consumable. The manufacturers probably think it is not a good idea to ask consumers to top up urea often. That would be construed as a marketing failure. In fact, the owner of the Passat we hired for the tests did not know there was a urea tank that could be accessed from the boot. In this case perhaps Volkswagen would have liked to coincide the urea top-up with the replacement of the engine oil so that the customer would not have to bother even knowing about it. But conservation of the urea becomes critical in this scenario. Urea injection is critical in SCR technology. If you stop urea injection, the SCR catalyst will not work. This could have motivated the manufacturer to cut off urea injection in non-testing conditions. When they needed to pass certification, they would aggressively deploy urea to cut emissions, but on the road they would try and conserve urea.
The other possible motivation could have been the manufacturer’s realisation that some parts of the vehicle’s system could have come under pressure when emission-control systems were deployed stringently. For instance, exhaust gas recirculation, which is another means of cutting emissions, also puts a lot of burden on the engine components. These may have raised durability concerns for the manufacturer. But the most popular theory doing the rounds is that the defeat devices were used to achieve greater fuel economy. It is as if the manufacturer said: “Once on the road do everything to get the fuel economy right, don’t worry about the emissions.” But our data do not support the popular theory. We did not see a significant difference in fuel economy between the lab and on the road to support this hypothesis.
Personally speaking, my surmise is that not only was the manufacturer targeting fuel economy, there must have been other underlying reasons, which only the manufacturer can explain. I am just as curious to know why Volkswagen did what it did.
Can you tell us about your journey to the WVU and your research interests there, starting from your doctoral work? What have been the elements of continuity in that journey, and in what ways has it been path-breaking?
When I finished my bachelor’s degree in mechanical engineering in India [University of Madras] in 2004, I was pretty clear that I wanted to work in the field of automotive engineering. But I was pretty raw and had no idea of what I wanted to specialise in. When I started the process of applying to universities in the U.S., I quickly found that the field of automotive engineering, at least in the way I had understood it to be, was outdated. But I found that there was a lot of interest in the field of engines and emissions. This branch is devoted to the development of more efficient engines and on how to make the vehicles clean up the exhausts that are generated. I got into West Virginia, but it was a big struggle to get into the research group; many students were already in line ahead of me.
We worked on numerous projects, which involved writing the proposals and arranging funding. This was pretty unique for a PhD student because not many doctoral students would have put their own research on the back burner in order to focus on work for a research group.
Can we talk a bit about your own doctoral work?
Well into my doctoral wor,my professor and I realised that I had worked on more than 20 projects but had not progressed on my own work for six years. I chose to work on emissions from vehicles running on natural gas and their impact on human health. I was primarily interested in providing an engineering viewpoint to the health effects.
I was interested in establishing the connection between the engineering and biological aspects of emissions. I evaluated the toxicity of natural gas exhausts. I mainly focussed on transit buses because they interact more with the population, unlike trucks on the highway. For instance, I was interested in how engines behaved or released emissions when buses pulled into a bus stop or when they accelerated away from one.
That is interesting, especially if your findings ran counter to the prevailing “consensus” that natural gas is a less polluting fuel.
Natural gas is just a white handkerchief trick. When you look at the tailpipe it looks comparatively clean; you don’t see visible smoke, unlike in a diesel vehicle. So the perception is that natural gas is clean. But for a natural gas vehicle to be clean, it has to be designed right. My own experience says that since 2012 natural gas vehicles in the U.S. have become truly clean.
But the technology used in most of the rest of the world is certainly not as clean. This is primarily because of the technology and the quality of natural gas that is used for vehicles in the U.S.
Natural gas is a very difficult fuel to burn in an engine; it is impossible to get complete combustion. When a fuel does not burn completely you have the formation of unwanted pollutants, most of which are carcinogenic. The EPA classifies them as toxic air contaminants; nearly 200 species of pollutants have been classified as such. We have found that if proper emission controls are not deployed, a large number of such chemicals are released by natural gas vehicles. So, the idea of phasing out diesel vehicles and replacing them with natural gas may end up causing a bigger problem, especially if suitable technology is not used to control emissions by natural gas vehicles.
When we talk about petrol vehicles, we talk about carbon dioxide emissions, when we talk of diesel we talk about nitrogen oxide emissions, but what are the problems posed by natural gas vehicles?
Natural gas vehicles release carbon monoxide. They also release hydrocarbons, otherwise called volatile organic compounds. These are basically unburnt or partially burnt fuel. Moreover, there is the problem of smog, which is caused by the mixing of hydrocarbons and NOx in the presence of sunlight.
Thus, by converting vehicles from other fuels to natural gas is not a good idea if you want to solve the problem of smog.
The first vehicles in the U.S. to start using natural gas came around 2005. When I compared emissions from vehicles of this vintage to the new ones in 2012, I found there was a 98 per cent reduction in all emissions. The constraints of the newer breed of gas-run vehicles are not the availability of clean technology. The constraints are more in terms of the range of these vehicles; they cannot operate over long distances without frequent refuelling. I am pretty sure that the natural gas technology in most of the world is similar to what was used in the U.S. around 2005.
Do we require a significant step-up in investment to adopt the newer, cleaner technologies for natural gas?
No, not necessarily. First, we have to see when the first class of such vehicles was introduced in a country. Then you need to have a programme for phasing out the vehicles and replacing them with new ones. As vehicles deteriorate, they behave worse and worse. And, if you do not have any active programme to monitor how these vehicles are behaving on the road, you are just hoping that they are behaving now as they did when they were introduced years earlier. Moreover, natural gas engines—and even those running on petrol—deteriorate pretty fast; in comparison, diesel engines are extremely robust, it is almost as if they never die.
Were the issues of your research ever conditioned from your time in India?
Whenever I have done my work here, I have personally been struck by the differences in standards here and in India. I am always struck by the fact that the emission standards in a relatively much more sparsely populated country like the U.S. are much more stringent compared with those in a densely populated country like India. But it is not just India; standards in China or even in Europe are, compared with the U.S., pretty obsolete. I always wonder about being able to work with organisations in India [and elsewhere] to help them solve problems there. Maybe what we have done on diesel cars would provide us such an opportunity.
India has a large, and rising, proportion of diesel vehicles in its automobile population. You have also spoken about the need to have independent test standards. Why is that important?
A standard should reflect the problems faced in a certain region. The reason why California has such stringent standards is that they have a problem of “non attainment”. India and China, for instance, face a totally different problem, when compared with Europe or the U.S. Both countries need to upgrade regulations to address the problems they face.
The fundamental thing is that standards should not be set from a vehicle standpoint. Instead, they ought to be set from an air quality standpoint. If you recognise the right of citizens to clean air, you would then set a norm for the air quality first, keeping in mind the health issues involved.
Then you [the regulator] would see the contribution of different categories of polluters to the overall emissions. The biggest source of pollutants are the mobile sources—cars, trucks, trains, etc. So, if you want a certain air quality, you would need to set targets for different categories of polluters, depending on how much they contribute to the overall problem. Currently, India adopts the European standards, but with a lag. When they go to Euro 6, India would move up to Euro 5. The point is that Europe does not have to be as aggressive in improving air quality as India or Thailand does. Of course, new regulations will induce technology adoption, but it will also result in the price of everything going up. But that is inevitable.