The diesel emissions scandal has already cost Volkswagen 25 billion euros, and no end is it sight. But how did it start? In a corrugated iron shack in the forests of West Virginia, discovered by a trio of university students.
Arvind, Hemanth and Marc actually only came to the United States to attend university. Arvind Thiruvengadam and Hemanth Kappanna are both from India, from Chennai and Bangalore, respectively, while Marc Besch is from Biel, Switzerland. They all ended up in West Virginia, not exactly the America you dream of when you come from Chennai or Bangalore. Probably not even when you come from Biel.
Attached to West Virginia University is an institute for emissions research – also, perhaps, not the field of study you dream of when you’re around 30 and aspiring to a career in auto engineering. The institute is called the Center for Alternative Fuels Engines and Emissions (CAFEE), located in an unprepossessing corrugated iron structure in a clearing in the hills of West Virginia. In other words, in the middle of nowhere. The nearest town is Morgantown, the nearest place you might have heard of, Pittsburgh.
This is where Arvind, Hemanth and Marc began measuring emissions. First truck emissions and then passenger cars – until they accidentally uncovered a scandal that brought the world’s biggest carmaker at the time to its knees. The emissions tests carried out by Arvind, Hemanth and Marc have already cost the Volkswagen company around 25 billion euros, mainly in buybacks, fines and settlements, and that is by no means the end of it. Because of a study written by these three students, former VW managers are wanted by the FBI, one has been arrested in the U.S. and others are in custody in Germany. One German politician called the diesel scandal “the biggest industrial scandal since World War II.”
All in a Day’s Work
Marc and Arvind park their cars in front of the hangar. Arvind arrives in a BMW with two huge exhaust pipes while Marc drives an old diesel jeep. Hemanth has come straight from the airport.
It’s late summer and the three engineers have come back to the place where it all began, here in the building in a forest clearing, to speak to DER SPIEGEL. It has been four years since they began testing the emission values of two VW diesel vehicles. They weren’t expecting much to come out of it.
Since then, their careers haven’t really advanced in any significant way. Marc, now 34, works as a research assistant. His sporty clothing makes it like he’s about to run up a glacier in the Swiss Alps. Arvind is also 34 and has attained the title of “research assistant professor.” In his dark suit and gray shirt, he looks less like an emissions specialist and more like a South Asian IT expert. Only Hemanth, a bearded 39-year-old, has switched sides and now works in industry. He has a job at General Motors in Detroit and the other two like to tease him about it.
At most, their work has afforded them a degree of fame on the emissions testing scene, but they’re not bothered by the modesty of their fame. “At conferences, they know us by name now,” says Marc. That’s something. Hemanth points out that the three of them are to be honored as exceptional former students that weekend at halftime in the football game between the West Virginia Mountaineers and Texas Tech. And Arvind remembers that they’ve been invited to a conference in Berlin in mid-October.
Thanks to the role he played in the diesel scandal, their boss, CAFEE director Dan Carder, was included in Time magazine’s list of the 100 Most Influential People in 2016. He’s also here for the reunion, but stays in the background because he doesn’t want to overshadow his former students. They all know that the VW scandal has tied their biographies together, probably forever. In the evening they’re planning to throw a few burgers on the barbecue then go for beers in a sports bar in Morgantown, a student hangout.
Culturally speaking, the two Indian engineers, Hemanth Kappanna and Arvind Thiruvengadam, could hardly have traveled further from home. In the U.S., West Virginia is associated mainly with forests, the heroin epidemic in its cities and its backwoods population, who refer to themselves as hillbillies. West Virginia is definitely not associated with environmental protection. “Rolling coal” is a popular hobby in these parts – it involves removing the particulate filters from a diesel pickup truck, driving to the nearest city, waiting until a Prius or a cyclist goes by, gunning the engine and roaring off, leaving a cloud of foul-smelling exhaust in its wake. That’s what passes for humor in West Virginia.
Gearheads, Not Idealists
Hemanth, Arvind and Marc likewise don’t do what they do for ideological or environmental reasons. They’re mechanical engineers. And there are plenty of interesting jobs for mechanical engineers in the car industry, from teasing a bit more horsepower out of engines to finessing the aerodynamics of a chassis to developing ultrapowerful turbochargers or working on e-car studies.
Or one is given the task of ensuring that emissions aren’t too damaging. That entails working with diesel particulate filters, catalytic converters, nitrogen oxide reduction and urea dosing systems. That’s what Hemanth, Arvind and Marc do. Deep in their hearts, though, they’re gearheads too. They love tinkering with cars. Marc, whose parents owned an Opel dealership in Switzerland, has souped up his own diesel jeep and, judging by appearances, its emission values probably aren’t great. Dan Carder also says he began restoring classic cars back when he was a teenager. Hemanth, Arvind and Marc never really intended to work in emissions technology.
The building in the hills of West Virginia bears the sign “Vehicle & Engine Testing Laboratory.” It belongs to the mechanical engineering faculty and is the nerve-center of the institute. Inside are a number of emissions testing systems. Metal rollers are built into the floor, on which the vehicles are placed so they can be driven without going anywhere. Huge fans simulate airstream, while tubes in all shapes and sizes lead to chrome-plated, cabled cupboard units fitted with measuring armature. A few engines are jacked up in the corners and there is an unmistakable smell of, well, of exhaust.
This is where Hemanth Kappanna was sitting one gloomy November day in 2012 when the International Council of Clean Transportation’s (ICCT) invitation to tender came in. The organization carries out technical and scientific analyses it then makes available to environmental watchdogs. This time, the ICCT was looking for a lab where the so-called clean diesel technology used in German car manufacturing could be tested.
Clean diesel? Really?
“The term ‘clean diesel’ sounded somewhat fascinating in our ears,” says Hemanth. “It came with a certain ambiguity.”
The Mobile Laboratory
Unlike in Germany, the Volkswagen Group of America ran a high-gloss marketing campaign making VW and Audi diesel cars out to be super-environmental. One famous Audi commercial, which ran during the world’s most expensive TV slot, the live broadcast of the Super Bowl, showed America being terrorized by “green police.” A shiny white diesel Audi breaks out of an “eco-check” roadblock and speeds off. The message was: The consumption and emission values of a VW or Audi diesel car were just as good as Toyota’s Prius hybrid, but with superior engine power and performance.
The biggest problem with diesel engines was not carbon emissions, as it is with gasoline engines, but nitrogen oxide (NOx). The U.S. permits much lower nitrogen oxide emissions than Europe does, but certain technologies made it possible to lower emissions to American NOx emissions standards. They, on the other hand, brought a different set of problems – more on them later.
Volkswagen was eager to overtake Toyota once and for all and become the No. 1 carmaker in the world, but to do so, the U.S. market was important. Toyota hybrids had proved extremely successful in North America, but the hope was that the clean diesel strategy could supersede them. The thinking was that if VW’s diesel cars had to abide by such strict emissions standards in the U.S., then the company might as well make a virtue of necessity and promote its turbo diesel engines as “green.”
“It’s not that there was suspicion. On the contrary,” says Arvind. “We were just very curious. These diesel emission technologies were quite new. And we knew they worked on paper. But they had never been tested in passenger cars under real driving conditions.”
As soon as he saw the ICCT’s call for tender, Hemanth knew that his lab had a good chance to be awarded the project. The institute had a lot of experience testing diesel engine emissions, even if it mostly come through working with trucks.
And it had another advantage – one that they couldn’t have known would play such a crucial role: The facility had portable measuring equipment, some of which had been built in-house, allowing the technicians to measure emissions in real-life driving conditions, even over longer distances. Usually, cars are tested in simulated conditions, running on rollers in the lab. Testing passenger cars on longer trips with portable measuring equipment is not standard practice. Hemanth didn’t even know how the equipment would fit in the car. This innovation – the testing of cars on the road over longer stretches rather than in the laboratory – was something that apparently no one at VW had reckoned with. It took a portable emissions-measuring system to expose the carmaker’s cheating.
Hemanth began doing the math. He worked out that a budget of $200,000 would be enough to test three or four German diesel cars and their emissions technology. The center was awarded the contract, but given a budget of just $70,000.
The German carmakers had two ways of lowering NOx emissions. One involved something called a Lean NOxTrap and the other relied on selective catalytic reduction (SCR). Both reduced NOx levels but also had some major drawbacks.
Research on the Cheap
It’s complicated stuff, but Hemanth, Arvind and Marc talk about such technologies like others chat about football. They interrupt each other and finish each other’s sentences. Ultimately, it is their fault that the world has been forced to come to grips with terms such as “AdBlue tank” or “defeat device,” terms that have appeared in countless newspaper articles and studies.
The way the trio explains it, the processes work a bit like this: With SCR, urea fluid is added to the exhaust fumes, breaking down NOx into oxygen and harmless nitrogen. The drawback: Cars need to be equipped with a tank for the urea solution, the AdBlue tank. This takes up space and needs to be refilled every few thousand miles. If you’ve paid $70,000 for a car, it’s likely not the kind of thing you’re eager to do.
The second method is less efficient but at least it doesn’t require an AdBlue tank. A NOx trap, or adsorber, chemically binds NOx during engine operation. When the adsorber capacity is saturated, diesel fuel is injected to purge it, with the NOx made to desorb into nitrogen. The drawback is that the technology only works at sufficiently high temperatures and entails a high fuel consumption.
Marc, Arvind and Hemanth felt that these methods posed a set of fascinating technical problems. How had the German carmakers solved them?
The contract with ICCT required them to test both technologies, SCR and the NOx trap. But where were they supposed to get the cars from? Arvind called Volkswagen in Michigan, but VW failed to see why it should make its cars available to a bunch of students. Car rental companies in West Virginia didn’t have any diesel vehicles in their fleets and no private car owners responded to ads they posted either.
The students eventually decided to relocate their operation to California. Marc and Arvind had previously worked there testing trucks and the West Virginia lab was on good terms with the California Air Resources Board (CARB), which defines vehicle emissions standards. Most importantly, though, they could find the cars they needed in California. Volkswagen’s “Clean Diesel” campaign had struck a nerve in the more liberal coastal states and, unlike in West Virginia, Californian car rental companies did include options for environmentally-aware customers.
The students rented a VW Jetta 2.0 TDI from Green Car Rentals, a BMW X5 3.0d from Exotic Cars San Jose and VW Passat 2.0 TDI from a private owner. The Jetta had an NOx trap, the Passat a SCR system and the X5 had both. The students wanted to test a Mercedes too, but the owner, one of those Hollywood types, suddenly wanted too much money. It seems likely that Mercedes owes him a debt of gratitude.
Taken for a Ride
When they went to pick up the Passat from its private owner, Marc Besch asked him what he thought of the urea tank – how often he had to refill it and if was an annoyance. The owner didn’t know what he was talking about. Refilling a urea tank? Never heard of it.
“That could have made us suspicious even before the testing started,” says Marc.
“But it didn’t,” says Hemanth.
They took the cars to El Monte near Los Angeles, where CARB has a test laboratory that’s surrounded by barbed wire. All three cars passed the emissions tests easily, with NOx emissions levels low enough to comply with the strict Californian standards. Such were the results under laboratory conditions: nothing out of the ordinary.
The road tests, which proved so momentous, were more complicated. The problems began with the power supply. The emissions-measuring device consumed a lot of electricity, but connecting them to the car’s own power supply would have interfered with the readings. There were battery-powered devices, but they tended to run out of juice after just an hour.
The students went to a hardware store and bought gasoline generators, which they screwed onto particle board panels in the backs of the cars. They channeled the generator exhaust through a tube out of the car window. It was now impossible to hold a conversation inside the car, because it was too loud, and the generators had to be refilled constantly. With the backseat of the car stuffed with measuring devices, one of the students sat in the passenger seat with a laptop and another drove.
To begin with, they went on a trip in and around Los Angeles lasting several hours: in heavy LA traffic, along highways and up Mount San Antonio. The nitrogen oxide output of the VW Jetta was up to 35-times higher than permitted levels. The Passat, which had the SCR system, fared a little better, but its NOx emissions were still 20-times higher than they should have been. Only the BMW’s emission levels showed no discrepancy.
Arvind, Hemanth and Marc initially viewed the aberrations as an interesting technical problem.
They started by dismantling their own equipment on the search for the problem. They then turned their attentions to the engines. After all, the excessive emissions levels weren’t the only curiosity in the results – the overall behavior of the two VWs was strange. Usually, harmful emissions drop as soon as the engine has warmed up. Not so with the VWs.
The students embarked on longer trips, to San Diego and San Francisco. Arvind and Marc even drove the Passat up the coast to Seattle and back, covering some 6,000 kilometers. With tubes attached to the exhaust pipe and the roar of the generator in the trunk, they were quite a sight motoring along I-5, the interstate between San Diego and Seattle. Outside San Francisco, Marc and Arvind were questioned by puzzled police officers about the tubes connected to the exhaust pipe after they had stopped at an exit to recalibrate their equipment. They spent the first of the road trip in the parking lot of a hardware store in Oregon desperately trying to get the generator running again. They stopped off at the Boeing museum south of Seattle and bought accessories from the souvenir store. In the evenings in the hotel room, Arvind often cooked dinner for the others.
Hemanth, the oldest of the three, would sometimes laugh at the others’ enthusiasm. Marc never wanted to stop working; Hemanth had never met anyone with a work ethic like his. Hemanth himself was convinced that the work would still be there the next day if they decided to knock off for the evening. Meanwhile, Arvind was a go-getter, always managing to do whatever he set out to do – he was the one who got hold of the three test cars, for example.
Yet at this point, their suspicions still hadn’t been aroused.
Mechanical Engineers, Not Detectives
They knew of the existence of defeat devices – hidden software that can detect testing conditions. In a lab test, the wheels will turn but the steering wheel doesn’t move. Speed picks up at regular intervals and a test usually only lasts 20 minutes. For this period of time, the software activates the emissions-reduction equipment before then turning it off again once the test is over.
A few years earlier, a number of major truck manufacturers had used defeat devices and been caught. The students, who had performed numerous tests on trucks, were well aware of this. Still, it didn’t occur to them that defeat devices might be behind the measurements they were getting. They were mechanical engineers, not detectives.
So they continued to look for answers elsewhere. The most probable explanation, in their view, was an unrecognized flaw in the exhaust system. Dan Carder, their supervisor, says: “To me, that seemed like the most probable cause. I thought Volkswagen might have to conduct a recall and it’s going to cost them some money.”
After a few weeks, they’d run out of things to test. They kept getting the same results. Hemanth, Arvind and Marc went home.
Their next task was to analyze the results and write them up. None of them felt like it. They hadn’t come up with an explanation and they had all moved on to other projects that seemed more interesting. They did nothing at all with the results for the time being.
Six months later, Marc finally sat down to write up the research. He came up with a 117-page report with the title “In-Use Emissions Testing of Light-Duty Diesel Vehicles in the United States.”
In March 2014 he presented it at a conference in San Diego. It was held at the Hyatt Regency and 200 people attended – executives from the oil industry and representatives from automobile manufacturers but also agents from regulatory bodies. Delegates from VW’s compliance department, who were responsible for ensuring that VW abided by American standards, were also there.
Marc didn’t think the presentation would make too many waves, he was mainly just relived he’d finished it on time. He didn’t actually identify the cars in the study, referring to them simply as Vehicle A, Vehicle B and Vehicle C.
After he’d delivered his paper, the audience began trying to guess which cars he’d been talking about. For car experts, it wasn’t hard to figure it out on the basis of the engine types and the emissions technologies in question. Vehicle A and Vehicle B were obviously Volkswagens.
The Rest is History
One member of the audience was Alberto Ayala, deputy head of CARB and a former faculty member at West Virginia University. He had known Dan Carder for 20 years and he was the man who had allowed the students to use the CARB laboratory in El Monte. He knew about the test results and he knew that Dan Carder’s students were reliable.
In winter 2013, even before Marc presented his paper in San Diego, Ayala had visited Wolfsburg and told people at VW about the strange numbers the students had found.
Once Marc Besch, Arvind Thiruvengadam and Dan Carder had left the conference and gone home, the ball started rolling. CARB put together a team of experts that began closely inspecting privately-owned VW diesel vehicles all over California.
The figures compiled by Arvind, Hemanth and Marc began to snowball into something entirely different – and much bigger. The emissions problems were now also a health issue – nitrogen oxide is linked to bronchitis, lung disease and heart problems. In bright sunlight, NOx emissions turn into smog.
NOx emissions are a serious problem in California, a state that’s home to a population of 39 million people and 25 million cars, and where the sun shines on most days. That’s why California’s emissions standards are so strict. And Ayala’s California Air Resources Board has made sure that, unlike in Germany, failure to comply with these standards will be investigated. Investigators started demanding answers from Volkswagen. The rest is history.
Back at the CAFEE testing facility in the hills of West Virginia, life continued after Marc Besch’s San Diego presentation as though nothing had happened. The investigation underway in California was confidential and for quite some time, the three students had no idea that their test results had sparked an extensive investigation into VW.
Hemanth, Arvind and Marc had moved on to polishing their portable emissions-measuring method. Since putting a generator in the back of a car hadn’t exactly been an ideal solution, they were now experimenting with installing the equipment on a trailer attached to the car. But there were problems with this approach too, from air turbulence to altered driving behavior and tricky trailer couplings.
The three would have liked to have continued working as a team. But then Hemanth went to Detroit. Marc also got an offer from General Motors, but chose to stay in research, and Arvind and Marc became increasingly close. As Marc says, they’re almost like brothers and they write all of their papers together.
On Sept. 18, 2015, when the United States Environmental Protection Agency revealed VW ‘s diesel dupe to the world, it had been awhile since Hemanth, Arvind and Marc had even thought about Volkswagen.
Moral outrage is entirely dependent on the standards one expects to be upheld. The three students were mainly glad that an explanation for their results had finally come to light – and they were relieved that, in the end, their figures had proven accurate. Hemanth, Arvind and Marc have high standards when it comes to their own research but they don’t necessarily hold Volkswagen to a high moral standard. The technology, they say, is actually pretty interesting.