The German trade magazine Auto Bild reports that road tests conducted by the International Council on Clean Transportation (ICCT) found the diesel BMW X3 20d Xdrive guilty of emitting more than eleven times more NOx gases than the limits of Euro 6, Europe’s latest diesel engine emission legislations. Oh boy.
If true, this proves that the dieselgate disaster is far from being a Volkswagen-specific issue. While BMW X5 passed its test in California done by the same people who found out about about Volkswagen’s cheat software, the ICCT suggested before that BMW, Opel and Mercedes-Benz can be in trouble as well.
The data Auto Bild is going to print tomorrow shows that while the U.S. market Volkswagen Jetta 2.0 TDI exceeds Euro 6 NOx levels by a whopping 22.61 times, the Euro-spec BMW X3 20d Xdrive performs worse than the American VW Passat 2.0 TDI, exceeding the target figure by almost 12 times.
The relevant Euro 6 standard for diesel emissions is 0.08g/km (or 0.12g/mi). The EPA’s current relevant standard (Tier 2, Bin5) is 0.05g/mi (or 0.03g/km). That makes European standard about two and a half times as lax as the American one, although it must be noted they run different test cycles.
The X3 20d’s 2.0-liter turbodiesel four-cylinder engine is also sold on several U.S. BMW models, including the X3 and 3 Series, but whether they might be in violation here is unclear.
So far, BMW issued a statement saying that their cars are not equipped with a cheat software, while Daimler says they work together with all the relevant authorities and disavow manipulative accusations by any organizations.
A European engine expert who we shall just refer to as Mr. Camel wrote in to explain what’s going on here:
Today that the German government announced that this VW issue effects many of European diesel engines too. They’ve identified the 1.6 and 2.0 diesels, but it is just a matter of time before they realise that a different algorithm is limiting the urea flow in the 3.0 diesel. The ICCT had data from Emissions Analytics from PEMs testing of an Audi A8 with the 3.0 diesel. This was x22 times over the 80mg/km NOx limit.
Yesterday the EU voted in an amendment to the EU6 legislation, to enforce the Real Driving Emissions tests on passenger cars from 2017. This was voted on in the Environmental council, and apparently supersedes the authority of the body that was supposed to handle this matter. The car industry (predominantly VW) had been pushing for a deviation factor of x3-x5 times the legal limits for “in-use” testing. Yesterday’s decision will bring the deviation factor to X1.5 times the emission limits in the test cycle. This will be difficult for vehicles with lean NOx traps, or SCRs with small urea tanks.
The reason why being eleven times over the EU emissions is mostly ok is a little complicated.
Basically, different speed load profile in the EU cycle to the US cycle - US cycle is very tough, and runs the engines up to full load in most cases. EU cycle is more like 50% load and max 2200rpm.
Warm up profile of engines from cold start (25degC ambient temp) means that the engine will behave differently outside the specific EU cycle and at different ambient temps. This is called cycle beating but it is currently legal. Most PEMs measurements are taken from a hot engine - this is not the case during the drive cycle. Engines are optimized to lower NOx emissions during the warmup phase, and the hot fast section is at higher engine loads.
The defeat devices that VW are using turn off the emissions controls when the steering wheel is being used.
There is a ‘real world’ measurement using PEMs by Emissions Analytics of a EU6 compliant Audi A8 with the 3.0 diesel. This was x22 times over the EU limit of NOx, hence it will have a defeat device restricting the urea flow when not driven on a chassis dynamometer.
In summary, BMW is just playing by the rules to be competitive. VW is cheating.
Okay, so here’s my idea for a solid test they should all use:
Tester gets in car 20 miles from a big city.
Tester starts up the car and drives it to the big city’s center immediately, without warming up.
Tester parks the car, and we have the figures. Done.
Contact the author at firstname.lastname@example.org.