A diesel emissions test you can’t game? We try it out

One of the rather well-used VW Golf GTD test cars.

Jonathan Gitlin

One of the rather well-used VW Golf GTD test cars. Jonathan Gitlin

Andreas Kufferath, head of the powertrain technologies development department at Robert Bosch GmbH.

Jonathan Gitlin

Andreas Kufferath, head of the powertrain technologies development department at Robert Bosch GmbH. Jonathan Gitlin

A look at the Portable Emissions Measurement System fitted to the back of the diesel Golf. The bit that's mounted to the side of the car measures things like air temperature and pressure.

Jonathan Gitlin

A look at the Portable Emissions Measurement System fitted to the back of the diesel Golf. The bit that's mounted to the side of the car measures things like air temperature and pressure. Jonathan Gitlin

Andreas Kufferath, head of the powertrain technologies development department at Robert Bosch GmbH. Jonathan Gitlin

A look at the Portable Emissions Measurement System fitted to the back of the diesel Golf. The bit that's mounted to the side of the car measures things like air temperature and pressure. Jonathan Gitlin

The PEMS unit and the associated gubbins in the trunk.

Jonathan Gitlin

These boxes tell you what's coming out of the tailpipes.

Jonathan Gitlin

BOXBERG, Germany—Diesel is a dirty word in the auto industry these days. The fuel was once viewed—particularly in Europe—as a potential savior, since diesel engines offer much better fuel economy and emit much less CO₂ per mile than engines that run on gasoline. But that changed once Volkswagen Group was caught cheating its emissions tests, resulting in billions of dollars of fines and a loss of public trust. Automotive-component maker Bosch had a hand in the mess, too—it provided the code on the Engine Control Units in VW Group's offending diesels. The supplier ultimately paid out several hundred million dollars in settlement in the US, although it was not required to admit any wrongdoing.

But the fallout from the scandal has been more wide-ranging, as the artificial nature of emissions testing came under the spotlight. As a result, from September this year, the European Union is adding a Real Driving Emissions (RDE) test to the Euro 6 regulations that govern what comes out of a car's exhaust. It's a test that should be extremely difficult to game, and on our visit to the company's proving grounds in Germany for the Bosch Mobility Experience (a technology showcase put on by the German engineering firm), we had a go at a slightly abbreviated version of the test ourselves.

Perhaps surprisingly for a company with a central role in the VW scandal, Bosch is a big fan of the RDE test, according to Andreas Kufferath, head of the powertrain technologies development department at Robert Bosch GmbH. Kufferath thinks diesel still has a valuable role to play and that, even with existing technology, it's possible for diesel engines and air quality to coexist.

Noxious NOx

"Current diesel engines have excellent fuel economy, and further improvements with or without electrification are possible," Kufferath told us. Data from a network of urban emissions monitoring stations like the one at Stuttgart Neckartor has shown that diesel passenger cars only contribute six percent of PM10 particulate pollution, thanks to the older Euro 5 emissions regulations and the filters that it mandated. Likewise, unburned hydrocarbon and CO emissions are handled by technologies like lean combustion and oxidation catalysts in cars with functioning emissions control systems. The real problem—and the one that got VW in trouble—involves nitrogen oxides, or NOx.

For NOx, there has been quite the mismatch between the amount that actually comes out of a diesel-powered car's exhaust on the road versus the levels emitted in the lab during type approval—and not just on diesels with cheat code, either. For example, while the older Euro 3 regs limited NOx to 500mg/km, those cars spewed out twice that much in the real world.

Euro 4 dropped the limit to 250mg/km, but real-world testing showed cars were still pumping out 800mg/kg, a figure that didn't change with the introduction of Euro 5 and its 180mg/km limit. This discrepancy is caused by loopholes in the regulations that relax the emissions standards during real-world driving, allowing manufacturers to turn emissions controls off while a diesel is driving in "cold" weather, for example.

But Euro 6 regulations are even more stringent at 80mg/km NOx, and soon all new diesel cars in Europe will have to prove they can actually achieve that out on public roads.

Real-world conditions

The RDE test is encouragingly thorough. For one thing, it has to include an equal mix of urban roads at speeds under 37mph (60km/h), rural roads at speeds from 37mph to 56mph (60km/h to 90km/h), and highways at speeds from 56mph to 90mph (90km/h to 145km/h). Each of those sections has to be at least 10 miles (16km), and they can't be flat—there should be up to 3,937 feet/62 miles (1200m/100km) of incline, for engines work much harder when climbing hills. The tests have to be run in all weather, from 19 degrees Fahrenheit (-7 degrees Celsius) to 95 degrees Fahrenheit (35 degrees Celsius), beginning with a cold start of the engine. Vehicle dynamics are also important; test drivers have to run the gamut of driving styles from ultra smooth all the way to lead-footed boy racer.

Bosch's RDE-complaint test route takes about 110 minutes to complete, starting from a Bosch facility in Feuerbach, running through Stuttgart, then looping around Sindelfingen and through Leonberg before finishing 58 miles (93km) later. On top of that, the company has a second, urban-only test route through Stuttgart. Test cars are fitted with a portable emissions-monitoring station that constantly samples how much NOx is coming out of the tailpipe, as well as recording ambient weather conditions, and the vehicles should be close to their maximum allowed weight. With that hefty PEMS unit hanging off the back, the vehicles are close to their maximum allowed weight. (The PEMS units are defined by the European regulations.)

We asked Kufferath if cheating through some form of geofencing would theoretically be possible. He told us that, since no actual fixed test route exists, there shouldn't be any way to optimize the test—beyond incorporating the right technology into the vehicle to allow it to pass. "Also, we tell the engineers that since they know the weak points of the system, they should go out and try and get the worst result possible," he said.

Beating 80mg/km is possible, even if you drive like me

The key to meeting the test is to combine engine optimization with exhaust gas treatment (selective catalytic reduction and NOx storage catalysts).

"There is no single component solution; it needs to be a well-balanced system," Kufferath explained. This system includes a 31,908psi (2,200bar) direct-injection system and "digital rate shaping" of the injection system (which divides the process up into many tiny injections of fuel to create a more gentle combustion process). Exhaust gas recirculation is used across almost the entire engine map. The engine's ECU constantly measures the temperature of the catalyst to prevent it getting too cool to work properly, tweaking the injection system to increase exhaust gas temperatures when necessary.