What VW most probably did to tweak their emission results

22 09 2015

It might become the end of an era, or might be yet another scandal. At the very least, it is very bad for the industry. VW has been caught cheating and they quickly (and wisely) admitted it. But what is the science behind this? 

The scandal is about diesel engines. Alongside CO2, water, oxygen and atmospheric nitrogen, the tailpipe of a diesel car also emits unburned hydrocarbon, CO (small amount but regulated), NOx and soot. Actually, there is a bunch of other gases that are not yet regulated but that would drive us too far for this article. 

For diesel engines, not only do you expect a low CO2 level but at the same time low soot and NOx. The CO2 emission is directly linked to the fuel consumption. Since these engines have a lower fuel consumption, they also have lower CO2 emissions. The standard way to limit the emissions of soot is to place a filter in the exhaust. This slightly affects the efficiency but reduces significantly the tailpipe emissions of soot. To avoid clogging, it is "regenerated" every now and then. The engine control unit (ECU) modify the injection settings so that hotter exhaust gases burns the soot away. The way VW and many other car manufacturers reduces the emissions of NOx in most passenger cars is to use exhaust gas recirculation. NOx is produced when very high temperature, nitrogen and oxygen are combined. The primary effect of doing this is to modify the specific heat of the mixture and reach a lower top temperature. To go much below the level of temperature where acceptable emissions are produced, EGR is not enough, you need to change technology. A second effect is on the oxygen content, by replacing air in the cylinder with exhaust gases that are recirculated back into the engine. In addition to this and to primary measures to change the combustion process (changing the injection, turbo, ...), secondary measures have been developed such as urea injection and NOx trapping. 

When talking about emissions, the word "trade-off" should be kept in mind. Even though, the strict norms have driven the industry towards very low emissions, everybody knows there is no free lunch. For example, going too far with the EGR would increase the amount of soot produced which might affect the filter, the EGR valve,... At the same time, trying to filter too much of the soot, would lead to dramatically increased fuel consumption. But something we did not yet mention is also very important: the driver experience. When using EGR, the torque available to the driver is significantly reduced. What do you think happens when a customer who just bought the latest fancy clean diesel car finds it less responsive that the previous he had? Well, he or she brings it back to the dealership and complains. The exact same thing happens when you have issues due to increased soot emissions or when the cost of secondary measures are excessive.  That's where VW has probably cheated. To avoid such kind of problems, they have programmed the ECU to recognise when the car is on a standard cycle. Since the conditions of the cycle are known in advance, it is rather straightforward to identify when the engine should behave and produce low emissions. Outside of these conditions, the ECU would reduce significantly the amount of EGR used (or secondary measures), which emits more NOx but avoid other problems. Since you don't see NOx, no problem... until someone thinks about measuring it on road.

That's exactly where I wanted to come. The new standard that will be implemented soon will require the real world emissions of the cars using PEMS (portable emission measuring systems). There is still a lot of debate on how do to it, but also of a lot of interesting research.

Francesco Contino 

Update on the link with CO2 thanks to my colleague Prof. Laurent Bricteux from UMons.
Updates on the secondary measures and many other aspects from my colleague Prof. Hervé Jeanmart from UCL.